For years, FORMEC has brought together thought leaders from the forestry mechanisation and engineering community to exchange cutting-edge research, practical experience, and forward-looking ideas. Next year, the conference arrives in Prague, where we will, in line with the motto of “Enabling resilient forest value chains through innovation,” explore how new technologies, smart operations, and adaptive strategies strengthen the resilience and sustainability of forest-based industries worldwide.
The identification of standing tree species, forest timber assortments, and on-site as-sessment of their quality and value using images would be widely useful for a variety of applications. In this study, we used the Orange data mining software to classify images of forest wood assortments by tree species and value. The software applies deep learning algorithms for image classification. Our results show high accuracy when classifying assortments by tree species, but lower accuracy when classifying assortments by market value. Classification accuracy for spruce assortment images by market value was only 39.1%. We conclude that further exploration of approaches to classify assortments by value is needed, as this task has the greatest practical significance but currently yields the least reliable results.
The current development of harvesters and harvesting systems makes it possible to calculate timber volume with a high accuracy, albeit over bark. Despite this, countries where timber is traded based on under-bark measurements, such as Poland, continue to measure volume manually or using photo-optical methods. Meanwhile in Finland, where over-bark measurement is in use, the volume calculated by harvester software can be used in trading. In the end, it is mainly bark or its absence that causes variability in timber volume calculation.
During processing, a certain amount of bark is reduced by the harvester knives, which, as a consequence, artificially lowers the measured diameter and calculated volume. Very careful calibration and accurate measurement of diameter and length does not guarantee that the log volume obtained represents the real volume. This is due to many factors, including season, e.g. in the spring, bark is easily removed or shaved with the harvester head knives, and the system is not able to recognise if the diameter was measured with bark or without bark.
The objective of this research was to find out how much bark is removed during log processing of pine to construct new model of bark thickness that needs to be subtracted when calculating log volume. Initial measurements are being caried out in Winter 2026, where bark thickness has been measured along logs after delimbing, with particular attention to thickness of bark partially removed (under knives) and thickness of natural bark (not shaved). Detailed methods, early data analysis and first results will be presented at the conference in the form of new subtraction of bark thickness for pine, to get close to real timber volume from harvester software.
The accurate measurement of individual, high-quality logs is a critical operational task in the timber trade, particularly during timber auctions where roundwood is traded individually and commands premium prices per cubic meter. In these high-stakes environments, even minor deviations in volume estimation can result in substantial economic losses for either buyers or sellers. Consequently, ensuring accuracy, transparency, traceability, and reproducibility is essential. Currently, log volume estimation predominantly relies on manual measurements using conventional stereometric formulae, such as those of Huber, Smalian, or Newton, in accordance with frameworks like the German Agreement for the raw wood trade. However, these traditional methods are constrained by simplified geometric assumptions, human error, tool calibration issues, and forestry rounding rules. Recent technological advancements in artificial intelligence , image-based methods, and three-dimensional (3D) reconstruction offer a promising pathway to overcome these operational bottlenecks. While small-scale devices equipped with Light Detection and Ranging (LiDAR) or Red Green and Blue (RGB) sensors have demonstrated potential, the integration of these technologies into fully automated, end-to-end workflows for individual log identification and volume estimation remains largely unexplored in operational practice. To address this gap, this study proposes an automated methodological framework that synergizes two-dimensional (2D) instance segmentation based on RGB information with 3D photogrammetric reconstruction, allowing log volumes to be estimated directly from precise geometric models. The proposed workflow comprises four primary stages. First, RGB imagery is acquired using consumer-grade mobile devices (specifically an iPhone 15 Pro and an iPad Pro mounted on a gimbal) providing a flexible, accessible data collection method suitable for auction yards. Second, instance segmentation of individual logs is performed using a YOLOv8-Seg model based on two-dimensional Convolutional Neural Networks (2D CNNs). This strategic use of 2D CNNs robustly isolates logs from complex backgrounds while bypassing the computational heavy lifting and extensive annotation requirements associated with fully 3D deep-learning architectures. Third, the segmented image data undergoes photogrammetric processing to generate detailed 3D virtual reconstructions of each log. Finally, the system automatically calculates the individual log volumes derived directly from the reconstructed 3D geometry. By closely integrating instance segmentation and photogrammetry, this pipeline transitions raw field imagery into highly accurate volume metrics with minimal manual intervention. The practical implications are significant across the forest-wood value chain. For forest owners and auction organizers, it ensures fair valuation and compliance with standardized trading frameworks through transparent, digital documentation. For timber buyers, it mitigates uncertainty and measurement disputes. Ultimately, this methodology reduces operator subjectivity, enhances measurement accuracy, and lays a robust foundation for the future integration of automated digital twins in forest operations and high-value timber trading.
Fungal sapwood discolouration is a major cause of value loss in Norway spruce sawlogs and directly affects grading outcomes according to EN 1927-1. Spores of fungi responsible for sap-stain are naturally widespread in forest ecosystems and spread via insects, wind, rain splash, and contaminated harvesting equipment, increasing the risk of quality loss during log storage.
This study evaluates a supervised deep learning approach for automated detection and quantification of sapwood discolouration in roundwood cross-sections. A dataset comprising 928 images of Norway spruce (Picea abies (L.) Karst.) was acquired. Stem discs were placed in a measuring frame and photographed using a digital camera (APS-C format) with a fixed 30 mm focal length lens, mounted on a tripod under varying natural lighting conditions typical of wood yard environments. Discoloured regions were delineated, and the masks distinguished heartwood, non discoloured sapwood, and discoloured sapwood. These masks served as ground truth labels for model training. Images and masks were converted to image resolutions of 128×128, 256×256, and 512×512 pixels, split into training (70%), test (20%), and validation (10%) subsets and augmented by rotations and flipping.
A U-Net was trained for supervised pixel-wise semantic segmentation (discoloured vs. non-discoloured wood). The encoder of the U-Net used convolutional blocks with 16-64 initial filters depending on the image resolution, and downsampling via max pooling. The decoder mirrored the encoder with upsampling and skip connections and a final 1×1 convolution with sigmoid activation produced binary probability masks. Training used the Adam optimizer, and the loss combined binary cross entropy and Dice. Evaluation metrics were the Dice coefficient and the Jaccard index (IoU). Models were trained using a batch size of 30.
Across model configurations Dice coefficients ranged from 0.871 to 0.908 and Jaccard indices from 0.783 to 0.838 on the independent test dataset, demonstrating robust segmentation performance despite heterogeneous illumination and non-standardized camera-to-object distances.
The results indicate that supervised CNN-based image segmentation enables objective, reproducible, and spatially explicit quantification of sapwood discolouration in roundwood. From a forest supply chain management perspective, such methods represent a step toward sensor-assisted quality assessment in harvesting and storage logistics. Integration into mobile camera systems, log-yard scanning setups, or future harvester- or loader-mounted imaging systems could support continuous quality monitoring along the supply chain. This would enable earlier detection of degradation processes, improved storage strategy decisions, and more data-driven timber allocation — contributing to the digitalization and increased transparency of mechanized wood supply systems.
Reliable, high‑throughput evaluation of internal wood quality is a critical enabler for modern mechanized forestry and downstream processing. We report a compact set of results from an imaging‑based workflow for automated analysis of roundwood, focusing on (i) robust separation of bark from solid wood and (ii) consistent detection of slices containing internal defects such as knots, cracks, or degraded tissue.
Data and approach: We trained a three‑class semantic segmentation model on annotated log images with voxel dimensions of 1 × 1 × 10 mm. The training set was deliberately enriched with difficult cases, especially slices with discontinuous or partially peeled bark. To stabilize edge decisions with minimal computational overhead, we added a one‑step adaptive relabeling rule that converts specific high‑intensity peripheral pixels—frequently misinterpreted as background—into the bark class. For defect presence/absence, we introduced a composite multi‑layer representation by merging three consecutive slices into a single input image, thus injecting short‑range spatial context without resorting to full volumetric modeling.
Results — bark segmentation:. On a 50‑image validation subset, the Sørensen–Dice score for bark reached 0.704 (wood 0.981, periphery 0.996). Applying the single adaptive relabeling step increased bark Dice to 0.750, while maintaining wood at 0.981 and periphery at 0.999. The improvement held for both slices with defects and defect‑free slices, indicating that the rule targets a systematic boundary issue rather than overfitting to specific textures.
Results — defect detection: Single‑slice classification achieved 93.94% overall accuracy across logs. Introducing the three‑slice composite raised accuracy to 95.39% and increased the worst‑case per‑log performance above 92%, evidencing greater robustness to inter‑log variability. When the inter‑slice distance in the composite was increased to 2 cm, the accuracy reached 94.72%, suggesting diminishing returns once the composite exceeds the typical continuity scale of internal features. Class‑wise diagnostics further showed that pronounced dark regions associated with unhealthy knots, decay, and major cracks were detected with near‑certain confidence, whereas subtle, healthy‑knot transitions remained the primary challenge.
Implications for technology and operations: Three pragmatic interventions—targeted inclusion of difficult bark cases, a single adaptive post‑processing rule, and composite multi‑slice inputs—deliver measurable gains in segmentation reliability and defect detection while keeping the computational and operational footprint low. Such lightweight improvements can be integrated into mechanized inspection chains to support more dependable sorting, planning, and value recovery decisions in forest‑based value chains.
Intensive forestry operations can lead to soil compaction, plastic soil disturbances, and rutting, which negatively impact soils, vegetation, and water bodies. Although many studies have explored the main factors influencing soil damage, it remains unclear whether wood extraction methods and driving direction affect the impacts of forest machinery. Specifically, the differences in soil damage between skidding (SKI) and forwarding (FOR) are not well understood, especially regarding the extent of soil compaction beyond wheel ruts and the influence of driving in curves or in straight lines. This study aimed to compare soil bulk density (BD), soil penetration resistance (PR), and rutting when driving in curves and on straight lines, while considering the total area of soil impacted by wheels and load. The same forest machine (a forwarder) was used during both FOR and SKI to keep machine parameters constant; in FOR, the machine transported logs as an ordinary forwarder, while in SKI, long trunks were fixed with chains to simulate a clambunk. Soil compaction was evaluated using BD and PR in the area impacted by wheel ruts, between the ruts, and by the load outside the wheel ruts. Rutting was quantified using drone imagery and Structure-from-Motion photogrammetry (SfM) to create 3D soil surface reconstruction, allowing for measurement of rut depth and soil volume changes.
In general, both BD and PR increased significantly after 20–25 machine passages on this sandy soil. Across treatments, PR increased by 49-120% within ruts and by 24-71% between ruts, indicating lateral propagation of compaction and a widening of the impacted zone beyond the wheel tracks. When comparing driving straight ahead with driving in a curve at the same transported wood volume (20 FOR and 25 SKI passes), PR within ruts was consistently higher in a curve for both extraction methods (FOR: 3.22 to 4.34 MPa; SKI: 2.82 to 3.85 MPa), and PR between ruts showed the same overall pattern (FOR: 2.45 to 3.07 MPa; SKI: 2.64 to 3.23 MPa). Photogrammetry-derived 3D soil reconstruction further showed that the area affected by the wheel ruts increased in curved segments for both FOR and SKI, and that the total area affected was larger in SKI due to the additional disturbance caused by dragged trunks, resulting in a wider footprint when driving in a curve. Rut depth increased significantly with traffic intensity in FOR; in contrast, in SKI, rut depth estimates were affected by rut concealment, thereby reducing measurement reliability. These results show important information on soil impact, useful for the development of a new wood extraction technique. 3D reconstruction measures soil volume changes and rut depths better than manual measurements.
Ground-based forest harvesting operations can substantially disturb soil through mechanisms such as compaction and rutting, posing long-term risks to forest productivity and sustainability. In boreal forest systems, soil trafficability maps derived from the depth-to-water (DTW) index are commonly used to identify areas vulnerable to machine-induced soil damage. However, the applicability and reliability of this approach in non-boreal forestry environments have not been sufficiently evaluated. The objective of this study was to assess the performance of the DTW index as a predictor of soil trafficability in plantation forests outside the boreal context
To address this objective, we examined DTW-based soil trafficability maps across three Pinus radiata plantation forests in New Zealand. Skid trails were selected within areas classified as potentially sensitive (DTW ≤ 1 m) and less sensitive (DTW > 1 m). Field measurements included soil moisture content, rut depth, and indicators of soil compaction, specifically bulk density, penetration resistance, and shear resistance. The data were analysed using linear mixed-effects models to account for site-level variability.
The results indicated no significant relationship between the DTW index and soil moisture at the time of sampling. This lack of correlation is likely attributable to temporal differences between harvesting activities and soil measurements, reflecting seasonal variability in soil moisture conditions. Despite this, rutting was consistently more severe in areas with DTW values ≤ 1 m, demonstrating that the DTW index is effective in identifying zones susceptible to surface deformation caused by machine traffic
Overall, the findings indicate that DTW-based trafficability maps are useful for highlighting areas where machine traffic is likely to exacerbate soil disturbance, but they should not be interpreted as defining zones that are entirely resistant or “safe” from impact. While DTW mapping represents a valuable planning tool for reducing logging-related soil damage in New Zealand plantation forests, its predictive capability could be enhanced by incorporating real-time soil moisture information and developing regionally specific flow initiation area (FIA) parameters. Future research should focus on refining DTW-based models to improve their robustness and applicability across a broader range of forest types, climates, and operational conditions.
Rut formation, entailing both a disturbance of soil functions and impaired machine mobility, is a common risk when operating ground-based forest machines under unfavourable soil conditions. As a complement to technical approaches such as flotation tires or bogie tracks as well as planning support tools such as trafficability maps, we present a concept of trafficability prediction for a forwarder building upon measured wheel slip of the preceding harvester. In cut-to-length operations utilizing a wheeled harvester, this machine typically has lower wheel loads than the loaded forwarder which also tends to require a higher number of passes over the machine operating trail. Thus, forwarder traffic commonly contributes more to rut formation than harvester traffic. In our approach, wheel slip data of the harvester combined with accurate position tracing shall be exploited in a prediction model aimed at avoiding severe impact of forwarder traffic on forest soils at locations with sensitive soil conditions. During an initial experiment, a harvester and a forwarder working in sequence were consecutively equipped with a set of sensors for real-time wheel-slip measurement and positioning, and rut depth was measured after machine traffic using a mobile laser scanner. As a work-in-progress report from an ongoing project, the sensor equipment and data processing methods of the concept are outlined, and results from the initial experiment are presented.
The increasing frequency of severe windstorms in Central European forests has led to extensive salvage logging operations that can have lasting effects on the integrity of the soil and related pedological and biological functioning.
This study quantified the extent and degree of soil disturbance following fully mechanized salvage logging in the University Forest Enterprise near Brno, Czechia, after a windstorm on 21 June 2024. It further explores changes in fine root biomass, root tip density, and mycorrhizal abundances. Five 50 × 50 m experimental plots were established in mature, broadleaf-dominated stands affected by the storm. Salvage logging was conducted in a cut-to-length fashion, and the resulting extent of machine traffic was mapped using high-precision GNSS.
The total disturbed area averaged 28 ± 3% of the plots when the entire machine operating trail was considered, and 22 ± 4% when only direct machine–soil contact was included. High-resolution digital elevation models derived from UAV imagery were used to estimate rut depths along 100 randomly distributed transects. Mean rut depth across sites was moderate (0.17 ± 0.11 m), with maximum values reaching 55 cm. Soil bulk density increased by 16% on average (from 1.22 ± 0.22 to 1.38 ± 0.25 g/cm³) in trafficked areas, with site- and depth-specific variations. Besides physical changes, machine traffic also affected biological conditions. Fine root biomass declined by 30%, and substantial reductions were observed for root tip density and mycorrhizal abundances.
Overall, results indicate average degrees of soil disturbance but high related biological impacts, with a striking areal extent—heavily exceeding existing standards and protocols designed for soil protection measures. This study emphasizes the rigorous maintenance of a permanent machine operating trail system, s, even under the unfavorable conditions of salvage logging operations.
For given wheel loads, the ground pressure exerted by off-road vehicles such as forwarders is to a large extent influenced by tire width and tire inflation pressure. Thus, wide flotation tires, but also flotation bogie tracks, are well-established means of reducing soil disturbance and maintaining machine mobility in forest operations. While reduced tire inflation pressures applicable in forest operations are still higher than those common for agricultural tires during field traffic, there is some development towards lower allowable inflation pressures for forestry tires as an alternative to flotation tracks in certain applications. In this context, the distribution of ground pressure under a driving forwarder was measured with two types of low-pressure tires with nominal section widths of 0.94 m and 0.75 m and with inflation pressures of 1.2 bar and 2.5 bar, as well as with standard tires (nominal section width: 0.71 m; inflation pressure: 5.5 bar) and with flotation bogie tracks (track-plate width: 0.92 m) installed on standard tires. Foil sensors with a spatial resolution of 9.7 mm were used for measurement, and measurements were made for the rear wheels of an eight-wheel forwarder which had wheel loads of 4.5 metric tons. Observed mean and peak ground pressures followed the expected trend and decreased with increasing contact width and decreasing inflation pressure. Thus, observations were consistent not only with previous findings from static measurements of ground-pressure distribution but also with results from field experiments.
Mechanized cut‑to‑length harvest operations with clear cutting as the final operation within the rotation is the predominant harvest method in Fennoscandian forestry. However, interest in continuous-cover-forestry (CCF) methods is increasing. CCF methods such as shelterwoods and patch cuttings can contribute to maintain stand continuity and can provide ecological and social benefits but may also introduce greater operational complexity. Forwarder work in CCF methods is likely to differ from conventional clear cutting due to lower concentration of harvested wood volumes along strip roads and longer extraction distances, but research on the topic is scarce. Therefore, the aim of this study was to determine if forwarder productivity and fuel consumption in fully mechanized shelterwood establishment or checkerboard patch cutting is more time or fuel consuming than in clear cutting, and if so, quantify how much.
The study was conducted on seven harvesting sites in central Sweden, in a block design. Each block contained three treatments: final felling, shelterwood establishment, and patch cutting, covering in total, 9–29 ha, with individual treatment areas ranging from 2.1 to 13.1 ha per block. The patch cuttings were done in a checkerboard pattern over the entire treatment area, in 33 by 33-meters square patches.
Data was collected with the machine control systems as StanForD2010 reports and were collected for each block and treatment area. The reports contained worktime, operator, fuel consumption, driven distance, and forwarded volume. Statistical analysis were done on the work shifts with mixed linear models to predict how time or fuel was explained by meters driven per extracted solid cubic meters under bark (m3sub), treatment and the random effects of operator and block.
Forwarder work in both shelterwood establishment and checkerboard patch cutting required significantly more time and fuel than clear cutting on the same blocks. As a result of different work methods, the mean extraction distance per m3sub in shelterwood establishment was 67 m and not significantly different to the 69 meters per m3sub in clear cutting. In patch cutting, on the other hand, the mean extraction distance was significantly longer (95 meters per m3sub). Based on those mean extraction distances, the patch cutting will consume 26% more time and 27% more fuel while shelterwood establishment will consume 10% more time and 2% more fuel compared in both cases to clear cutting. However, if the methods would be possible to implement with equal driving distances, the study indicates that there would be an 11% increase in forwarding time for both CCF treatments, whereas the fuel consumption would be 2.5% higher in patch cutting and 1.3% higher in shelterwood establishment compared to clear cutting.
These results will be useful when planning shelterwood and patch cutting management, by enabling the differences in fuel and time consumption and, ultimately, costs can be accounted for. Further research could complement the results by further investigating various types of shelterwood removal, patch cuttings of other shapes and sizes, and the total system economy of the management and operations.
Forest harvesting operations inevitably affect forest ecosystems, and in order to apply Sustainable Forest Operation a careful planning is essential mainly with high level of mechanization employed. This case study was developed within the framework of the AGRITECH project in one of the first living lab activated. The study area is located in Central Italy in a Turkey oak forest managed as coppice with standards. The area was divided into two plots: an experimental plot, where operations were planned using precision forestry techniques, and a control plot, harvested without preliminary planning, within each plot three replicates were identified. The same harvesting activities were carried out in both plots, CTL logging system, with felling and processing by chainsaw and extraction by forwarder. This research evaluates the feasibility of applying precision forestry to the planning and execution of logging operations in an area typically characterized by rugged terrain and limited permanent extraction roads or tracks. After field data collection, the two treatments were compared using soil surface impacted (percentage of soil affected by the passage of the machinery), working times and productivities and related GHGs emissions and costs. Overall, precision forestry applied during both the planning and operational phases reduced the total impacted surface area by more than 50%, while slightly decreasing working productivity, while there has been only a slight improvement in terms of reduced emissions and costs. These results highlight the good potential of precision forestry to balance productivity, safety, and environmental protection. At the same time, they point to current limitations that, if addressed in the near future, could significantly support the sustainable development of the Italian forestry sector.
Boom-corridor thinning (BCT) has been proposed as a promising alternative for treatment of dense stands with small-diameter trees, as it promote higher harvester productivity due to effective crane work in comparison to traditional selective thinning (ST). BCT can be performed using high to low degree of systematic approach, from strict (following a strict pattern) to flexible (adjusting the pattern to e.g. stand structures). BCT could also be combined with selective thinning of un-thinned zones in-between boom-corridors (BCTcombo), but which has yet not been studied. The objective was therefore study and compare harvester productivity in BCTcombo and ST in young dense first thinning stands.
The study was performed in North Karelia, Finland in two-storied birch stands. Time studies were executed in 18 study units, to which treatments (BCTcombo and ST) were randomly assigned. The stands were of 8-11 m in height with a mean stem density of 9877 stems/ha (range 2450-16100) of which, on average, 1806 stems/ha (20%) were > 6cm DBH and 80% were below 6 cm DBH, classified as undergrowth. The instructions for the operator were to perform quality thinning and target ca 1500 future crop trees. The harvester used was a Komatsu 901 featured with a 11 m crane and a multi-tree handling harvester head. The trees were harvested at full length or bucked into sections (i.e. tree parts) and bunched along the strip-roads.
Based on the harvester computer data, the harvesting removal for treatment BCTcombo was on average 65 m3/ha and 3333 stems/ha, which was 18% and 13% higher, respectively, compared to ST. The average tree size cut was however 20.5 dm for both treatments. The observed harvester productivity was on average 8.5 m3/PMh for BCTcombo, which was 14% higher compared to ST. Preliminary results from this study aligns well with previous research, showing that that BCT facilitate higher harvester efficiency.
In Mediterranean forestry, thinning of pine forests is mainly done with traditional harvesting system (THS) due to its versatility and low investment and relocation costs compared to specialized machinery. However, THS is based on winch coupled to farm tractor with a large share of manual work for felling and processing tasks. The increasing lack of workforce for these activities may be solved with the introduction of highly mechanized systems. In this context, light multipurpose machinery harvesting system (LMM) may be a valuable alternative, providing a fully mechanized solution with an affordable initial investment and limited relocation costs. To assess its feasibility the performance of a LMM constituted by a unique harvester-forwarder combi machine, has been compared to the THS. The results show highest overall productivities of THS (2.87 m3/PMH15) compared to LMM (2.3 m3/PMH15). Yet, due to the higher hourly cost of the traditional system overall harvesting costs are lower in the LMM (38.8 €/m3) than in THS (39.7 €/m3). Additionally, LMM provides a larger number of timber assortments, leading to a higher value recovery with a difference of 4.5 €/m3 compared to THS. This provides an economic balance per m3 over threefold higher in LMM. Concerning damage assessment, the THS caused higher number of injured trees (8.5%) compared to LMM (2.8%), yet in both cases overall damages to the remaining stand are limited.
This study analyzes and compares the productivity and economic efficiency of pull-through delimbers and processors in the processing of whole trees from partially mechanized timber harvesting in Baden-Württemberg, Germany. In forest-rich and topographically challenging regions such as the Black Forest, fully mechanized harvesting systems often reach operational limits. As a result, adapted semi-mechanized approaches are frequently applied, typically relying on the pre-concentration of timber. Within such systems, the pull-through delimber represents a practical alternative to processor-based processing. However, its performance and operational potential have so far received little systematic investigation.
The study combines qualitative and quantitative methods to evaluate operational workflows and machine performance. Expert interviews were conducted following the Grounded Theory approach to gain insights into practical applications, operational constraints, and decision-making processes. In addition, time studies were carried out to determine work productivity and cost parameters. These data were used to compare the pull-through delimber system with the processor-based processing system regarding productivity, cost efficiency, and operational suitability.
The results show that the processor-based system achieved higher and more consistent performance in terms of pure processing productivity. In contrast, the pull-through delimber system demonstrated considerable operational flexibility and adaptability. The pull-through delimber proved to be suitable for small to medium scale operations, under steep terrain conditions, structurally complex stands, and the production of treelength assortments. Furthermore, it was capable of handling larger stem diameters and could be deployed rapidly during salvage logging operations following disturbance events.
From an economic perspective, the analysis indicates that the pull-through delimber system can achieve competitive or even lower costs per cubic meter under favorable operating conditions. Cost advantages arise primarily from the machine’s high mobility, lower fixed cost risks during idle periods, and the potential to generate additional revenue from chipped biomass. Considering the established markets for treelength and wood-chip products in southern Germany, the results highlight a significant operational potential for pull-through delimber systems in semi-mechanized harvesting systems.
Located in the heart of the Western Carpathian Mountains, Slovakia has a high share of its forest land in mountainous or hilly terrain. In this morphologically and geologically difficult terrain, an adequate harvesting system is required to operate efficiently, prevent environmental degradation and secure forest workers' safety. A large proportion of machinery used in harvesting operations in Slovakia is relatively old. The continuous and still predominant use of tractors in wood extraction causes irreversible erosion, especially on steep and erosion-sensitive flysch slopes, and is also responsible for many accidents. In recent years, their share has been decreasing in favour of more modern and environmentally friendly harvesting systems such as harvesters and forwarders, as well as tower yarders. To support the expansion of innovative and sustainable harvesting systems in Slovakia, this study aims to assess their spatial extent of technical applicability and current application potential in Slovak forests.
The performance of seven state-of-the-art harvesting systems was evaluated for each forest management unit (JPRL) using Multi-Attribute Utility Theory (MAUT) in a spatial GIS-based framework. Terrain-related attributes were derived from a LiDAR-based digital terrain model of Slovakia with a resolution of 1 × 1 m. Other stand-specific data were obtained from forest management plans (PSL). First, the technical compatibility of each harvesting system was modelled based on limiting factors such as slope inclination, obstacle abundance, and skidding distance. However, technical feasibility alone does not ensure sustainable forest operations. To address trade-offs between environmental impacts, worker safety, and economic viability, a MAUT framework was applied. The normalisation of criteria indicator values to a 0–1 dimensionless scale allowed diverse criteria, such as soil disturbance, injury rates, exposure to workplace sickness, and system rentability, to be combined. This resulted in a single evaluation score for each alternative, enabling the suitability comparison of technically feasible harvesting systems within each forest management unit. Scenario-based analyses examined how changes in the relative importance of objectives affect harvesting system modeled suitability and potential application.
The resulting technological compatibility maps of harvesting systems show where each system can be applied and where terrain conditions prevent their use. Complementarily, the harvesting system suitability maps based on sustainability criteria reflect the current potential application of innovative harvesting systems, taking into account both current stand conditions and the relative importance of sustainable forest management objectives for forest managers. The study provides spatially explicit guidance for sustainable harvesting, contributing to the National Forest Programme 2025–2030. Forest district managers and harvesting machinery companies can benefit from this information.
Forest management faces increasing uncertainty due to climate-driven changes in forest dynamics, disturbances and growing economic constraints on timber harvesting, particularly in mountainous terrain and areas with limited forest accessibility.
For policy implementation and long-term planning, both in government and the timber industry, it is essential to have estimates about the future provision of ecosystem services (such as timber supply or protective service), because harvesting costs strongly determine economic viability of forest operations/forest management.
Most studies to date neglect or greatly simplify operational factors (e.g. by assuming constant net revenues or constant harvesting costs under all conditions). We present a framework that accounts for harvesting costs that vary with harvesting characteristics and terrain properties, supporting long- term forest growth and management modelling. It dynamically assesses best suitable timber harvesting methods (dynBEST) and estimates associated costs under future forest trajectories.
The framework can be integrated modularly with many forest growth and management models to support forest planning.
We will show an example in which the framework is applied to the full Swiss National Forest Inventory network, covering with more than 6,000 plots a broad gradient of biogeographic regions and topographic conditions: We integrate (i) climate-sensitive, long-term forest simulations with the forest model MASSIMO under alternative management strategies, (ii) assessments of technical feasibility of harvesting options, and (iii) productivity-based modelling of harvesting costs over a simulation period of 100 years.
Selecting the most appropriate harvesting system is a critical decision in forest operations. In practice, this choice is often based primarily on the experience of forest technicians or logging contractors. Increasingly, however, there is a need for more objective and transparent approaches that can support decisions in complex operational contexts. In recent years, the integration of Geographic Information Systems (GIS) and Multi-Criteria Decision Analysis (MCDA) has emerged as a promising framework for harvesting system selection. This decision is inherently multi-criteria, because operational, topographic, pedological, infrastructural, and stand-related factors all influence system suitability to varying degrees. At the same time, advances in remote sensing and the growing availability of high-resolution spatial datasets, such as Digital Terrain Models (DTM), Digital Surface Models (DSM), and soil trafficability maps, have expanded planning potential beyond the scale of individual stands and toward broader, property-level applications.
Despite this progress, GIS–MCDA planning tools are still not fully integrated into everyday forestry practice, particularly in regions where small-scale and fragmented ownership remains dominant, such as the Italian Apennines. A key challenge in developing a reliable Decision Support System (DSS) for this context is the degree of agreement among stakeholders. Forest technicians, researchers, owners, and logging operators all contribute to harvesting system decisions, but their priorities and perspectives do not always coincide.
To address this issue, we developed a GIS-MCDA-based DSS to identify the most suitable harvesting system at the Forest Management Unit (FMU) level across an entire forest property of approximately 1,000 ha. For each FMU, the DSS selects the most suitable option among three alternatives: forwarder, skidder, and all-terrain cable yarder. The system relies on six criteria that can currently be derived entirely from spatial and remote sensing data, including road network density, distance to the nearest road segment, soil bearing capacity (estimated through soil trafficability mapping), terrain slope and roughness (derived from high-resolution DTM), and harvestable biomass (estimated from DSM data, based on the prescribed silvicultural treatment).
The relative importance of the criteria was determined using the Analytic Hierarchy Process (AHP) through interviews with more than 100 experts representing four stakeholder categories: researchers, forest technicians, forest owners, and logging contractors. The DSS combines spatially explicit criterion values with stakeholder-specific weights to generate suitability maps for each harvesting system. These maps are then overlaid with FMU boundaries to identify the highest-suitability system for each unit. The model was run separately using the weight sets obtained from each stakeholder group, enabling a comparative assessment of agreement and divergence in system selection across the forest supply chain.
Results are currently being finalized and will be presented for the first time at the conference.
Mountain forests often provide protection against gravitational natural hazards and supply multiple ecosystem services but are increasingly vulnerable to climate-change-enhanced disturbances due to legacies of past spruce (Picea abies (L.) H. Karst.) promotion. To optimize the allocation of limited operational resources, forest managers must spatially prioritize interventions that balance management costs with timber revenue and forest resistance and resilience ensuring the long-term protective function of these ecosystems.
We present a novel approach for the spatial prioritization of management interventions over a mid-term horizon, considering both silviculture and harvesting operations. We demonstrate the process through a case study area of a Swiss mountain forest. The process includes: (1) a spatial optimization method to delineate operationally coherent forest management units as the basis for prioritization; (2) qualitative and quantitative assessments of silvicultural benefits for forest resistance and resilience under two scenarios with different harvesting intensities (25% and 50% removal of the standing volume); and (3) a quantification of the associated harvesting, planting and thinning costs and of the net revenue for each management scenario.
Our results indicate that the more intense scenario produced a negative net revenue five times higher than that of the less intense scenario. This substantially reduces the area that can be managed given the same limited financial resources. However, the more intense management scenario offered superior silvicultural benefits by reducing mid-term disturbance predisposition and enhancing the overall long-term resistance and resilience. These findings highlight the critical trade-off between maximizing silvicultural benefits for forest resistance and resilience and operating within economic constraints.
In even-aged young Scots pine (Pinus sylvestris) forests, first thinning plays a crucial role in maintaining sustainable carbon sinks. However, forest owners often postpone thinning due to the low operational productivity (m³ Eₕ⁻¹), which reduces profitability and, consequently, contractor and industry interest. To address this sustainability challenge, we modeled boom-corridor thinning with the aim of improving operational efficiency.
The productivity modeling framework was based on workplace-level data and combined time, motion, and principal component analysis. In addition to a traditional tree-level model (RMSE 5.57; MAE 4.30), a workplace-level model was developed to predict productivity and to compare boom-corridor thinning with conventional selective thinning. Productivity differences were primarily explained by a work-motion component, including crane movements, felling and processing activities, stem length, and the number of removed stems. Additional components related to stem size, harvester movement, and supporting work phases (planning, clearing, bunching, and handling of tops and branches) were also identified.
On average, productivity in boom-corridor thinning was 20–25% higher than in selective thinning. Harvester-operator-selected boom-corridor thinning emerged as the only significant predictor of productivity and proved to be the most efficient method. Model evaluation demonstrated improved predictive accuracy (RMSE 3.49–3.52; MAE 2.36–2.45), indicating more reliable productivity estimates.
The proposed modeling framework supports the development of more profitable thinning approaches. Based on these results, harvester-operator-selected boom-corridor thinning can be recommended as a viable method for sustainable forest management in practice.
Harvester data with accurate GNSS position provides a potentially valuable source of detailed information for forest inventory in first-thinning stands. In thinning operations performed with modern harvesters, all trees along strip roads are removed, forming a systematic line sample that represent approximately 15–20% of the stand. This study investigates how strip road trees can be identified from harvester production file (hpr) data and how strip road placement and within-stand spatial variation affect the estimation of pre-thinning forest attributes. Three different strip road tree identification methods were evaluated using data from seven first-thinning stands in western Finland: (1) a boom angle and length–based method, (2) a buffer-based method using reconstructed strip road centerlines in a global coordinate system, and (3) an angle-only sector method. The accuracy of the resulting diameter distributions, stem counts, and basal area estimates was assessed against detailed field reference data using Reynold’s error index and RMSE metrics. In addition, theoretical strip road networks were established and systematically shifted across each stand to quantify within-stand variation and sampling sensitivity. The results show that the buffer-based identification method consistently produced the most accurate estimates, while the angle-only method performed weakest. Offset strip road samples revealed substantial variation in stem count and basal area in some stands, indicating that strip road alignment and spatial heterogeneity can significantly influence representativeness. The findings demonstrate that harvester-based strip road sampling is a feasible and cost-efficient approach for estimating stand-level forest variables in first thinnings, if tree identification methods and spatial biases are carefully considered.
The digital transformation of forest inventories needs the validation of consumer-grade sensors over different phenological stages. This study evaluates the seasonal measurement agreement of stand-level basal area (BA) estimates and operational efficiency between the TreeScanner smartphone application and the FJD Trion P1 professional personal laser scanner (PLS). Research was conducted in fifty circular plots (300 m2) in heterogeneous forests (45–170 years) with rugged topography (slopes 8°–42°) in Brașov, Romania. The experimental design followed a five-dimensional logical order: inter-platform BA agreement within seasons, intra-platform BA variability across seasons for both devices, and seasonal stability of acquisition times for each platform independently. Results regarding inter-platform agreement for BA showed that winter (WI) achieved the highest precision (RMSE: 0.092 m2; Bias: -0.058 m2), while in autumn (AU) it reached statistical parity (paired t-test, p=0.0926). Conversely, spring (SP) exhibited the lowest agreement (RMSE: 0.391 m2; Bias: -0.184 m2) due to vegetative noise. Intra-platform analysis revealed that BA estimates derived from TreeScanner varied significantly between seasons (p < 0.05), primarily influenced by the leaf emergence cycle and annual diameter increments. In contrast, the FJD Trion P1 data (p < 0.05), was significantly impacted by seasonal environmental constraints, such as understory regrowth and climatic variables including snow, mud, or rain. In terms of efficiency of operation, the TreeScanner application, which records the time to scan each tree individually, maintained an inter-season stable non-different (p > 0.05) rate of acquisition (16s/tree approx.). Conversely, the FJD Trion P1, which records the total time to scan the plot, was found to vary significantly (p < 0.05), as its performance proved more sensitive to structural obstructions like growing shrubs and seasonal factors such as snow and rain-induced terrain difficult. It is concluded that Smartphone LiDAR technology is a mature technology for rapid forest inventories with professional-grade agreement during the leaf-off period and better temporal stability than the professional standard.
Keywords: Mobile LiDAR, forest inventory, measurement agreement, phenological cycle, operational efficiency, basal area, personal laser scanning.
Harvester-mounted mobile laser scanning (HarvMLS) is a novel approach to quantify forest structure in situ during harvesting operations. However, its reliability relative to established aerial- and ground-based laser scanning systems remains unknown. In this study, we quantified forest structure, assessed structural changes caused by thinning, and benchmarked HarvMLS against drone-based laser scanning (DLS) and handheld mobile laser scanning (HMLS). Point cloud data were collected in a boreal forest in southern Finland during thinning operations in 2024. Forty 10 × 10 m plots were processed using a grid-level approach. Forest structure was characterised using structural attributes (specifically maximum canopy height, mean canopy height, canopy cover (CC) and filled voxel proportion) and the coefficient of variation (CV) of the structural attributes. Results showed that thinning induced consistent structural changes across all three laser systems, which included a minor reduction in maximum canopy height, a moderate decline in mean canopy height, increased canopy openness and reduced filled voxel proportion. Increased CV values across attributes indicated greater heterogeneity post-thinning. No significant differences were observed between the scanning systems in estimating maximum canopy height, mean canopy height, CC or in their associated CV values. While significant differences were detected for filled voxel proportion; post hoc tests indicated no statistically significant difference between HarvMLS and HMLS. All systems consistently estimated the CV value of filled voxel proportion. These results demonstrate that HarvMLS provides reliable estimates of forest structural attributes and their variability during active forest operations.
Tree diameter at breast height (DBH) and tree height (H) are essential parameters for monitoring forest growth and estimating carbon sequestration. Conventional field-based measurements are labor-intensive, time-consuming, and costly. In recent years, LiDAR technology has increasingly been applied in forestry due to its efficiency and long-term cost-effectiveness. However, extracting DBH and H from LiDAR-derived point cloud data remains complex, and standardized methodologies are still lacking in Thailand. Furthermore, many existing workflows rely on commercial software, which increases operational costs.
This study proposes an open-source workflow for extracting DBH and H to reduce both cost and processing time. Data were collected in a eucalyptus plantation using a SLAM-based laser scanning system. Two alternative processing approaches were evaluated. The first approach integrates CSF and TreeISO plugins in CloudCompare and the AdQSM software to estimate DBH and H. The second approach employs the 3DFin plugin within CloudCompare. Field measurements were conducted to validate the results obtained from both approaches.
A total of 225 trees were analyzed. The results show that the first approach achieved higher accuracy in estimating DBH, with an R² of 0.788, RMSE of 0.726, and MAE of 0.583, compared to the second approach, which yielded an R² of 0.541, RMSE of 1.457, and MAE of 1.325. A similar trend was observed for tree height estimation. The first approach produced estimates closer to field measurements, with an R² of 0.826, RMSE of 0.824, and MAE of 0.689, whereas the second approach resulted in an R² of 0.701, RMSE of 1.055, and MAE of 0.848.
The findings demonstrate that the first approach provides estimates closer to field measurements than the second approach. Additionally, both approaches showed better performance in estimating tree height than DBH. Despite its lower accuracy, the second approach offers advantages in terms of simplicity and reduced processing time. Therefore, each approach presents distinct trade-offs, and the choice of method should depend on user objectives, required accuracy, and resource constraints.
Accurate single-tree volume estimation is fundamental to forest inventory and operational planning in timber harvesting. Conventional volume estimation relies on field-measured tree dimensions such as diameter at breast height (DBH), tree height, and occasionally upper-stem diameters, which are labour-intensive and costly to acquire at scale. Mobile laser scanning (MLS) offers an efficient means of capturing below-crown stem geometry in boreal forests, yet its potential for supporting volume prediction through integration with biometric models has received limited attention. This study evaluates the feasibility of single-tree volume estimation by combining MLS-derived stem diameter profiles with a semi-parametric taper model. Three MLS measurement configurations of increasing stem coverage were compared on 50 Scots pine (Pinus sylvestris) trees in a managed stand in Eastern Finland: (1) MLS-derived DBH only, (2) MLS stem profile from 0 to 2 m, and (3) MLS stem profile from 0 to 5 m. Stem volume was computed by numerical integration of the reconstructed taper curve. Reference volumes were obtained from a taper model calibrated with field-measured DBH, tree height, and upper-stem diameter (d6). As a cross-check, Laasasenaho’s three-variable volume equation using the same field inputs was also evaluated.
Volume prediction improved systematically with increasing MLS stem coverage. The DBH-only configuration produced a relative RMSE of 26.1% with a negative bias, indicating that DBH alone provides insufficient geometric information for reliable volume estimation. Extending the MLS profile to 0–2 m reduced relative RMSE to 13.4%, and further extension to 0–5 m achieved a relative RMSE of 10.2%, with 74% of trees predicted within 10% of the reference volume. Laasasenaho’s equation showed close agreement with the taper-integrated reference (relative RMSE 5.1%), supporting internal consistency of the field-based benchmark under standard Finnish volume modelling practice. These results demonstrate that MLS-based below-crown stem profiles, when integrated with appropriate biometric models, can yield operationally useful single-tree volume estimates. The approach offers a practical basis for developing efficient ground-LiDAR inventory workflows, with potential applications in pre-harvest assessment and stand-level growing stock estimation.
Production planning in a Nordic context involves scheduling volumes purchased from private woodlot owners for harvest by independent cut-to-length (CTL) contractors. Planning goals include operational targets for production volume fulfillment, relocation efficiency, and ground damage. The FORAC Harvest Arena provides an online training environment where managers and students can work towards higher goal fulfillment during independent training sessions of varying difficulty.
Each training scenario involves managing 6 CTL teams (harvester, forwarder) towards fulfillment of monthly production goals for 5-7 mill customers over 3 months. 1100 harvesting contracts are located across 3 climate zones where each zone has individual patterns for weekly operating conditions (4 weather classes). The user map interface uses click-and-drag functionality to route harvesting teams between harvesting sites. A 6-factor filter function determines which sites are displayed on the map interface.
The arena offer varying scenarios where the key functionality includes a weekly prognosis for trafficability versus the actual operating conditions met by the manager when the next week arrives. Monthly mill customer demand can be varied in terms for volume and log size. The training typically starts with an informal test-run during stable conditions (mid-summer or mid-winter). It then proceeds to more challenging conditions such as freeze/thaw transitions between ii) autumn-winter and iii) winter-spring-summer.
So far Harvest arena has been used in professional and university training. The primary challenge is high fulfillment of monthly mill production goals (+/- 10 % goal volume per mill customer). In general there has been greater within-group variation for professional training. Even though scenario difficulty increases as training progresses both cohorts increased their production goal fulfillement. Average production scores for professional training increased from 59 to 66 % between autumn- and spring conditions. For university training the scores increased from 77 % for stable conditions to 82% and 87% for autumn- and spring conditions.
Reflections on learning progression: The available selection of harvesting contracts for a single CTL team decreases as demands for site bearing capacity increase and relocation distances decrease. For cohorts with lower levels of production fulfillment, KPIs for both ground damage and relocation efficiency improved with training. However for the final and most challenging scenario (spring) both professional and student groups had similar KPI-levels for ground damage and relocation distances.
Pedagogical approach: Completion of each 12-week training scenario requires 3-4 hours. Each scenario is followed by a post-exercise presentation of results with discussion of i) basic principles and ii) further improvement potential. The follow-up sessions have proved to be particulary valuable opportunity for sharing of practical insights and analytical approaches.
Forest operations are increasingly exposed to the impacts of climate change. In northern regions such as Quebec, a significant amount of wood is harvested and transported during the winter, when the ground is frozen. These operations are environmentally friendly and less costly. However, climate change is expected to lead to shorter frost periods, prolonged spring thaws, and shifting hydrological patterns. As a result of reduced winter access, dependence on costly summer roads will increase, which will likely disrupt operations, raise costs, and add logistical complexity. The full extent of these disturbances remains uncertain due to the limited predictability of future climate conditions.
The objective of the study was to evaluate the performance of forest operations under various climatic conditions using an educational game, Virtual Wood Supply Arena. Two scenarios were evaluated, representing short and prolonged spring thaws. For each scenario, teams were tasked with fulfilling roundwood demand from mills while taking into account operational limits for harvesting and transport crews.
Under prolonged thaw, transport distances increased even while delivery volumes fell by 8 %. Similarly, roadside inventories increased by 150 %, and costs per cubic metre rose nearly 12 %. It is recommended that the researchers focus on adaptation measures to mitigate these impacts and maintain competitiveness.
Knowledge of costs and profitability are fundamental for planning of all forest operations. Skogkurs is developing a new generation of web-based decision support tools for harvesting, silviculture and forest road networks. These replace traditional Excel-based spreadsheets for increased availability and usage in the field.
For more widespread use such tools need functionality and interfaces tailored for multiple users and mobile devices. The latest example of this is the Productivity and Cost Calculation (PCC) package for CTL harvesting, developed in collaboration with forest owners’ associations, the forest entrepreneur association and NIBIO (Ny kostnadskalkulator for skogsdrifter • Skogkurs). It supports conventional thinning and final felling as well as CCF operations. The programming interface (API) enables integration with stakeholders' external applications. The latest review of PCC server use showed 20 000 database operations for one month (Sept 2025). This usage is far over the levels of earlier downloadable Excel-based versions. Current work focuses on a corresponding package for pre-commercial thinning (PCT). Future work extends to decision support for forest road networks.
Excel is the preferred tool for more advanced calculations. However, it is also associated with difficulty in tracking intentional or non-intentional changes in parameters and formulas by multiple users. The aim of the new toolbox is to facilitate consistent communication and collaboration among colleagues, clients, and contractors on a shared digital platform. Validation of empirical models is done with participating contractors and forestry organizations. Such tools typically suffer from outdated parameters. The new toolbox will therefore be supported by an automated updating with sector cost indexes.
The toolbox architecture provides visual introductions to concepts and parameters relevant for the respective user groups, based on web designers’ best practice. The target groups include forest entrepreneurs, operations planners, and forest managers, as well as educational institutions and forest owners. User testing ensures the desired adaptations to both professionals and less experienced users.
The web-based toolbox enables provides increased insight into cost drivers and maintains focus on productivity development. The current tools are used in education programs ranging from operator apprenticeships to master-level forest operations courses. All stakeholders gain from improved functionality, availability and quality assurance.
A study was conducted on dynamic modeling and efficiency optimization in tree cutting by logging machines. A mathematical model was established, considering continuous reduction of tree mass, dynamic response of the driving system, and time-varying acceleration. The variation of feed rate and its impact on cutting performance were analyzed, particularly during simultaneous pruning operations. Adaptive feeding strategies were examined for common branches and special branching structures. Through theoretical deduction and numerical simulation, it was demonstrated that under identical operating conditions, increasing the cutting length from 2.25 meters to 4 meters resulted in an approximately 33% improvement in average feed rate and a reduction in total processing time ranging between 15.6% and 19.6%. The simulation results were validated against actual operational data, with an error of less than 4.9%. The model's effectiveness was confirmed, and the feasibility of optimizing cutting length was supported. These findings provide a theoretical reference for control and process parameter optimization of logging machines.
Annual logging planning with optimization solutions relies on measurable criteria for felling area conditions and suitable weather. Increased data precision improves the reliability of the optimization assessing logging feasibility and simulates the effects of varying weather scenarios. Which leads to effective process management by centralizing planning of harvesting . The study aimed to improve the precision of regression-based forecasting models predicting logging conditions through the integration of satellite data with previously used meteorological center inputs, leveraging the satellite data’s higher spatial resolution, better spatial coverage and spatial variability detection of the satellite data. Secondly, geospatial information system (LVM GEO) data was integrated with meteorological measurements, implementing finer special data(10X10km) improving the precision of area-specific forecasts. The study was realized on data of 12000 final fellings and 5000 commercial-thinning of JSC “Latvijas valsts meži” (Latvia’s State Forests) from year 2017 to 2024. For each felling area were assigned points that characterize the proportion of bad and extreme logging conditions from total cutting volume in the last 7 days in each company region. The 7-day average of adverse logging conditions correlated best with weather constraints. Felling areas were additionally characterized by felling volume combined with the proportion of the area with groundwater levels up (according to the LVM groundwater model), and the distinct impacts of each forest type. Furthermore, for each felling area, satellite-derived temperature and precipitation data over 7-, 15-, and 30-day periods during logging obtained from ERA5 and Copernicus repositories, were analyzed in parallel with measurements from 26 Latvian State Geological and Meteorological Center weather stations. Further a regression analysis was performed to develop a predictive model for assessing logging conditions and assigning “points”(lower points – easier logging, higher points – harder logging) of each felling area. Geospatial information system data that has proven to be significant (p<0.05) is the proportion of the felling area where the groundwater level reaches up to 0.1 m in combination with the felling volume in addition to forest types (p<0.05). By regression analysis the weather condition impact on logging possibilities was set. Equation allows to calculate the maximum logging points under defined weather constraints. Satellite data that has proven to be significant (p<0.05) are daily maximum air temperature at 2 m height, daily minimum air temperature at 2 m height, daily mean air temperature at 2 m height, soil moisture at depth, soil moisture index, precipitation sum, snow depth, soil temperature at depth. The additional factors allowed to increase the coefficient of determination by 1% on average for the felling data equation, and by 8.5% on average for the weather conditioning equation. As a result, using seven years of weather data, three meteorological scenarios—optimal, moderate, and suboptimal for logging—were defined per region. Monthly 15- and 30-day average temperatures and 30-day precipitation were used to calculate logging condition indices via two predictive models contained in information system and optimization system providing objective criteria to identify the most suitable month for harvesting each area.
In Mediterranean forest regions such as the Pyrenees, Continuous Cover Forestry (CCF) is widely applied to maintain structural complexity and keep trees standing after harvesting. This approach often requires extraction systems that are flexible, minimally damaging, and capable of operating under variable terrains and stand structures. Among ground-based machines, the skidder is one of the most commonly used timber extraction tools due to its versatility, particularly when equipped with a winch for ground retrieval of stems from uneven terrain.
Traditionally, basic double-drum winch skidders (DDS) have been the dominant machinery in Mediterranean small-scale forests. However, more advanced configurations such as clambunk and crane skidders (CCS) are becoming increasingly adopted due to their higher load capacity and improved efficiency during loading and unloading operations. The literature on Mediterranean logging highlights significant differences in extraction productivity between configurations, and emphasizes the importance of machine type, skid trail conditions, and extraction distance on performance metrics in ground-based systems (e.g., winch vs grapple configurations).
The objective of this study was to compare the productivities of DDS and CCS machines under two forest management scenarios: a final felling in mountain pine stands and a close-to-nature thinning in Scots pine stands. In treatments where the CCS relied exclusively on the crane for loading and used the clambunk for transport, results showed a significant productivity advantage for CCS (6.4 m³/PMH) compared to DDS (4.6 m³/PMH). Conversely, when both skidders operated primarily with the winch — simulating conditions typical of selective or uneven-aged harvests — productivity differences were not statistically significant (3.88 vs. 3.83 m³/PMH for CCS and DDS, respectively).
These findings suggest that in operational contexts where crane-assisted loading is feasible and efficient, clambunk and crane skidders offer a clear productivity advantage over conventional winch-focused skidders. When the extraction context does not allow effective crane use (e.g., dense stands or complex ground conditions), machine productivity converges, reinforcing that operational conditions and silvicultural objectives influence the optimal machinery choice.
Selective logging in structurally complex stands requires technologically flexible skidding systems capable of operating under pronounced environmental and operational uncertainty. Variability in terrain morphology, soil bearing capacity, timber assortment dimensions, and dynamic resistance to motion creates a stochastic decision environment in which conventional deterministic design approaches prove insufficient. In practice, this often results in either excessive energy consumption and underutilised traction potential or increased soil disturbance due to inappropriate parameter settings of technological equipment. Therefore, a scientifically grounded framework for the justification of skidding machine parameters is required, capable of integrating technical performance, environmental constraints, and incomplete information within a unified decision model.
This study proposes a multi-criteria optimisation methodology based on Single-Valued Trapezoidal Neutrosophic Numbers (SVTN-numbers) to support the design and operational justification of technological equipment for skidding machines in selective logging. The neutrosophic formalism enables simultaneous representation of truth, indeterminacy, and falsity components of expert evaluations, thus explicitly incorporating uncertainty inherent in forest operations. The trapezoidal structure allows interval-based modelling of key operational parameters, such as tractive effort, travel speed, and energy intensity, which are strongly influenced by slope gradient and micro-relief variability.
A decision matrix was constructed to evaluate three technological configurations for selective logging: (A1) a conventional skidder equipped with a grapple, (A2) a tractor fitted with a hydraulic manipulator, and (A3) a skidding system with a winch mounted on a front loader. Alternatives were assessed against productivity, energy consumption, and manoeuvrability in restricted stand conditions. Criteria weights were derived using a neutrosophic soft-matrix approach to reflect fluctuating external loads and partial expert knowledge. Defuzzification through a neutrosophic score function enabled the transformation of imprecise assessments into comparable scalar indicators and the identification of a Pareto-efficient solution space.
The integrated evaluation demonstrated that configuration A3 achieved the highest aggregated performance index (0.625), exceeding conventional systems (0.48 and 0.55 respectively). This result is causally associated with the structural integration of skidding and loading functions within a single machine platform. The frontal positioning of the winch improves mass distribution, enhances operational responsiveness, and reduces auxiliary handling time, thereby increasing overall system productivity while maintaining lower energy intensity per cubic metre extracted. From an environmental perspective, reduced machine passes and improved manoeuvrability contribute to decreased soil compaction risk, which is critical in selective logging systems oriented towards sustainable forest management.
The proposed neutrosophic optimisation framework transforms the parameter justification process from an intuitive engineering choice into a transparent, analytically substantiated decision-support tool. By embedding uncertainty directly into the mathematical model, it enhances the robustness of technological design under variable forest conditions. The approach provides a methodological basis for adaptive skidding machine configuration and can support strategic equipment selection in forest enterprises seeking to balance economic efficiency with ecological integrity.
In Sweden, half of the forest area is owned by approximately 300 000 non-industrial private forest (NIPF) owners. While most harvested timber is produced by companies using harvesters and forwarders, NIPF owners still contribute a substantial share of the total time spent on forest work. A large part of this work concerns logging, where tractor-based equipment, using chainsaws, tractor-mounted processors, and grapple loader trailers, represent a low-investment alternative for small-scale logging operations. Additionally, this equipment may be economically feasible for low-removal harvesting. However, studies on tractor-mounted processors’ work performance have been limited in recent decades, resulting in a lack of updated data on time consumption and cost estimates. Additionally, to the authors’ knowledge, no previous study has evaluated the combined tractor-based system during low-removal harvesting.
To address this knowledge gap, time studies were conducted in Southern Sweden during a low-removal harvest operation with a removal intensity of 10%. The studied system consisted of a Husqvarna 550 XP G chainsaw, a Hypro 450 XL processor, and a 101 4WD FTG Moheda grapple loader trailer, where the processor and grapple loader trailer were attached to a Valtra A105 tractor. The same operator executed all work in the study. Based on the time study data, regression models were developed that explained time consumption for work elements, and machine costs were estimated using the COST model.
In total, the studied logging operation comprised 36.2 productive machine hours excluding delays (PMH₀). Data was collected for 157 trees for chainsaw work and 161 trees for processor work, where the discrepancy was caused by equipment errors. Mean diameter at breast height (DBH) was 23.4 cm. For forwarding, 12 loads were recorded with a mean forwarding distance of 1 260 m and a mean load volume of 6.6 m³ob. Observed productivities for the chainsaw, processor, and grapple loader trailer work were 8.7, 4.3, and 6.3 m³ob PMH₀⁻¹, respectively. Costs, when excluding wages, were 0.26, 11.61, and 6.88 € m³ob⁻¹ for the chainsaw, processor, and grapple loader trailer work.
For chainsaw work, inter-tree distance, DBH, manual bucking, and tree species had a significant effect on time consumption. For processor work, distance between tractor stops, winching distance, DBH, manual bucking, and tree species were significant predictors. For the grapple loader trailer, forwarding distance and load volume significantly affected time consumption. Overall, the resulting models showed clear and strong relationships between predictors and time consumption and provide updated data for evaluating tractor-based harvesting systems. However, caution is warranted as the study was based on the observations of a single operator.
The aim of the study was to characterise the effects of timber skidding using a prototype battery-powered Salamander 600 4×4 micro-skidder and the impact of this operation on the topsoil layers along the skid trail. The tests were carried out in a young oak stand growing in a mountain broadleaved forest habitat, on sloping terrain with an average slope of 15° (26-27%) and numerous obstacles in the form of stones.
The timber was skidded in two stages. In the first stage, the long logs were pulled individually from the cutting area towards the skid trail. The average distance of timber transport in this stage was 42 m (39-44 m). Then, each load was transported along a 70-metre trail to a storage site by the forest road, where short assortments (bolts) were cut and loaded onto a transport vehicle. The slope gradient in the cutting area was 13° (23%), and on the trail in the longitudinal direction it was 9° (16%). On the skid trail surface, there were many boulders protruding from the ground and several transverse depressions, which effectively limited the speed of the micro-skidder both with and without a load.
During the work, continuous time measurements were performed using the Timing 1.1 application. The application allows the duration of all observed activities to be recorded with an accuracy of 1 second. The duration of each successive observed operation and delays were recorded in a PSION Workabout MX portable field recorder. Before the observation began, the length and diameters of the logs were measured. The Huber formula was used to calculate their volume. The operator had two years of experience both in forestry operations and in operating micro-skidder.
Soil parameters were measured after the first and tenth skidding cycles on three transects perpendicular to the trail axis, located at the beginning, middle and end of the trail. Within each transect, at five points (in ruts, between ruts and at a distance of 1 m from ruts on both sides of the trail), soil compactness was measured with a FieldScope SC 900 static penetrometer, and soil moisture and conductivity were measured with a Spectrum Technologies TDR 350 meter. In addition, soil samples were taken to determine bulk density and moisture content under laboratory conditions.
The average volume of a single load was 0.16 m³ (0.10–0.27 m³). The time required for skidding in productive work time, in the first stage over the described distance within the cutting area, amounted to over 22 min/m³, while the time required for skidding along the trail amounted to over 24 min/m³. The micro-skidder operator needed a total of 47 minutes to skid 1 m³ of timber over an average distance of 112 m. The skidding efficiency in the first stage was 2.89 m³/h and was about 13% higher than the skidding efficiency along the skid trail. A statistically significant increase in soil compactness occurred on the trail in ruts at a depth of up to 20 cm. Changes in other parameters proved to be statistically insignificant.
Harvesting operations in short-rotation eucalyptus plantations present increasing challenges, particularly when balancing diverse operational practices and performance expectations. Variability in field operations further complicates decision‑making, underscoring the need for models that capture realistic productivity dynamics. This study used a simulation-based approach to evaluate how the performance of two commonly used farm tractors modifications —differing in size class and equipped with different types of backhoe grapples—is influenced by payload characteristics and extraction distance in Thailand’s very short‑cycle eucalyptus plantations. Both tractors (T1 and T2) were monitored using GNSS, from which operational events and travel speeds were derived. Payload size was estimated through manual field measurements. Simulation experiments were then used to quantify how moving speed and payload size affect productivity across a range of extraction distances. Significant differences (α = 0.05) in travel speed and payload capacity between the two machines resulted in pronounced disparities in productivity as extraction distance increased. While the number of daily work cycles for both tractors remained similar up to an extraction distance of 100 m, differences emerged in daily production (m³), productivity (m³/h), and efficiency (h/m³), with the larger T1 consistently outperforming T2. At longer distances (e.g., 1000 m), the combined effect of reduced speed and smaller payloads tripled the unit extraction time for T2. These results have important implications for operational planning in Southeast Asian short‑rotation forestry. Because these tractors are often used interchangeably in local practice, keeping extraction distances below 50 m minimizes performance variability. At greater distances, T1 maintains notably higher production and a greater number of work cycles, emphasizing the importance of selecting an appropriately sized tractor based on site layout. These tractor modifications were introduced for the first time in international context and presented new sights for small-sized timber extraction performance.
Accurate timber measurement is essential for forestry and the wood industry. Lower-value assortments such as pulpwood or energy wood are commonly assessed by manually measuring the gross stack volume and converting it to net volume using conversion factors. These traditional methods are labor-intensive and prone to human error, highlighting the need for fast and objective digital alternatives.
This study evaluates three LiDAR systems (a terrestrial laser scanner (TLS), a personal laser scanner (PLS), and a LiDAR-equipped tablet) and four commercial photo-optical applications (iFovea, LogStackLIDAR, PolterMAX, and Timbeter) for measuring gross volume, net volume, and conversion factors of wood stacks under operational conditions representative of Austrian forestry practice. In total, 22 wood stacks were measured at a timber yard in Austria. Reference net volumes were obtained from electronic 3D log measurements at a sawmill, while reference gross volumes were derived by simulating measurements according to the German roundwood trading framework on high-precision TLS point clouds.
For LiDAR-based measurements, log ends were extracted from the 3D point clouds, rasterized, and analyzed separately for both stack faces. Gross volume was estimated by constructing an alpha-hull around the rasterized log ends to determine the face area and multiplying it by the assortment length. Net volume was calculated from the summed raster area representing solid wood, and also multiplied by log length. The conversion factor was derived as the ratio of net to gross volume.
The LiDAR-equipped tablet showed the highest combined accuracy (RMSE_gross = 2.60%; RMSE_net = 3.58%; RMSE_conversion = 3.33%), outperforming both the photo-optical applications and even the TLS. This result demonstrates that low-cost, consumer-grade LiDAR sensors can provide high measurement precision under practical field conditions. The TLS also produced reliable results, though minor occlusion effects likely reduced its performance in net volume estimation. The PLS delivered accurate gross volume estimates but showed substantial variability in net volume and conversion factor results, mainly due to fluctuations in point cloud density linked to scan duration.
Photo-optical applications achieved high accuracy for gross volume (RMSE_gross = 4.73% – 8.78%). However, net volume accuracy varied more (RMSE_net = 5.28% – 10.74%). The widely used quadrant method, simulated in this study, showed accuracy comparable to photo-optical approaches but lacked their efficiency and automation advantages.
Regarding field efficiency, photo-optical methods were fastest on average per stack, with Timbeter (0:53 min) and PolterMAX (1:14 min) requiring the least time. The PLS (1:42 min) and LiDAR-equipped tablet (2:39 min) required comparable field times, whereas TLS measurements were substantially slower (6:54 min). Field time variability was particularly high for iFovea and LogStackLIDAR due to occasionally required manual corrections caused by wood stacks in the background.
Overall, the results confirm the strong potential of LiDAR-based approaches for accurate, efficient, and automated wood stack measurement. By enabling reliable estimation of gross volume, net volume, and conversion factors, LiDAR technology can enhance transparency, traceability, and data continuity across the forestry supply chain. The demonstrated performance of the consumer-grade tablet suggests promising opportunities for cost-effective implementation in forestry operations.
The accuracy of round wood volume determination is fundamental to planning and accounting for individual processes in the wood trade. The rapid technological development of mobile devices, including computers, digital cameras, and automated systems, has enabled the emergence of practical data acquisition tools. This also applies to various areas of forest management. One option in this regard is the measurement of wood raw material using photo-optical techniques.
The functioning of the different applications is quite similar and proceeds in the following phases:
1. Taking a series of images of a stack of medium or large-sized round wood.
2. Merging of the images into one panoramic image.
3. Analysis of the panoramic image in order to:
4. Identification of the cross-cut forehead of round wood and counting them,
5. Inscribing circles in the cross-cut forehead sections,
6. Making a diameter distribution table,
7. Inscribing the pile into the polygon and calculating the area of the polygon,
8. Determination of the replacement coefficient from the ratio of the area of the cross-cut foreheads of wood to the area of the polygon, or use of a preset coefficient.
9. Calculation of the stack volume and preparation of the report.
The State Forests had its first experience with photo-optical measurement of wood raw material between 2013 and 2015, commissioning a research project for the SF Development and Implementation Centre in Bedoń. Three applications were analyzed: iFOVEA by FOVEA GmbH, AFoRS by Scheller Systemtechnik GmbH, and sScale by Dralle A/S. Approximately 25,000 m3 of stacked wood were measured. Compared to the volume obtained from manual measurement, the values for the individual methods were -0.58%, +2.17%, and +4.07%, respectively, and the labor intensity ratio was 1:0.47:0.63:0.23. Later, only the sScale application was used more widely, measuring approximately 1.6 million m3 of stacked wood from post-hurricane areas in 2017-2018. In 2020, tests with the Timbeter application began. In 2024-2025, as part of the second phase of implementing photo-optical roundwood measurement, this application was deployed in over 100 forest districts (primarily in the Piła and Poznań regional directorates). Test results indicated that the application provides sufficient quality results and the required accuracy compared to the traditional reference. It enables seamless data transfer to the State Forests' IT system. As of October 2025, approximately 650 forest districts are using this application in production.
Accurate, early and cost-effective field information is a bottleneck for both forest and orchard production systems, especially when decisions depend on reproductive potential and individual tree or hedge structure. The ARGUS consortium develops and validates operational UAV and terrestrial LiDAR workflows to (i) automatically quantify reproductive organs in maritime pine crowns and chestnut orchards, and (ii) estimate dendrometric variables for maritime pine trees and hedge-scale descriptors for hazelnut orchard hedgerows.
For maritime pine flowering, ARGUS implements an end-to-end processing chain from acquisition to indicator delivery. UAV image acquisition was evaluated both with autonomous waypoint missions and targeted crown imaging, and the retained workflow relies on standardized, close-range crown imagery to capture high-resolution images of reproductive organs. A deep-learning detection model, trained with expert annotations and multi-site acquisitions, delivers automated flower counts that are robust, enabling faster, more consistent and scalable estimations. These outputs support monitoring of flowering dynamics and improve decision-making for forest management and seed supply planning. In addition, ARGUS delivers an automated method to detect and quantify chestnut burrs from UAV imagery under variable orchard conditions. The outputs can be produced at orchard and block scales and are directly usable for yield anticipation and operational management.
For dendrometry, terrestrial LiDAR is used to derive structural metrics tailored to each production context. In maritime pine, individual-tree attributes including stem diameter (DBH) and height provide objective phenotyping variables that support genetic selection and evaluation. In hazelnut orchards, hedge-oriented dendrometry (height, width and volume proxies) enables rapid characterization of canopy structure and provides actionable indicators linked to orchard productivity and monitoring.
Forestry operations are under growing pressure to supply sawlogs as quickly, cost-effectively and customer-specifically as possible. In fully and highly mechanised timber harvesting systems, harvesters and processors take on a key role at the front end of the wood supply chain, as they determine the assortment lengths directly in the machine. The commonly used length measurement systems rely on a tactile measuring wheel that is pressed against the trunk during feed. However, under the extreme operating conditions, these systems are increasingly reaching their technical and methodological limits.
The current measuring principle records the distance travelled along the log surface, which corresponds to the shortest end-to-end connection only in the ideal case. In addition, variable contact conditions affect the measurement: contact pressure, wood and bark properties as well as weather conditions (wetness, ice, temperature) change the effective rolling circumference. Furthermore, slippage, missing bark, branches and branch stubs can interrupt continuous contact with the log surface and lead to measurement errors.
In practice, machine manufacturers therefore prescribe regular manual calibrations, often several times per day, to detect and reduce systematic deviations. This additional step, however, reduces productivity, is itself prone to errors and has little effect on the variation of the measured data. To reliably meet the required minimum length and avoid the risk of quality downgrades due to undersized lengths, additional overlength is therefore often applied. This results in an economic loss of value for forest owners, as sawmills usually only pay for the required nominal length.
Within this context, a systematic literature review is being conducted to examine the current state of research on non-contact alternatives for length measurement in harvester and processor heads. The aim is to identify existing approaches, compile and compare them and thereby develop a clear insight into the challenges and potential of individual measurement methods in view of the extreme operating conditions in timber harvesting.
Alternative contactless measurement methods have been investigated since the early 2000s. Optical approaches are considered promising, as are methods using microwave or laser technology. Key limitations are interference from contamination, vibration and occlusion. At the same time, advances in sensing technologies and analysis algorithms are expected to further increase measurement accuracy and robustness. In light of these findings, subsequent research will focus on developing and evaluating a novel, non-contact, multi-sensor system to achieve more precise, calibration-independent length measurements.
Accurate estimation of harvested wood volume is one of the most crucial processes for billing. It makes the roundwood trade a key task in forest management. Globally, in many countries, measurements are conducted using the traditional volumetric sectional method with analog tools. Currently, many digital alternatives, such as photo-optical methods supported by smartphones or ad hoc-designed devices, or data collected by harvesters, are available.
The main aim of this study was to analyze and compare the results of roundwood measurements obtained using the traditional sectional method and digital methods such as the Timbeter app, the LogStackLIDAR app, and harvester data from the John Deere TimberMatic™ system for measuring stacked wood in forest conditions. In total, the gross volume of 75 stacks of pinewood was estimated. The total estimated volume was over 4100 m3. Subsequently, each stack was classified according to volume ranges: I − smaller than 30m3 (st), II − from 31 to 60m3 (st), III − from 61 to 90 m3 (st), and IV – larger than 90m3. For each analysis, the traditional measurement was considered the reference value to evaluate the accuracy of the digital methods.
The accuracy was then determined as the deviation between the average volumes estimated with the apps and the average volume obtained by applying the traditional method (55.93 m3). Thus, the highest accuracy was obtained with the LogStackLidar app (56.79 m3, +1.5%), followed by the harvester measurement (59.30 m3, +6%) and the Timbeter app (65.00 m3, +9%). Given the classified stack sizes, LogStackLIDAR demonstrated stable results and high accuracy across all stack size classes. Timbeter showed systematic overestimation across all classes. Harvester measurements revealed variable performance characteristics: significant errors in small stacks, excellent agreement in Class III, and higher overestimation in Class IV.
To summarize, a detailed analysis shows that both LogStackLIDAR and harvester measurements are highly accurate relative to manual measurements. Overall, the findings indicate that measurement accuracy is method-dependent and, in some cases, influenced by stack volume. The results highlight the importance of method selection in roundwood trade and billing processes, particularly when systematic errors may significantly affect financial settlements.
Reliable end to end traceability of individual sawlogs from the forest to the sawmill is a key enabler for digitalisation, process optimisation, and value recovery in the forest value chain. Code based log identification using computer vision offers a scalable alternative to physical tagging, but its practical deployment is challenged by variable visual conditions, partial occlusions, and classification uncertainty typical of forest and industrial environments. This study investigates the robustness of Reed–Solomon–based symbol encoding for code based sawlog identification under realistic detection and classification noise.
We present a simulation based evaluation of a computer-vision classifier system using an Reed-Solomon RS(8,4) symbol code to encode log identifiers. The code consists of four data symbols and four parity symbols, providing a minimum Hamming distance of five and enabling correction of multiple symbol errors or erasures. To emulate realistic classifier behaviour, synthetic confidence matrices are generated by jointly modelling per symbol detection probability and inter class discriminative power. This framework enables controlled analysis of identification performance across a wide range of noise levels, symbol loss rates, and dictionary sizes, up to the full RS code space of 23⁴ possible identifiers.
Code recovery performance is evaluated using both hard decision decoding, which selects the most likely symbol at each position, and soft decision decoding, which exploits the full confidence distribution across symbol candidates. Results show that soft decision decoding consistently and substantially outperforms hard decoding across all tested scenarios. In low noise conditions, correct identification rates exceeding 99% are achieved even for per symbol detection probabilities as low as 0.8, without applying any dictionary constraints.
Under moderate noise, identification rates above 98% are maintained for dictionary sizes of up to 20 000 identifiers, demonstrating strong scalability for industrial scale deployments. At high classification noise levels, and symbol detection probabilities as low as 50%, performance becomes increasingly dependent on both code design and dictionary filtering. Here, 74% of identifiers are correctly recovered using dictionary-constrained decoding, primarily due to the strong minimum distance properties of the RS(8,4) code. This performance is primarily attributable to the code’s minimum Hamming distance of five, which guarantees unambiguous recovery in the presence of up to four missing or erroneous symbols. Notably, even when more than four symbols are missing or misclassified, the dictionary-constrained decoder frequently identifies the correct code candidate, highlighting the benefit of combining error correcting code structure with probabilistic decoding.
Overall, the results demonstrate that Reed–Solomon–based symbol encoding, combined with soft decision decoding and optional dictionary constraints, enables robust and scalable code based sawlog traceability under challenging visual conditions. The proposed approach supports reliable digital identification of individual logs in forest and sawmill environments without reliance on physical tags or manual intervention.
The forestry–wood sector faces a growing need to increase efficiency, transparency, and trust across the supply and processing chain. In roundwood (log) trading, however, a pronounced information asymmetry often persists: buyers make decisions based on limited or non-comparable data on the quality of individual assortments, which increases transaction costs, raises the risk of disputes, and may lead to inefficient utilization of wood raw material. At the same time, pressure is increasing for faster, data-driven decision-making that better links input quality with the economic outcome of processing.
In this context, we present iWOOD—a secure online solution for auction-based log sales, designed to connect modern scanning technologies with an electronic marketplace and an auditable trading process. The key innovation is the integration of 3D log scanning, including CT, enabling the creation of a digital log profile: a data-supported description of selected quality parameters and a visualization (or indication) of internal defects. This profile provides a standardized information basis for comparable bid evaluation and reduces uncertainty in purchasing. From a processing perspective, it also enables improved planning and optimization of operations to maximize yield and product value from available raw material.
iWOOD is conceived as a modular system supporting user registration and role management, creation of auctions with detailed raw material descriptions, and uploading of multimedia (photos, video, 3D/CT data). The result is a clearly defined auction item complemented by a digital log profile that can be selectively disclosed to relevant potential buyers. The platform covers the entire auction workflow—from authentication and authorization, through auction execution and notifications, to bid evaluation and result recording. Strong emphasis is placed on data security and integrity: encrypted communication, two-factor authentication, secure integration with other systems via access tokens, clearly defined user roles, and logging of all key events for traceability and audit.
The platform uses rule-based automated bid evaluation according to predefined criteria and includes analytical tools to support pricing and purchasing decisions. Development is carried out within the LignoSilva Centre of Excellence, which provides both the technological infrastructure (3D/CT scanning) and expert capacity for process digitalization in the sector. The solution builds on a concept validated in a relevant environment and is progressing toward a pilot demonstration under operational conditions with market actors (forest owners, traders, and wood processors). The expected benefits include increased transparency and efficiency in the roundwood market, reduced transactional risks, stronger trust between parties, and a better linkage between raw material quality and processing decisions.
Value-based bucking is a central optimisation task in cut-to-length harvesting systems, where stem segmentation into assortments determines both product distributions and economic performance.
In Norway, harvester price matrices are typically harmonised within procurement regions, even though considerable variation exists in stem form, internal wood quality, and other tree-level characteristics that influence product value realisation. This study examines how tree-level quality attributes affect optimal bucking decisions and how these shifts propagate along the supply chain under Norwegian operational conditions. Using detailed harvester recorded stem measurements (.hpr) combined with a dynamic programming bucking simulator (OptBuck), we simulated assortment outcomes across multiple scenarios that incorporate variation in wood properties.
Results show that including quality related attributes in bucking optimisation systematically alters assortment composition and net stem value, with pronounced impacts on the proportion and value of sawlogs, being the most economically significant assortment. These changes further influence downstream logistics flows and overall supply chain economics. The findings emphasise the need to integrate tree-level quality information into bucking parameterisation and procurement planning to enhance value capture and operational efficiency in modern forestry.
Traditional measurement methods have limitations in data accuracy and compliance with international traceability standards. In the context of forest management and supply chain digitalization, there is a need to develop and test new tools that improve the effectiveness of wood sourcing and tracking. This study evaluated the performance of a digital wood sourcing, tracking, and quality evaluation system developed within the Horizon Europe SINTETIC project. The system integrates Radio Frequency Identification (RFID) of trees and logs, Light Detection and Ranging (LiDAR) for biometric measurement, and acoustic stiffness assessment, all within a client-server digital ecosystem. Two sites were considered based on extraction type: a thinning operation in a spruce-dominated stand and a group shelterwood cut in a mature beech-dominated stand. More than 250 standing trees and 560 logs were surveyed using both manual inventory procedures and the developed digital system. At both sites, detailed measurements were taken regarding biometrics and time consumption for each item; at the second site, digital quality assessment was also included. The data were used to estimate the RFID reading rates, the success of quality assessments, biometric estimate comparisons, and the efficiency and cost of the compared methods. The RFID reading rate was nearly 100%, with only one tag failing due to external factors. The log quality assessment had a success rate of 95%, with failures mainly due to logs with obvious cavities caused by decay. Bias in DBH and tree height was very low, indicating average differences of less than 0.06 cm and 0.4 m, respectively. For logs, differences in volume estimates were less than 0.013 m³. The manual system yielded an efficiency of 0.014 hours per item, while the digital system required 0.020 hours per item (assuming measurements were taken on one tree and four logs sourced from that tree). Within the digital system, tagging added approximately 0.011 hours per item, while tagging and quality checking combined added 0.022 hours per item. Economically, the manual system was estimated to cost 0.37 €/tree, whereas the digital system cost 0.23 €/tree, primarily due to lower labor input. Tagging added 0.24 €/tree, while tagging plus quality assessment added 0.42 €/tree relative to the digital system's base measurement capabilities. These results demonstrate the performance of the digital system which, in addition to lower base costs, provides a fully digital workflow, more precise item-based biometrics, a tamper-proof tracking method, and integrated quality assessments.
The application of large language models (LLMs) in forestry is still experimental to large extent, specifically in highly specialized subdomains such as the European castaniculture. This scientific initiative presents CastaneaGPT, a domain-specific digital assistant which is designed to support sweet chestnut (Castanea sativa) management in Italy and beyond. CastaneaGPT is conceived as both a working prototype and a methodological case study, and aims to illustrate how a professional AI system can be developed within a forestry context.
The system follows a modular architecture with a clear separation between backend “reasoning” and frontend presentation. The backend is based on a Retrieval-Augmented Generation (RAG) pipeline that combines OpenAI LLM models with LlamaIndex for document ingestion, indexing, and retrieval. The integrated text corpus includes curated scientific literature, grey literature, and web-based materials, processed through structured PDF and URL ingestion workflows with metadata extraction and citation control. Upgrade measures such as enhanced retrieval strategies (e.g., MMR, hybrid BM25), chunk-size optimization, conversation memory, and a forestry-related glossary are implemented to consolidate the factual grounding of the tool. FastAPI serves as the orchestration layer, enabling containerized deployment (Docker) and flexible frontend integration (e.g., Glide). The user interface is planned to integrate three main features: dashboard, map (PostGIS-based), and chat components.
Beyond the technical realization, the project is driven by two principal scientific interests: (1) the methodological challenges in building a reliable forestry-specific RAG corpus, and (2) the empirical testing and evaluation of different interaction modes between user and AI assistant (e.g. for professional castaniculturalists; seekers of decision support when reactivating abbandoned chestnut plantations; playful users such as children) in order to evaluate the influence of these interaction modes on the comprehension and the trust of the users.
CastaneaGPT is conceived as a digital platform to connect the castanicultural community, support decision-making, and strengthen knowledge dissemination, while offering a reproducible blueprint for similar domain-specific AI tools in forestry.
Soil compaction caused by forestry machinery poses a significant challenge to the sustainability of forest ecosystems, affecting not only physical soil properties such as pore space and water retention but also biological activity—a key driver of nutrient cycling and soil regeneration. Traditional methods for assessing biological activity are often invasive or labor-intensive. A promising alternative is ecoacoustics, which uses soil microphones and indices like the Acoustic Complexity Index (ACI) to non-invasively and continuously monitor the acoustic activity of soil organisms. This approach offers a way to evaluate soil health and the impacts of different management intensities. This study addresses two central questions: 1. Long-term effects of skid trails: How does machinery traffic during timber harvesting influence biological activity in sandy forest soils, both immediately after compaction and years after last use? and 2. Comparison of forest road types: Do different forest infrastructure types (natural trails, skid trails, forest roads) differ in their impact on acoustically measurable biological activity?
The hypotheses suggest a reduction in biological activity due to compaction (measured via ACI) and differences between road types based on usage intensity and substrate composition. Data collection took place in two study areas within the Schorfheide-Chorin Biosphere Reserve, locates in the federal state of Brandenburg, Northeast Germany:
• Skid trails: Acoustic measurements (soil microphones, ACI calculation) and soil physical parameters (density, pH, humus content) were recorded on freshly trafficked and long-unused skid trails.
• Forest road types: Natural trails, skid trails, and forest roads were compared in terms of acoustic complexity (ACI) and microclimatic factors (soil/air temperature, moisture).
Preliminary results reveal no significant differences in acoustic complexity (ACI) between trafficked and untrafficked areas or among road types—despite visually apparent trends (e.g., higher ACI values on skid trails). This raises questions about the sensitivity of the method: While classical soil parameters (density, humus content) responded to compaction, acoustic activity remained largely unaffected. Possible explanations include the overall low biological activity in sandy soils, spatial heterogeneity of soil organisms, limited sample size, or methodological constraints (e.g., sensor sensitivity, noise interference).
However, these findings highlight the potential of ecoacoustic methods as a non-invasive tool for assessing soil health. At the same time, they underscore the need for methodological refinements, which will be included in our further research. In the long term, ecoacoustics could play a bridging role between research and practice, enabling forest managers to monitor the impacts of machinery traffic or infrastructure measures in real time and across large areas—a critical step toward adaptive, soil-conserving management strategies
Soil degradation resulting from mechanised timber harvesting on mountain slopes represents a critical challenge to the long-term sustainability of forest ecosystems. In the Ukrainian Carpathians, the extensive network of skid trails established during logging operations frequently becomes the primary source of erosion, soil compaction, and hydrological disruption. Despite increasing recognition of these environmental impacts, the adoption of effective, context-specific reclamation strategies remains limited. This study synthesises current scientific knowledge and field-based observations to develop an integrated framework for restoring soil functionality and ecosystem resilience on degraded skid trails in this region.
Field investigations were conducted on selected logging sites across the Ukrainian Carpathians, examining skid trails of varying ages and usage intensities. Particular attention was paid to sections located on slopes exceeding 20-25°, where gravitational processes exert the most pronounced influence on trail degradation. These observations were complemented by analysis of slope gradients, surface hydrological pathways, and the morphological characteristics of degraded trail segments.
The findings demonstrate that skid trail degradation arises from a combination of mechanical soil displacement, loss of structural stability, and altered surface hydrology. On steep slopes, these factors accelerate the formation of rills and gullies, substantially delaying natural recovery processes. Three principal reclamation approaches were identified as relevant to mountain conditions: technical interventions involving mechanical restructuring and water diversion; bioengineering techniques utilising fascines, root mats, and natural materials for slope stabilisation; and biological restoration through revegetation with native species, including nitrogen-fixing plants and deep-rooted shrubs. However, the effectiveness of any single method applied in isolation remains limited. Sustainable restoration requires their synergistic integration, tailored to local slope gradient, soil characteristics, and erosion risk. When appropriately combined, such integrated treatments have been demonstrated to reduce erosion intensity by up to 40-60% compared to untreated controls, significantly accelerating the recovery trajectory of degraded slopes.
Based on these findings, a tiered reclamation strategy is proposed. First, proactive planning of skid trail networks during harvesting operations can substantially reduce initial soil disturbance. Second, post-harvest interventions should combine shallow mechanical loosening to alleviate compaction with the installation of bioengineering structures along critical drainage lines. Third, biological restoration using site-adapted native species accelerates humus formation, root reinforcement, and long-term soil stabilisation. Systematic monitoring of vegetation establishment, erosion dynamics, and soil recovery is essential to evaluate reclamation success and inform adaptive management practices.
The proposed integrated approach aligns with nature-based solutions principles and contributes to restoring the water regulation, carbon sequestration, and biodiversity support functions of mountain forest soils. It offers a practical pathway for forest managers in the Carpathians and analogous regions to enhance ecosystem resilience following mechanised logging. Further research should focus on long-term field trials to validate the cost-effectiveness and ecological outcomes of different method combinations under varying site conditions.
Forest roads are a critical component of forest management infrastructure. They ensure access to timber resources but may also support recreation, conservation and access to rural areas. In Norway, extreme precipitation events have lately caused widespread road failures and erosion and have highlighted limitations in existing drainage design standards. This study presents final results from an integrated assessment of forest road drainage performance under extreme rainfall conditions with respect to culvert design, density and placement.
We chose two case study areas heavily affected by extreme precipitation events in recent years to study the efficiency of current design standards under present and future climate conditions. We combined detailed field surveys with digital terrain data to model flow paths, estimate surface runoff and evaluate the hydraulic performance of existing drainage structures for different return periods and under climate-adjusted precipitation scenarios. These analyses were complemented by cost–benefit assessments to quantify the relationship between drainage capacity, infrastructure damage, and investment costs.
The results indicate that minimum design standards based on 25–50-year return periods are often insufficient under current and projected climate conditions, particularly in steep terrain. Culvert capacity, inlet and outlet configuration, road alignment, and longitudinal slope were identified as key factors influencing failure risk. The study further demonstrates that targeted upgrades of critical drainage elements, including increased culvert dimensions and improved inlet protection, can substantially reduce damage while remaining economically viable.
Based on the empirical findings and model simulations, the project delivers updated, evidence-based recommendations for forest road drainage design, upgrading, and maintenance. A practical risk assessment framework and a spreadsheet-based decision-support tool for culvert dimensioning were developed and validated in the study areas. These tools enable forest road planners and managers to identify vulnerable sections, prioritize interventions, and balance technical performance with economic constraints.
By integrating field observations, hydrological modelling, and economic analysis, this research provides concrete guidance for improving the resilience of forest road networks under increasing climate variability. The outcomes support adaptive infrastructure planning and contribute to strengthening sustainable and resilient forest value chains.
Forest road infrastructure plays a multifunctional role in mountain forest management, with a focus on enabling efficient forest management while potentially altering catchment hydrology and flood response. Despite their importance, the quantitative hydrological effects of forest road networks under varying maintenance conditions are not well understood, which limits evidence-based planning and prioritisation of infrastructure investments. This study combines a comprehensive digital road network assessment with hydrological scenario modelling to quantify the influence of forest road condition and drainage functionality on runoff dynamics in the Rindbach catchment in the Northern Limestone Alps in Austria.
First, the entire network of forest roads accessible to trucks within the catchment area was digitally recorded. Then, selected sections were inspected through systematic on-site surveys. These surveys included a detailed assessment of culverts, the condition of the road surface, longitudinal drainage systems and the load-bearing capacity in both dry and wet conditions. The results revealed significant deficiencies in the infrastructure. For instance, only 59% of culverts were fully functional, while 22% were non-functional and just 18.2% of longitudinal drainage systems were operating at full capacity. However, the load-bearing capacity remained adequate, even in wet conditions (an average of 72.7–73.9 MN/m²), indicating the prevailing geology (limestone). This suggests that weather-related usage restrictions are primarily due to drainage deficiencies rather than structural deficiencies.
Hydrological simulations were then performed using NASIM a deterministic precipitation-runoff simulation tool to evaluate three infrastructure scenarios: (i) the natural state without roads; (ii) the optimal state with fully functional drainage; and (iii) the actual state with documented deficiencies. These simulations were conducted under a 10-year, 12-hour design precipitation of 118 mm. In the sub-catchment areas, peak discharge (Q_max) decreased from 8.90 m³/s in the natural state, to 7.50 m³/s with optimal drainage infrastructure, and to 6.80 m³/s in the actual state. This demonstrates that functional drainage systems could act as effective buffers, flattening hydrographs and reducing flashiness indices. Deficient structures prolonged concentration phases and increased baseflow duration through uncontrolled retention effects. These findings challenge the conventional assumption that forest roads universally accelerate runoff, revealing that properly designed and maintained drainage infrastructure could mitigate rather than exacerbate flood risk.
The practical implications emphasize the targeted improvement of culverts and drainage ditches, the strategic closure of little-used roads to restore retention areas, and the implementation of decentralized water retention measures such as retention ditches and infiltration zones. The digital road network provides a solid basis for maintenance planning in line with requirements, hazard assessment, and prioritization of forest road infrastructure investments. This integrated approach demonstrates that hydrologically sensitive forest road planning, combining regular functional monitoring and maintenance with strategic design, improves both operational stability and flood protection at the watershed level in mountainous regions.
Climate change has altered weather patterns, resulting in longer dry periods followed by heavy rainfall, which is increasingly impairing the water and soil protection functions of our forests. The storage of precipitation water and the delay of runoff processes are fundamental forest functions that should be taken into account in forestry measures in silvicultural and technical forest operations. The opening-up of forests by access roads and the structuring of forest stands by skidding lines contribute to the deflection of precipitation water on the surface and reduce the amount of water that can potentially infiltrate the forest stand. Specifically targeted water retention measures, both silvicultural and, above all, technical, are supposed to catch flood peaks, reduce the risk of soil erosion and contribute to an overall improvement in the water balance of forest stands.
Based on a systematic literature review, natural factors affecting the hydrological water balance and technical factors relating to forest access were analysed in order to illustrate the relationship between water balance, surface runoff, water retention and forest road networks.
In a recent project study, GIS-supported local drainage lines were recorded, runoff volumes were modelled, sensitive areas were identified, and finally technical measures for water retention were planned and dimensioned. The results of this case study are reported here.
Climate change mitigation requires transformative changes in high-emission sectors, particularly the construction industry, which is responsible for a significant share of global greenhouse gas emissions. The increasing interest in wood-based construction reflects its potential to reduce the global warming potential (GWP) of the built environment while strengthening forest-based value chains.
This paper presents a structured literature review of existing research on timber construction as a climate change mitigation strategy. Drawing on life-cycle assessment (LCA) studies, forest operations research, and industrial case analyses, the review synthesizes evidence on three primary mechanisms: (1) biogenic carbon storage in long-lived wood products, (2) substitution effects that reduce embodied emissions compared to concrete and steel, and (3) innovation in forest mechanisation and supply chain systems that improve efficiency, traceability, and adaptive capacity.
The review also examines how expanding wood-based construction can enhance the resilience of forest value chains by stimulating sustainable forest management, promoting technological modernization in harvesting and processing, and supporting rural economies. At the same time, the literature highlights critical challenges, including sustainable sourcing, biodiversity safeguards, carbon accounting uncertainties, and the need for digital integration across forest-to-building systems.
By consolidating current scientific evidence, this study positions wood-based construction as both a climate mitigation strategy and a systemic innovation capable of enabling resilient forest value chains. The findings contribute to ongoing discussions in forestry mechanisation, engineering, and sustainable forest industry development under changing climatic conditions.
Greenhouse gas emissions from forest management operations are currently gaining global attention for tackling climate change by reducing the carbon footprint of forestry activities. This study aimed to quantify the carbon footprint of a full rotation teak (Tectona grandis) plantation, identifying key emission sources in silviculture and harvesting This study assessed the carbon footprint of teak plantation management in the Mae-Moh Forest Plantation, Lampang Province, Thailand, over a 30-year period. The analysis consisted of two major components: silvicultural practices and harvesting operations. The life-cycle emissions of carbon dioxide from silvicultural practices were estimated at 1726.5 kg CO₂eq ha-1, with annual firebreak construction 1176.0 kg CO₂eq ha-1 and followed by fertilizer application accounting at 338.0 kg CO₂eq ha-1. The first year of establishment generated the highest emissions owing to the implementation of all silvicultural activities. Harvesting operations resulted in 14.2 kg CO₂eq/m³ in final felling, with short-distance transportation identified as the largest contributor (5.3 kg CO₂eq/m³). When aggregated over the full rotation, the 15th year exhibited a peak in annual emissions associated with plantation management. Potential mitigation strategies include optimizing machinery size and operation, implementing regular equipment maintenance, and applying fertilizers according to the soil nutrient status and plant-specific demand. Therefore, an integrated nutrient management system that emphasizes organic amendments is recommended to reduce the carbon footprint of teak plantations.
In Finland, the forest industry has taken several steps toward more sustainable wood procurement. For example, biogas-powered timber trucks have been introduced, and large companies now require that biodegradable chain oil is used in harvester heads. Biodegradable hydraulic oils entered Finland’s market in the late 1980s and were adopted in forest machinery largely due to industry demand. However, serious malfunctions and mechanical failures were reported, and the requirement to use them was gradually withdrawn in the late 1990s and early 2000s. In Sweden, by contrast, biodegradable hydraulic oils are compulsory in forest machines, guided by a national standard (Svensk Standard SS 155434).
This study examined perceptions, experiences, and future prospects of biodegradable hydraulic oils in Finnish forest machinery. The research combined a nationwide survey of logging contractors with semi-structured interviews of forest machine manufacturers and an importer of biodegradable hydraulic oils. The aim was to assess whether biodegradable hydraulic oils are perceived as a threat or an opportunity in Finland and how historical experiences influence current attitudes.
An electronic survey was distributed in February–March 2025 to more than 900 forest machine members of the Trade Association of Finnish Forestry and Earth Moving Contractors (Koneyrittäjät ry), yielding 121 responses. In addition, interviews were conducted with representatives of Ponsse Plc, John Deere Forestry Oy, Komatsu Forest Oy, and Ajotek Oy.
The results indicate a lack of trust among logging contractors. Of the respondents, 61% considered biodegradable hydraulic oils more as a threat than an opportunity, while 27% viewed them positively. Forty percent had personal experience using such oils, and 83% reported observing or hearing about technical problems. The most frequently mentioned issues were failures of hydraulic pumps and motors, hose breakages, accelerated component wear, and poor cold-weather performance. Only 5% reported clear advantages, such as longer oil change intervals and broader operating temperature ranges.
Cost and reliability were key concerns. Biodegradable hydraulic oils were perceived as substantially more expensive than mineral oils, and their performance was often considered inferior. Nearly half of the respondents stated they could not imagine using them in the future unless reliability improved, prices became competitive, and performance matched that of mineral oils.
Manufacturers and the importer emphasized both risks and opportunities. Biodegradable oils can reduce the environmental impacts of hydraulic leaks and improve the sector’s public image, but successful adoption requires careful product selection, adherence to recommendations, and investments in maintenance infrastructure. Negative experiences with early vegetable-based oils in the 1990s continue to shape attitudes and contribute to persistent distrust.
The findings suggest that biodegradable hydraulic oils represent both a threat and an opportunity, depending on stakeholder perspective. Wider adoption in Finland will require independent research, testing under local climatic conditions, harmonized standards, and clear economic incentives. Rebuilding trust is essential for enabling environmentally sustainable technological transitions in forest operations.
Mixed species forests have been increasingly recognized as an important pillar of sustainability driven ecosystem management under uncertain and more frequent extreme environmental and social conditions. Although knowledge about the resilience of mixed stands has remained somewhat fragmented, the growing frequency of climate related risks has indicated that the importance of mixed stands is increasing and that the long term resilience of mixed forests is generally higher than that of pure stands. This review synthesized available scientific research on the diversity of benefits when replacing pure stands with mixed forest stands, especially focusing on stand productivity, resilience, and product diversity in mixed stands compared with pure stands, integrating findings from different forest ecosystems. The review concluded that mixed stands generally achieved higher stand level productivity through more diverse and complementary resource use, sometimes resulting in increased growth efficiency. Furthermore, the review found that mixed stands expanded the range of available forest products, improving timber assortment diversity, supporting higher biodiversity (including fauna), and increasing the diversity of non timber products such as fruits and other edible, medicinal, and bio based materials. By combining international research with regionally relevant insights, the review added value for forest management, where the improved resilience associated with species mixing has been acknowledged but insufficiently examined in relation to forest value chains and long term ecological and economic stability. The synthesis also contextualized the important role of replacing pure stands with mixed stands for strengthening forest ecosystem services, enduring climate related stressors, and supporting a more diversified and resilient forest based silviculture.
Growing environmental concerns and increasing pressure on natural resources have intensified the demand for robust sustainability assessment tools in industrial sectors. Life Cycle Assessment (LCA) enables the systematic quantification of material and energy inputs, as well as associated environmental emissions, across the entire life cycle of products and technologies. In mechanized forest operations, LCA provides a comprehensive framework for evaluating the environmental implications of emerging solutions and informing the transition toward lower-emission technologies, such as hybrid harvesting systems.
This study evaluates the environmental performance of hybrid harvester engines compared to conventional diesel-powered harvester engines using a cradle-to-grave approach. The system boundary includes component manufacturing, machine assembly, operational use over the engine lifetime, maintenance, and end-of-life treatment. The functional unit is defined as one engine over its operational lifetime. The LCA model was developed in SimaPro using the Ecoinvent database and IPCC characterization factors. Environmental impact categories considered include global warming potential (GWP), acidification, eutrophication, human toxicity, and resource depletion.
Field measurements from operational trials in Eastern Canada indicate that hybrid harvesters consume approximately 20% less diesel fuel under comparable stand and working conditions as compared to diesel harvesters. Preliminary results for the conventional diesel engine show that the use phase dominates total life cycle impacts, contributing most emissions across impact categories. Although hybrid engines require additional components such as electric motor and power pack units, preliminary scenario modeling suggests that operational fuel savings offset these additional embodied impacts over the engine’s lifetime. These early findings indicate that hybrid harvesting technology has strong potential to reduce the environmental footprint of mechanized forest operations, particularly where full electrification remains constrained.
Globally, road transportation is a major source of greenhouse gas (GHG) emissions. One possible option to lower GHG emissions of road transportation is to use alternative fuels, for instance liquefied biogas (LBG) and electricity, on vehicles. During recent years, LBG-powered trucks have grown in number even on the road transportation of the forest sector, such as timber trucks and woodchip trucks. In this study, we utilized a holistic approach to consider all possible impacting variables to explain energy consumption and further GHG emissions.
Dataset consisted of vehicle information from seven heavy-duty (gross vehicle weight of 75 tonnes) woodchip truck combinations. Three of the trucks were LBG-powered and rest were diesel-powered vehicles (references). Additionally, multiple different data sources (i.e., weather information, route properties and driver-related variables) were included in the study. The dataset had 7310 payloads and the total mileage of the vehicles during the study was over 1.3 million kilometers.
According to our preliminary results, on transportation cycle, including 100 km empty driving, loading payload, 100 km laden driving and unloading payload, the LBG combinations used 6% more energy compared to the diesel vehicles when a measurement of MJ tkm-1 was applied. However, the LBG combinations had 60% lower emissions (g CO2eq. tkm-1) compared to the diesel vehicles when the default emission factors of the ISO 14083:2023 standard were utilized.
Moreover, the variables affecting energy consumption were detected. From weather variables rain occurrence, ambient temperature and snow cover accumulation on the road had statistically significant influence on energy consumption. Furthermore, from road properties, mean grade between starting and ending points of the trip and elevation variation during the route affected energy consumption. Respectively from driver-related variables, time since the driver has completed eco-driving course for Driver Certificate of Professional Competence renewal influenced energy consumption.
Efficient wood logistics is essential for minimizing operational costs and reducing greenhouse gas (GHG) emissions. In Poland, several symptoms of inefficiency—including the fragmentation of forest operations leading to frequent machine relocations, extended transport distances, and the need to move machines to designated parking locations after each working shift—highlight the need for a systematic and comprehensive assessment. To date, existing studies have predominantly focused on individual supply chain segments or specific stakeholder perspectives. A holistic assessment of the wood supply chain efficiency in Poland remains limited.
This study assesses the efficiency of the wood supply chain in two Regional Directorates of the State Forests distinguished by contrasting spatial forest structures. The first one is characterised by concentrated forest stands covering 294 thousand hectares, with 56 thousand cutting treatments per year resulting in 1.2 million m3 of harvested wood supplying 170 customers. The second is characterised by fragmented stands covering 188 thousand hectares, with 30 thousand cutting treatments annually resulting in 746 thousand m3 harvested supplying 181 customers. The analysis encompasses the entire process, from harvest planning to delivery to the processing plants, including harvesting, extraction, storage, assortment allocation to mills, and transportation. The study is based on historical data from 2019–2023, including the spatial distribution of forest compartments, harvested volumes by assortment, silvicultural treatments, the road network, and the demand for assortments from wood processing plants located throughout the country.
The business-as-usual (BAU) scenario derived from historical data was compared with three levels of improvement scenarios using a simulation–optimization approach implemented in ArcGIS Pro 3.4.0. The scenarios are structured hierarchically according to their decision-making impact: Level A addresses strategic logistics, Level B focuses on tactical infrastructure, and Level C deals with operational routines. Level A (three variants) examines alternative allocations of loading points to processing plants to minimize transport distances and applies harvesting concentration strategies to reduce machine relocations. Level B (two variants) evaluates different configurations and densities of loading points (storage sites). Level C (two variants) analyzes alternative machine destinations after the end of the working shift.
In total, twelve scenarios, including the BAU scenario, were developed and evaluated in terms of total operational costs, incorporating GHG emission costs per m³ of harvested, extracted, and transported timber across all feasible equipment combinations under Polish conditions. The results of this study will be presented and discussed, with particular emphasis on the performance of the analysed scenarios and the challenges associated with implementing the most cost-effective solutions.
Wood supply chain and transportation logistics are critical drivers of resilient forest value chains, enabling the reliable, sustainable, and cost-effective flow of a unique renewable resource. In recent years, operational volatility has increased due to disturbance-driven salvage wood peaks, tighter labor markets, and rising expectations about emissions. Despite the sector’s relevance to resilience-oriented innovation, empirical findings on timber trucking remain fragmented and limited to single-country settings, thereby constraining the availability of transferable evidence for research, management, and policy.
We present insights from a representative empirical survey of more than 300 wood transportation enterprises across eight European countries. Our study was designed for meaningful cross-country comparison while respecting national institutional and operational contexts in Austria, the Czech Republic, Finland, Germany, Italy, Latvia, Poland, and Portugal. The selected countries reflect Europe’s diversity in forest ownership structures, workforce conditions, climatic zones, disturbance histories, transport infrastructures, and modal splits. Our overarching objective is to provide a robust empirical baseline on constraints, practices, and transition priorities in timber transportation logistics, with an explicit focus on resilience-relevant leverage points.
Our survey development followed a structured co-design process. First, we established a shared conceptual approach linking transportation logistics to resilience and sustainability. This framework included contingency planning, coordination interfaces, regulatory conditions, workforce constraints, and digitalization. Second, we drafted a standardized questionnaire with more than 60 questions. We refined it through multiple rounds of feedback from national experts and stakeholders. Third, we piloted the instrument to ensure measurement equivalence across countries, translated it into national languages, and deployed it as a harmonized online survey.
Across Europe, entrepreneurs report a strong convergence in perceived bottlenecks. Timber trucking resilience is widely constrained by economic pressure, persistent labor shortages, and an aging driver workforce. Infrastructure limitations, especially at the interface between forest roads, terminals, and stockyards, are frequent. Preparedness for disturbance-driven peak volumes is often perceived as insufficient. Disruption risks are well recognized, yet contingency planning and coordinated crisis routines across value chain actors remain underdeveloped. Respondents identify actionable transition pathways. Stronger vertical coordination between forest owners, harvesting, carriers, and industry, together with horizontal collaboration among transport actors, is seen as essential to stabilize flows and improve utilization. Multimodal and multi-echelon unimodal transport are viewed as strategic options where terminal access and service quality are adequate. These approaches offer additional capacity for longer distances and peak periods, and support emissions reduction.
Building on these insights, interconnected leverage points emerge. These include faster digitalization and data-driven planning; workforce development, training, and retention; crisis-ready governance with clear routines and pragmatic regulatory flexibility during disturbances; investment in infrastructure and buffering capacity enabling operational flexibility and multimodal integration; and targeted investment support for modernization, green technologies, and collaborative innovation partnerships. Together, these pillars can translate recurring bottlenecks into resilience-oriented breakthroughs in European timber transport logistics.
Forest operations managers handle multiple sources of uncertainty when planning wood supply. Wood sourcing for Nordic forest industries typically include both internal and external suppliers, where the first-line for solving operational deviations from delivery plans involves re-planning of internal supply operations.
This paper presents an adaptation of the typical monthly wood supply plan in order to quantify the risk of varying supplier portfolios.
The approach is used in masters-level instruction in industrial wood supply. The student exercise is done in 3 steps over 2 teaching days. The first step consists of selecting a supply structure aimed to secure year-round supply to a generic pulp mill. The mill supply structure is defined by students’ selection of annual wood orders from 5 supplier categories (2 internal and 3 external). An initial tactical plan is drafted for monthly deliveries per supplier category to maintain monthly stock intervals. At this stage the planning is purely determinstic, and can be assisted by an initial optimization, including consideration to seasonal trends for wood availability. The second step is then to test the draft plan using Monte Carlo simulation with supplier-specific distributions of monthly demand fulfillment. The resulting frequency of stock levels exceeding defined goal intervals is then used to adjust the annual order from the respective supplier categories. After adjustment, the third step is to explore internal responses to reduce the risk for exceeding the goal stock intervals. The response options include i) time horizon for response and ii) strength of response.
The distributions of monthly demand fulfillment (%) for the respective supplier categories are described with a 2-parameter gamma-distribution. The estimation of the gamma parameters is based on 2 years of pulpwood deliveries to Swedish SCA from 25+ external supply organizations and 6 internal supply regions. Multivariate clustering was used to segment the suppliers into six main categories, including spot contracts. The respective distributions were estimated in relation to two base-lines: i) originally contracted monthly volumes and ii) revised monthly contract volumes.
The approach enables adaptation of a typical monthly master plan to simulate various supply system risks. The approach serves well as a practical introduction to understanding uncertainty and risk in wood supply.
To meet climate goals and limit global warming there is an urgent need to decarbonize road freight transport, including the forestry sector’s transport. Previous studies have shown that electrification is an effective means for carbon abatement in freight transport and will play a key role. In the project TREE (Transition to efficient electrified forestry transport) 12 electric timber and wood chip trucks are operated in commercial operation at seven different sites in Sweden, illustrating the feasibility to electrify forestry sector’s transports. The TREE project has pronounced that access to reliable charging is critical for efficient electrification. A single charger failure, due to technical malfunction, limited power supply, or waiting times at the charging point, can render an entire route infeasible or that an entire region is not accessible. This creates a form of systemic fragility, where the operational viability of electric trucks is highly sensitive to disruptions at individual charging points. Such vulnerabilities reduce system reliability, increase operational risk, and may slow down the adoption of electric trucks. In this study, we aim to analyze and improve the robustness of charging infrastructure for heavy electric trucks in forestry transport. We propose indicators for assessing the robustness of a charging network. To assess the robustness of a charging network, we combine a fleet planning tool for mixed fleets of electric and diesel trucks with Exploratory Modeling and Analysis (EMA) to run many simulations with different potential charging failures implemented and different combinations of trucks. Based on the results, we explore the robustness of a charging network in a region. The aim is to use the results to propose improvements in the charging network.
In Finland, thinning operations are carried out annually on approximately 0.5 million hectares. Due to the labour‑intensive and costly nature of field inspections, official harvesting quality assessments produced by the Finnish Forest Centre are conducted on less than 1% of this area. For many years, the aspect with the greatest need for improvement has been thinning intensity. Harvester operators also require real‑time information and feedback systems, because feedback on thinning quality is given in many cases after weeks or even months after the operation. It is assumed that harvesting quality can be improved through tree‑level operator guidance, while at the same time enabling the collection of large amounts of forest‑related data.
This study evaluates the feasibility of conducting follow-up harvesting quality assessments using only digital data sources: StanForD 2010-based harvested production (hpr) messages and stand-level tree maps derived from Ponsse Thinning Density Assistant (TDA) point clouds combined with external forest information databases (Metsakanta.fi service and Metsaan.fi data provided by the Finnish Forest Centre). A new methodology is introduced to automatically assess harvesting quality indicators, including residual stand structure, thinning intensity, and strip road geometry along machine trails.
The objective is to determine whether harvesting quality can be digitalised and assessed at scale without field measurements, enabling coverage of all harvested stands instead of small inspection samples. Additionally, the study examines whether the use of TDA-based data collection supports improved harvesting quality through operator guidance and feedback. By linking quality assessment with machine-generated data streams, the proposed framework represents a step towards scalable, transparent, and adaptive monitoring of forest harvesting operations.
Reliable forest stand delineation is a fundamental prerequisite for forest management, harvesting system allocation, and spatial planning. Conventional stand boundaries are typically derived from visual interpretation in the field or from aerial imagery, making them time-consuming, experience-dependent, and only partially reproducible. In response to increasing disturbance dynamics and the demand for consistent spatial planning units, this study develops and evaluates a reproducible, data-driven approach for automated stand classification and segmentation based on airborne 3D data.
The methodological framework integrates airborne laser scanning (ALS), UAV-based laser scanning (ULS), and image matching from aerial imagery as alternative data sources. The workflow was tested in two independent Austrian study areas where all three airborne systems were acquired within a narrow temporal window: a 130-hectare forest management unit in Styria and a larger test area exceeding 1.000 hectares in Tyrol. Structural and site-related predictors were aggregated on hexagonal grid cells (314 m²) derived from point clouds and digital terrain models. These included vegetation height quantiles, mean and median canopy height, height variability, elevation, slope, aspect, and distances to forest roads and watercourses.
Forest development stages were classified using gradient boosting models trained on independent inventory data. Model performance was assessed via k-fold cross-validation. In the Styrian test area, strict class agreement reached 56.3% (Cohen’s kappa = 0.49). Allowing a deviation of one adjacent development stage to account for structural transition zones increased overall accuracy to 82.3%. Moreover, the automated classification resulted in a 22.7% increase in within-class structural homogeneity compared to conventional stand delineation by forest experts, measured as a reduction in intra-class variance of canopy height metrics. Important predictors were the 75th percentile of vegetation height, median vegetation height, and elevation-related variables.
Pixel-level predictions were subsequently aggregated into operationally usable compartments using spatially constrained hierarchical clustering and graph-based region-growing algorithms. Small and isolated patches were reassigned to neighboring units, resulting in coherent and practically manageable stand polygons. The standardized workflow enabled direct comparison of data-source effects. ALS ensured homogeneous coverage over large areas, ULS provided higher structural detail at stand scale, and image matching represented a cost-efficient alternative but showed limitations in canopy penetration and terrain representation, influencing predictor stability and class separability.
The resulting digital stand maps provide a standardized spatial reference for deriving forest metrics such as growing stock, stem density, and diameter distributions. These compartments support harvest planning and facilitate rapid map updates following disturbance events. Overall, the integration of airborne 3D sensing and machine-learning-based segmentation provides a quantitative and scalable foundation for data-driven forest management and operational planning.
Accurate tree species identification in young boreal forests is critical for sustainable forest management, biodiversity monitoring, and bioenergy resource assessment. However, current remote sensing approaches predominantly rely on single-sensor methodologies, limiting classification accuracy in complex young stands where trees exhibit similar heights, small crowns, and subtle spectral differences.
This research aims to develop and validate ForestFusionNet, a scalable multimodal deep-learning framework that fuses LiDAR-derived 3D structure, high-resolution digital aerial photogrammetry (DAP), Sentinel-2 multispectral time series, and Planet high-resolution multi-temporal imagery to classify tree species at the individual-tree level in young boreal forests. The work pursues two main objectives: (1) to build a harmonized multi-sensor dataset and extract complementary features, and (2) to design a unified fusion deep-learning architecture and rigorously evaluate its accuracy and computational efficiency.
The study will be conducted in two young boreal forest sites in Juva (27.865E, 61.730N) and Heinävesi (28.736E, 62.509N), Finland. The proposed ForestFusionNet architecture will employ sensor-specific encoders: a point-cloud network for LiDAR and a CNN for aerial imagery.
The proposed ForestFusionNet architecture will utilize sensor-specific encoders, including a point-cloud network for LiDAR, a CNN for aerial imagery, and a temporal module for satellite time series. Different fusion mechanisms (late, intermediate, and deep) will be explored to combine modalities optimally. The model will be trained and validated using field-collected individual-tree reference data. The model's performance will be compared with single-sensor results by evaluating how well it identifies each species using metrics such as precision, recall, F1-score, and overall accuracy. Ablation studies will quantify the contribution of each modality, and computational costs (training/ inference time, memory usage) will be reported to assess deployability for large-area mapping.
We anticipate that ForestFusionNet will achieve significantly higher classification accuracy compared to any single-sensor approach by leveraging the complementary strengths of 3D structure, high-resolution texture, and phenological spectral patterns. The project will deliver (i) an open-access harmonized multi-sensor dataset hosted on Zenodo; (ii) the complete ForestFusionNet code released on GitHub; (iii) empirical evidence on the accuracy-efficiency trade-offs of multi-sensor fusion; and (iv) a scalable operational framework for precision inventory of young boreal forests.
The results will directly benefit forest owners, small- and medium-sized enterprises, and policymakers by enabling cost-efficient, high-precision forest inventory. Moreover, improved individual-tree species information will support more reliable biomass and carbon stock estimates, thereby enhancing national and EU climate reporting under the LULUCF sector and supporting the targets of the European Green Deal.
Keywords: Deep learning, multimodal fusion, tree species classification, young boreal forests, remote sensing
Bark stripping is a major disturbance factor in Central European forests and leads to reduced growth, increased susceptibility to fungal infection, and elevated mortality risk. In operational forest management, severely damaged trees often need to be removed in order to avoid economic losses and secondary infections. Reliable and objective detection of bark stripping is therefore highly relevant for precision forestry and technology supported harvesting decisions. This study investigates whether bark stripping can be detected automatically using Light Detection and Ranging (LiDAR) data and Convolutional Neural Networks (CNN) and evaluates the potential integration of such an approach into LiDAR based forest inventory workflows.
Data acquisition was carried out in Styria in Austria on 37 sample plots with and without visible bark stripping. In total, 71 trees with bark damage were recorded across these plots, containing 172 individual bark stripping wounds. Three laser scanning systems were used, including a terrestrial laser scanner (TLS) Riegl VZ600i, a mobile laser scanner XGrid Lixel L2 Pro (MLS), and a consumer grade device Apple iPad Pro. The TLS point clouds provided intensity, reflectance, amplitude, and RGB attributes for each measured point, whereas the MLS point clouds delivered RGB and intensity attributes, and the tablet based system RGB attributes only.
Tree stems were unwrapped from the point clouds to generate two dimensional stem representations. Bark wounds were manually delineated and served as reference data. From these labeled stem surfaces, 100 × 100 pixel image patches with a spatial resolution of 2.5 mm per pixel were extracted to generate training and validation datasets. Initial model development focused exclusively on TLS data, as this dataset provided the most complete geometric and radiometric information. A pre trained ResNet 18 architecture was fine tuned for binary classification of damaged versus undamaged stem patches.
First experiments using grayscale patches derived from TLS point clouds achieved an overall classification accuracy of 76 percent on the independent test dataset. Precision exceeded recall, indicating a conservative model behavior with a tendency to classify uncertain patches as undamaged. The resulting F1 score reflected this imbalance between omission and commission errors. Ongoing work investigates multi channel input configurations combining RGB and LiDAR derived features as well as larger spatial contexts in order to increase detection sensitivity.
The presented workflow demonstrates the feasibility of automated bark stripping detection from LiDAR data and provides a foundation for integrating damage detection into digital forest inventory and decision support systems. Such an approach can support automated timber harvesting in identifying trees that require priority removal and contributes to the advancement of precision technology and remote sensing in forest operations.
Automated Deep Learning for Forest Resource Mapping to Drive Efficient Operational Planning
The efficient planning and execution of mechanised forest operations rely heavily on near-real-time, spatially explicit intelligence regarding forest resources. While the rapid growth of multi-source data—from national inventories to satellite imagery—presents a major opportunity to optimize these operations, it demands scalable, automated modelling workflows. To bridge this gap, we present an automated deep learning (DL) framework designed to generate actionable maps of forest resources for operational planning. The system ingests Sentinel-2 surface reflectance mosaics, Shuttle Radar Topography Mission (SRTM) terrain data, and climate variables, automatically optimising neural network architectures for two critical operational products: (i) precise estimation of key forest stand attributes—standing stock volume (Vol), basal area (BA), diameter at breast height (DBH), and mean height (H)—and (ii) high-resolution (10 m), national-scale tree species mapping.
Pairing harmonised field plots and inventory data from Romania and Finland with remote sensing predictors via Google Earth Engine, the workflow utilizes autoencoder-based bottleneck representations and Hyperband-driven hyperparameter tuning (via Keras Tuner) to compress high-dimensional inputs and automatically optimize network configurations. On independent test plots, the resulting models achieved R² values of 0.52 (Vol), 0.48 (BA), 0.72 (DBH), and 0.75 (H). Furthermore, the automated tuning strategy yielded a supervised bottleneck multilayer perceptron for species mapping that achieved macro F1 scores above 84% and overall accuracies of approximately 87%. By delivering wall-to-wall, consistently tuned maps of both forest structure and composition, this DL framework provides the foundational spatial data required for efficient forest engineering. The outputs directly support critical operational tasks, including optimized harvest planning, targeted machine and workforce allocation, and efficient road and transport logistics, while offering an extensible basis for operational forest digital twins.
Pack-mules are the archaic logging system that prevailed in the Apennine mountains until few decades ago. Today, despite the dramatic reduction in the number of teams, this system still survives, contrary to all expectations. Given that the system has its merits and cultural value, there is an interest in preserving and enhancing it, perhaps even expanding it.
To do so, however, we must first determine the conditions that have allowed this archaic system to survive to this day, the unique assets that can be leveraged to maintain it, and the hurdles that must be removed in order to support a possible expansion – before all know-how is forgotten and the purposedly-selected breeds degenerate and die.
In order to achieve this goal, CNR teamed up with the main draught animals’ associations active in Italy and developed a network of contacts with mule loggers – active, retired and newly equipped (yes, there are a few rare examples).
The research was based on face-to-face interviews and defined the characteristics of those businesses, the motivations of their entrepreneurs, the possibilities and methods for the preservation and eventual revival of mule logging in Italy.
Focus points were: the integration of animal and mechanized logging; the reduction of physical effort; the importance of animal selection and training; the capacity of achieving a minimum use intensity.
Two-stage trucking systems are increasingly used in steep-land harvesting in Aotearoa New Zealand to reduce the need for the quality of roading and landings required to allow access for on-highway high productivity motor vehicles (HPMV’s). In these systems, stems or processed logs are trans-shipped between a constrained extraction/processing landing and a larger storage and load-out skid using a dedicated intermediate truck. This study synthesises elemental time study findings from nine two-stage systems based on eight different intermediate truck configurations, spanning two widely used platform types: (1) modified 6WD articulated dump trucks, and (2) off-highway/on-highway log trucks (6x8 and 8x8). Data were collected across multiple harvesting contexts (ground-based and cable), with cycle elements separated into loading, travel loaded, unloading, travel unloaded, and delays. Intermediate road distance and grade were measured in segments, and payloads were estimated using truck scales or log-count/volume methods.
Across case studies, haul distances ranged from 350 m to 5,440 m, with average loaded speeds typically between 2 and 7 m/s depending on road profile and surface condition. Productivities ranged from 19 to 44 tonnes per productive machine hour (t/PMH), with utilisation 73–96%. Results indicate that intermediate-truck productivity is primarily determined by (i) haul distance and road profile, (ii) effective payload capacity (including the ability to use a trailer and the degree of load sorting), and (iii) loading/unloading system design—particularly access to loaders and interference with other landing/skid activities. Where long haul distances combined with constrained payload (e.g., inability to use a trailer), travel time became dominant and reduced throughput, sometimes requiring additional truck hours to avoid bottlenecking the prime mover. Conversely, higher payload capacity and/or shorter leads could offset challenging grades, and well-integrated landing/skid layouts reduced interference delays.
These findings support a practical design implication for harvest planners: intermediate trucking should be treated as a system component whose performance is jointly shaped by truck configuration, road geometry, loader availability, and landing/skid layout, rather than by truck platform alone. When these elements are aligned, two-stage trucking can provide access to steep or erosion-prone settings while potentially maintaining or improving overall harvesting productivity and reducing safety and environmental risk.
In 2026, an Italian machine manufacturer developed a new light yarder designed to match the specifications drafted in 2024-2025 by CNR, through a survey of small-scale logging contractor preferences (presented at FORMEC 2025). That study stressed the importance of practicality and low cost, and directed design efforts towards a lightweight, compact machine that can cope with poor infrastructure (narrow landings) and can be moved with a pickup truck. At the same time, the machine must present innovative characteristics, in order to compete with other small-scale yarders and qualify for public support. The result was an extremely light tower yarder, powered by an Otto-cycle engine, which is light, efficient and exempt from many of the regulatory limitations imposed on diesel engines. The new yarder features an electric carriage weighing only 60 kg (including 30 m of 8 mm cable), but capable of carrying a 1000 kg load. To maximize operational flexibility and convenience, the machine is a three-cable all-terrain type and is equipped with its own folding tower. Operation is obtained through a modern remote control. The proposed presentation will offer a detailed description of the machine and of the results from the first field trials, conducted in 2026.
Steep-terrain forestry ($>30^\circ$) demands a radical departure from conventional machinery to balance operational efficiency with soil protection. This study evaluates a dual-solution framework developed by MSE: the "Shin-Raptor II," a specialized 6,400 kg class lightweight winch-assisted (WA) base carrier, and the "Feller Buncher Saurus (FBZ)," an innovative attachment for integrated earthworks and harvesting.The Shin-Raptor II is a specialized platform that redefines steep-slope felling by eliminating the conventional crane arm. Its compact 6,400 kg design, combined with a unique chassis-integrated felling architecture, maintains an exceptionally low center of gravity and minimizes soil impact. Powered by a fully synchronized Winch Assist (WA) system, this lightweight machine achieves superior vertical mobility and high-traction performance on $45^\circ$ slopes where traditional heavy harvesters fail. Its $10^\circ$ engine-oscillation mechanism further ensures operational reliability under extreme tilt.Complementing this specialized carrier is the FBZ, designed to transform standard $0.45~m^3$ class hydraulic excavators into multifunctional infrastructure units. The FBZ uniquely integrates a $0.45~m^3$ excavation bucket with an eccentric pull-cut blade, enabling a single machine to perform slope profiling, stump removal, and road construction while simultaneously executing high-speed felling. This integrated workflow significantly reduces machine traffic and operational costs by eliminating the need for separate road-building equipment. Furthermore, its "pull-cut" geometry allows for the efficient processing of high-density broad-leaved timber. Together, these technologies provide a scalable framework for precision forestry, balancing extreme mobility with comprehensive operational versatility.
The study will present the implementation of a smart monitoring system based on LoRaWAN technology to support the improvement of safety and operational efficiency in cable yarding on steep slopes.
As a case study area, a typical harvesting intervention with cable yarder in the Italian prealps was chosen.
An LTE-enabled LoRaWAN gateway was mounted on the tower yarder to ensure a stable wireless communication across the harvesting site, with a theoretical data transmission capacity of the payloads up to 1km away from the gateway itself. Preliminary network coverage analyses were performed in GIS to investigate how terrain morphology interacted with the signal transmission. To this, a detailed methodology involving the use of visibility analysis tools allowed to evaluate areas potentially lacking
network coverage. The methodology was then validated on field with a two-antenna GNSS-RTK system exchanging corrections with LoRa transmission protocol.
Then, high sensitivity motion sensors (i.e., inclinometers and accelerometers) were installed on key structural components, including the carriage, tail anchor tree, and tail spar, to measure inclination and rotational movements under working conditions. Specific warning and critical thresholds were defined to detect abnormal movements and potential instability. An automated alarm system was integrated with a Telegram chatbot to provide real-time notifications to operators.
The preliminary results confirm the feasibility and robustness of the solution and open to a wider discussion on LoRaWAN technology applied to monitor and connect forest operation in real time also in complex mountain terrain. Canopy occlusion was confirmed not to be a factor correlated to the signal quality. Transmission of the payloads were possible up to 500m from the tower yarder. Further improvements as the use of a directional antenna, or changing the SDR of the radio transmission may
further extend this approach to even longer distances.
Ongoing climate change and associated damaging events are leading to a significant need for large-scale forest regeneration. The key challenge is to efficiently address the increasing demand for reforestation despite a shortage of skilled workers by means of automated processes. In Germany, reforestation is still predominantly carried out manually and is associated with high labour intensity. First planting machines as well as excavator-based planting attachments are available; however, due to limited productivity, their adoption in operational practice remains low. In parallel, approaches for drone-based seeding are being explored, but these currently achieve only limited establishment success due to insufficient soil contact and the lack of precise positioning.
The establishment of the next tree generation represents the starting point of resilient forest value chains. Here planting operations are of central importance. Automated and digitally supported planting systems offer the potential to perform this key task more efficient, reproducible, and less labour-intensive.
Against this background, the ForestBots research project, funded by the German Federal Ministry for Research, Technology and Space within the funding initiative “Digital GreenTech – Environmental Technology Meets Robotics” (FKZ: 02WDG1760A), addresses the development of a collaborative system for automated reforestation. The objective of the project is to demonstrate a semi-autonomous forest operation using a small swarm of fully electric tracked vehicles for seedling planting.
This contribution focuses on the systematic analysis of the process chain as a foundation for the development of automated and digital planting systems based on autonomous mobile robots (AMR). The requirements for an automated planting process were derived based on a comprehensive analysis of the state of research and technology, a patent and market review, as well as expert interviews and workshops with forest nurseries and forestry stakeholders, including the Lower Saxony State Forests and the Chamber of Agriculture. These include, among others, seedling characteristics, operational timing, and relevant environmental and site conditions.
Based on this, the overall process of seedling planting was broken down into its essential functional subprocesses. These include system coordination at the global level, seedling handling and logistics, as well as a detailed analysis of the planting process itself. For each subprocess, dependencies, interfaces, and critical tasks with particular relevance for automation were identified.
The results provide a structured description of functions and processes as a basis for further concept development and for the integration of sensors, actuators, and data-driven decision-making. The study thus contributes to the development of robust, semi-autonomous, and digitally connected systems for reforestation.
Forestry operations are facing increasing structural challenges, including a shrinking skilled workforce, heightened safety requirements, and rising expectations for productivity with reduced environmental impact. These create a clear need for automation technologies that can support or eventually perform demanding tasks autonomously. In 2021, our AORO platform marked a significant milestone as the first fully autonomous unmanned forestry machine demonstrated live. While AORO successfully achieved autonomous navigation and partial crane automation in a forwarding task, it lacked the capability to autonomously complete the full log handling cycle. In particular, the crane could not operate all degrees-of-freedom autonomously due to the absence of grapple sensing and a dedicated vision system, restricting the system’s ability to execute a complete log handling cycle that required complex motions of the grapple.
This work presents a new add-on system that closes this gap and enables full control over all crane’s degrees-of-freedom. The solution extends the AORO machine with an integrated system that combines log detection, precise crane control, and coordinated motion planning. Together, these components allow the machine to perform a complete work cycle: navigate to a work area, identify scattered logs, grasp them securely, and pile them in an organized manner along the roadside.
To perceive the work environment, a dedicated stereo camera is mounted on the crane, providing a clear view of the logs within the operating range. A Convolutional Neural Network (CNN)-based detection model was trained to recognize and localize individual logs under realistic field conditions, including variable lighting and cluttered backgrounds. Once detected, the position of each log is estimated relative to the crane base using calibrated geometric transformations. This ensures that the crane can directly act on the detected targets without additional adjustment.
To achieve reliable and precise handling, the crane control system has been upgraded to operate all joints simultaneously. The grapple has been equipped with new sensors to provide direct feedback: a wired encoder measures rotation, and a wireless sensing solution monitors claw position. These additions significantly improve grasp accuracy and allow the crane to coordinate its movements more effectively during both pickup and placement.
A new motion planning approach based on S-curve velocity profiles coordinates all crane joints throughout each movement. This ensures smooth acceleration and deceleration, reducing abrupt motions that can destabilize logs or impose unnecessary stress on the machine. The result is controlled, efficient motion both when approaching a log and when transporting it to the piling location, contributing to stable handling and reduced mechanical wear.
The integrated system was validated in a public demonstration in which the unmanned machine autonomously navigated along a forest road, detected randomly scattered logs, grasped them, and piled them to one side of the road. The full cycle, from perception to final placement, was executed without human intervention. The demonstration confirmed robustness under realistic field conditions and provided tangible proof of system effectiveness, closing a critical gap in AORO’s manipulation autonomy.
Forest environments pose significant challenges for automated driving due to dense vegetation, irregular terrain, varying illumination conditions, narrow skid roads, and limited or unreliable GNSS availability. In contrast to structured road environments, skid roads often lack clear boundaries and may contain unpredictable obstacles such as branches, uneven ground, and irregularly spaced trees. Reliable automated driving in such environments therefore requires robust perception, mapping, and navigation components that remain stable under highly unstructured and dynamically changing conditions.
To address these challenges, a multi-modal perception and navigation framework has been developed and deployed on a full-scale forestry machine. The framework integrates deep learning-based visual perception with LiDAR-based mapping and localization and extends the navigation pipeline with occupancy-aware path planning.
For semantic perception, a state-of-the-art YOLO26 model is trained using a combination of newly annotated forestry datasets and previously collected data. The dataset includes skid roads, tree trunks, and typical environmental structures encountered during timber extraction. The trained model is deployed directly on the forestry machine and enables real-time detection while driving along skid roads. This visual module provides semantic understanding of the driving corridor and surrounding objects, supporting corridor recognition and obstacle awareness under varying lighting conditions and partial occlusions, thereby establishing a foundation for vision-supported driving automation functions in forestry applications.
In parallel, a LiDAR-based landmark extraction method was implemented for geometric mapping and localization. Trees with distinct cylindrical trunk characteristics are detected in real time and parameterized as geometric primitives. These tree landmarks are incorporated into a SLAM framework to build a consistent map and estimate vehicle pose during operation. While this landmark-based approach provides accurate localization when trunk geometry is clearly observable, performance can degrade in scenarios where trunk features are irregular, partially occluded, or less distinguishable.
To improve robustness, an occupancy grid mapping module was introduced as a complementary environmental representation. Rather than relying exclusively on explicit trunk landmarks, the occupancy grid captures the spatial distribution of nearby objects along skid roads and provides a denser representation of occupied space. This representation compensates for situations in which landmark detection becomes unreliable and offers improved environmental awareness in complex forest conditions.
Building upon the perception and mapping modules, a multi-modal path planner, referred to as FlowPlanner, is developed. The planner integrates information from tree landmarks, tree clusters, and reference paths to generate a combined guidance flow field. Based on this flow representation, multiple candidate trajectories are generated online and evaluated with respect to alignment, safety margins, and kinematic feasibility. The most suitable candidate is selected as the local goal path. This approach enables real-time generation of feasible driving trajectories and supports exploratory navigation in partially unknown forest environments.
Initial field experiments in real forest conditions indicate the functional feasibility of the proposed framework. The modular system architecture supports incremental refinement and integration of additional perception and planning components, providing a scalable foundation for progressive automation in demanding forestry operations.
Forestry cranes play a critical role in mechanized logging operations, yet their core mechanical architecture has changed little over the past decades. As a result, they consume substantial energy, much of which is continuously required to support their own structural mass and payload against gravity, even when not performing productive work. While prior research has focused extensively on automation, trajectory planning, advanced control strategies, and improvements in hydraulic and actuation systems, these approaches primarily optimize motion execution rather than addressing the underlying mechanical sources of energy demand.
This work proposes a model-based optimization framework to minimize overall operational energy consumption. Passive mechanical elements are incorporated as design variables within the mechanical architecture, enabling systematic exploration of energy-efficient configurations. Although such elements are widely used in robotics to improve energy efficiency through gravity compensation, their integration into forestry crane design remains largely unexplored.
The optimized configurations are experimentally validated using reduced-scale crane prototypes, enabling direct comparison with the original design. Results demonstrate energy reductions exceeding 45% during dynamic tasks while maintaining functional performance and operational capability.
These findings show that substantial gains in energy efficiency can be achieved by rethinking the mechanical architecture of forestry cranes, rather than relying solely on improved control strategies or actuation technologies. Integrating passive elements into the structural design significantly reduces inherent energy demand, offering a practical pathway toward more sustainable and energy-efficient forestry operations.
Our project focuses on the development of a teleoperation system for a forestry delimber operating in full-tree applications in Canada. In many regions, full-tree processing involves repetitive tasks: feeding stems, delimbing, cutting to prescribed lengths, handling, and stacking. These operations require precision, are physically demanding, and take place in remote environments. The forestry sector faces a persistent shortage of skilled operators, limiting production capacity and increasing pressure on crews. Operational costs are significant, particularly due to long distances between cities and harvesting sites, often exceeding 120 km. Transporting operators to remote sites adds logistical complexity and expense.
A major constraint is the lack of reliable cellular connectivity. Harvesting sites are commonly located outside LTE or 5G coverage in Canada, preventing the use of conventional teleoperation solutions that depend on terrestrial infrastructure. This connectivity gap, combined with labor shortages and high costs, motivates the need for an alternative approach.
To address these challenges, we are developing a satellite-based teleoperation system for forestry delimbers. The objective is to enable a remote operator to control a machine in real time from an operations center near a mill or service hub. By centralizing operators in a controlled environment, this approach reduces travel time, improves working conditions, optimizes workforce allocation, and enhances operational efficiency. It also extends opportunities for experienced operators who may no longer wish to work full-time in remote forest conditions.
A key technical challenge lies in communications performance. Teleoperating a forestry machine requires continuous transmission of video streams and sensor data, including onboard cameras and lidars used for environmental perception and positioning assistance. Satellite networks introduce constraints in bandwidth and latency. To ensure feasibility, we developed advanced data compression and stream optimization mechanisms. The system architecture maintains low end-to-end latency to preserve precise and safe machine control while minimizing bandwidth usage. Our operational target is to keep total data consumption below 4 terabytes per month, including video and command streams required for teleoperation.
The platform incorporates adaptive bitrate strategies that adjust stream quality according to network conditions. Safety mechanisms guarantee predictable machine behavior in the event of communication degradation or temporary link interruption. Edge processing capabilities are integrated on the machine to prioritize critical control data over auxiliary sensor streams, ensuring stable command responsiveness under fluctuating satellite link conditions.
An initial validation phase is conducted using a Caterpillar delimber equipped with our teleoperation system on a controlled private site. These trials evaluated sensor integration, satellite communication robustness, control stability, and overall system performance under realistic conditions. Quantitative metrics collected during testing included end-to-end latency, packet loss resilience, and bandwidth utilization.
The next phase includes operational trials in an active forest environment scheduled for summer 2026, assessing performance under real production constraints and varying weather conditions.
This project demonstrates that reliable, safe, and economically viable satellite-based teleoperation of forestry delimbers is achievable in remote environments, supporting a structural evolution in heavy forestry machinery operations.
According to our Roadmap 2035 Vision for Forest Machine Data and Sensor Technologies in Finland, forest resource data in 2035 will be substantially more accurate in terms of tree volume, quality, changes, harvestability and soil information. Accurate, single-tree interpreted forest resource data will provide high-quality pre-operational information for cost- and energy-efficient, climate-resilient, low-carbon planning and implementation of the wood supply process. These pre-operational data include information collected in advance via national airborne laser scanning programs, drones and other unmanned aerial vehicle platforms. By 2035, development accelerated by artificial intelligence (AI) will have introduced more automated functions into the wood supply planning systems, thereby enabling optimal selection and timing of harvesting sites. Digital forest twins will allow AI-optimised advanced planning of forest machine operations, taking into account harvesting conditions and seasonality. They will also enable the transfer of optimised harvesting plans with single tree maps to systems that assist machine operators.
By 2035, precision positioning will be the standard for all harvesters in Finland. Forest machine sensors incorporating laser scanning and machine vision will increasingly enrich the pre-operational data sent to machines. While they are not yet universal across the entire Finnish fleet, these technologies are becoming fundamental to modern forest operations. Pre-operational data will also be augmented in real time using StanForD 2010-based forest machine data. Simultaneously, forest machine manufacturers will provide automated assistant systems that will enhance operator performance, improve harvesting quality and support operator well-being. These systems will also help address the shortage of skilled operators. For example, tasks such as cutting strip road networks on harvesting sites, monitoring thinning intensity in real time, identifying poor-quality or diseased trees, preserving biodiversity-relevant trees and planning forwarder payloads will be almost fully automated, requiring minimal attention from the operator, who will act more as a process supervisor.
By 2035, the quality of harvesting operations will be transparently reported to forest owners and authorities. Authorities will no longer conduct nationwide field inventories themselves; instead, field data will be automatically generated during harvesting and forest management operations with single-tree accuracy. After forest operations, harvested timber data will be transferred seamlessly to mill customers, thereby enabling batch-level traceability of industrial roundwood. Guidelines and rules regarding ownership and use of forest machine and sensor data will be updated, and data will be actively collected, shared and utilised according to established recommendations in Finland.
Implementation of the Roadmap 2035 Vision requires coordinated actions from multiple stakeholders across the wood supply value chain. Key actors include forest companies, forest machine contractors and manufacturers, forest owners, and forest and environmental authorities and research scientists. To measure and document forest biodiversity and assist the operators that use machine and sensor data, actions and investments are needed to define metrics, develop and standardise technologies and processes, and ensure competence. Many technologies can be adopted relatively quickly for full-scale operational use. Some technologies, however, will require prominent RDI investment to be ready within the ten-year target. Achieving this Vision demands commitment and a forward-looking mindset from all stakeholders.
Heavy forestry machinery like harvesters or forwarders may damage machine operating trails during their operation by causing rut formation. This damage is mostly due to a combination of the loaded weight of these machines and soil condition, which can also result in increased wheel slip. Therefore, having a method to calculate slip in real time will result in a better understanding of caused soil damage.
The purpose of this presentation is to estimate wheel slip of forest machines in real time on machine operating trails by proposing a combined radar and wheel-encoder approach. The proposed solution is independent of machine make and model since it relies completely on external sensors making it modular for integration into existing systems. The core of this approach relies on using one wheel encoder on each side of each bogie axle to measure wheel rotation speed. A radar is then used to measure the real velocity of the ego vehicle, which is independent from wheel slip. As of current knowledge, there are no control systems on forest machines that use RADAR to calculate ego velocity. By evaluating both of these values, wheel slip can be numerically computed in real time. The evaluation of this method is currently being assessed in an ongoing project.
By providing real time slip calculation, advanced systems (like operator assistance or autonomous driving modes) can utilize it and indicate if operation of the machine can continue or if it should be halted, as predicted rut depth exceeds regulatory limits.
Despite increasing fully mechanized timber harvesting, motor-manual operations are still required, for example for the selective removal of high-value individual trees, in sensitive stands or difficult terrain, in calamity patches, or in small private forests. Besides physical strain, these motor-manual tasks entail an elevated accident risk for workers, caused for instance by deadwood, limited experience in assessing trees, or unsuitable choice of tools. While the flow of information along the wood value chain is already digitally supported in harvester–forwarder systems through standards such as StanForD2010, ELDATsmart, or papiNet, many steps in motor-manual harvesting are still recorded by hand and are frequently passed on with media discontinuities.
The research project SmaSiKaFE (Smart and Safe Calamity Area Harvesting) addresses these challenges in motor-manual timber harvesting and designs a continuous digital process. Forest workers are provided with a smartphone app that supplies the necessary information and offers recommendations for safe felling and value-optimized bucking. Using the app, work progress can be documented and transmitted to downstream actors.
The technological basis for data exchange is the Smart Forestry concept, which follows the Industry 4.0 approach. A so‑called Asset Administration Shell (AAS) enables standard-compliant data exchange along the entire value chain. The AAS represents physical objects (forest stands, machines, stem sections etc.) as digital twins, encapsulating their data, functions, and parameters and thus enabling manufacturer-independent interoperability. Data is exchanged in a decentralized, trustworthy, and protected way via a messaging hub, the Smart Systems Service Infrastructure (S³I).
Digital work orders can be created using standard software connected to the AAS. In addition to the task description, aspects of work organization as well as emergency numbers and rescue points can be specified, which users can directly view or dial in an emergency.
For the safe felling of backward-leaning trees with wedges or winches, estimating the tree’s center of gravity is crucial. The proposed method records a point cloud of the stand via terrestrial laser scanning, either in advance of harvesting or continuously during routine work. Individual trees are extracted from the point cloud, followed by spatial recognition of stems and branches and automatic detection of possible damage (e.g., crown or stem breakage). The stem is then decomposed into segments to which typical density values are assigned, allowing the center of gravity to be derived. This information is provided in the app via the tree’s digital twin, enabling the calculation and display of the required forces, attachment heights, and insertion depths of auxiliary equipment. As an alternative to laser scanning, photo-optical methods for determining tree parameters are considered, which may be particularly advantageous in small private forests.
Based on the tree data, precise suggestions for assortment bucking can be generated from the taper curve and actual stem form, analogous to harvester-based optimization. After felling and processing, stem sections and work progress are recorded in the app; the data is transmitted back, updates the virtual stand, and forms the basis for subsequent extraction and logistics.
High-quality, diverse annotations remain a bottleneck for deploying computer vision in
forestry, where scenes are highly variable and targets are defined by operational context rather than appearance alone. A preliminary, fully autonomous dataset creation workflow is presented that links harvester-mounted RGB video to StanForD production records. The pipeline uses visual language models (VLMs) to ground operationally relevant targets in each clip, then applies segmentation with tracking to produce temporally consistent, pixel-accurate instance masks. Because VLM outputs can be non-deterministic, the pipeline includes a consensus-based reliability layer with rejection and fallback logic to reduce grounding errors. The system outputs structured training data aligned with operational metadata and preservs traceability to machine events, enabling reproducible regeneration of labels as models and prompts evolve. Early field-case results are reported for standing trees, a key target in cut-to-length operations.
In an initial clearcutting case study in a pine stand with limited undergrowth, one hour of video processing yielded 52 StanForD-linked cutting events, each represented by a 15-second video clip. The pipeline successfully localized, segmented, and tracked the target tree in 40 of 52 clips (77%). Failures were primarily caused by confusion with high stumps, crane components, stones, and other misdetections. The results indicate that context-aware grounding plus tracked segmentation can generate operationally aligned labels with minimal manual intervention in simple harvesting conditions, while exposing specific failure modes that guide future robustness work.
Resilient forest value chains depend on the adaptive capacity of forest managers at all scales. Small-scale forest owners, who manage a substantial share of Europe’s forest area, play a critical role in ensuring sustainable timber supply, safeguarding biodiversity, and maintaining multifunctional landscapes. However, their decision-making environment is becoming increasingly complex and uncertain. Climate change intensifies disturbance risks such as drought, storms, and pest outbreaks; timber markets are volatile and globally interconnected; and regulatory as well as societal expectations regarding ecosystem services, carbon sequestration, and habitat protection continue to grow. At the same time, many small-scale forest owners operate with limited access to specialized advisory services or advanced analytical tools. While innovation in harvesting technology, digital monitoring, and smart operations has advanced considerably, decision support solutions tailored specifically to the realities, constraints, and objectives of small-scale forestry remain underdeveloped.
A prototype AI-based decision support system designed to strengthen adaptive forest management at the holding level will be presented. The system builds on a systematic review of recent advances in generative and agentic artificial intelligence and translates these developments into a practical forestry context. Functioning as an interactive and context-aware assistant, the AI supports decision-making in complex and dynamic environments. By integrating stand-level data, owner-defined objectives, risk considerations, and regulatory constraints, the system generates context-sensitive management options aligned with multifunctional forest management and biodiversity goals. It structures trade-offs between economic performance, ecological resilience, and long-term sustainability, thereby enhancing transparency in strategic planning.
The study is showcasing how agentic AI can translate complex data, expert knowledge, and scenario assumptions into actionable guidance for decentralized forest managers. By empowering small-scale forest owners with adaptive, user-oriented digital tools, this approach contributes to strengthening resilience and sustainability across the whole forest value chain.
The demand for domestic energy wood and pulpwood in Finland has been high due to the increased use of forest chips and the cessation of wood imports from Russia. The forestry operations in the supply chain still face a persistent and long-standing challenge: young, often untreated first‑thinning stands are characterised by high stem densities and small‑sized harvested trees, which result in low hectare-based removals and quality products, generally pulpwood and energy biomass. This, in turn, lowers forest machine productivity and increases harvesting costs, highlighting the need for improved harvesting systems and solutions, and accurate productivity estimates of those systems.
The objective of this study was to examine the time consumption, productivity, and profitability of two harvesting system alternatives for those stands, and namely: separate or integrated harvesting of the two main products (e.g. pulpwood and energy wood). For the test, we used an excavator‑based system (Doosan DX140 LCR‑5) equipped with a TMK‑300 felling head for whole-tree harvesting and Ponsse Ergo 8w/H7, Ergo 6w/H6, and Fox/H5 harvesters adapted for multiple-tree handling through dedicated software.
Time study data were collected from five forest stands. In total, 4,519 stems (309.1 m³ [solid over bark, sob]) were harvested with harvester‑based systems and 3,210 stems (49.4 m³ sob) with excavator‑based systems. The harvesting systems tested were:
1) Whole‑tree harvesting,
2) Delimbed stem wood harvesting,
3) Pulpwood harvesting (using single-tree cutting & multiple-tree handling), and
4) Integrated energy wood (delimbed stem wood) and pulpwood harvesting with two-piles cutting.
When the size of removal was 5–13 cm (diameter at breast height), the effective hour productivity (E0) was 0.3–11.8 m3/h with separate whole-tree cutting and 0.4–19.3 m3/h with separate delimbed stem wood cutting. Correspondingly, the productivity of separate pulpwood cutting using single-tree and multiple-tree handling was 0.3–17.3 m3/h and 0.5–18.3 m3/h, respectively. With integrated delimbed stem wood and pulpwood cutting, using a two-pile method, the productivity was 0.4–17.8 m3/h. Cutting and extraction costs were calculated, and the profitability and sensitivity of different harvesting systems under various stand characteristics were examined and discussed.
Electrification is emerging as a key technological pathway for future forest harvesting systems to address existing challenges. These include reducing greenhouse gas emissions, improving energy efficiency, and strengthening the resilience of the forest sector to external disruptions, as it is currently heavily dependent on fossil fuels for harvesting and in-forest operations. Major manufacturers such as Komatsu Forest and Ponsse have recently introduced hybrid drivetrains in concept and prototype machines, paving the way also for a transition towards fully electric machines. However, battery-electric solutions in forestry face substantially different operational demands compared with on-road vehicles. Harvester and forwarder duty cycles are characterised by short, repetitive high-power peaks, complex crane operations and highly dynamic load profiles. At the same time, machines in northern regions routinely operate at temperatures down to –25 °C or colder. Such conditions strongly influence battery power output, usable capacity, charging behaviour and long-term degradation. While most battery research is conducted under standard laboratory conditions or according to automotive driving cycles, dedicated studies for forestry applications under cold-climate conditions are essentially non-existent. This represents a critical knowledge gap for harvesting system design, system selection and reliable winter operation planning and therefore a boundary for a future implementation of electric forestry machinery.
To address this gap, we are establishing a commercial-scale battery testing laboratory at the Swedish University of Agricultural Sciences (SLU), Department of Forest Bioeconomy and Technology in Umeå. The facility will enable controlled performance and aging tests of state-of-the-art lithium-ion and sodium-ion cells under forestry-specific duty cycles and temperatures ranging from –40 °C to +40 °C. Test protocols will be co-developed with machine manufacturers and forest companies to accurately replicate real harvesting and forwarding work cycles, ensuring relevance for mechanized cut-to-length systems operating in boreal climate conditions. Two main experimental blocks are planned: (i) characterisation and performance tests, including pulse-power and representative load-cycle experiments at varying temperatures, and (ii) long-term degradation studies at sub-zero and reference temperatures to quantify capacity fade, power loss and safety-relevant phenomena under realistic use. By generating high-quality, application-specific data, the laboratory will support improved battery integration into harvesting systems, inform cold-mitigation and charging strategies, provide inventory data to support upscaling in environmental and economic analyses (e.g., LCA/LCC) and enable more reliable uptime and lifecycle predictions. Beyond strengthening research infrastructure in northern Sweden, the initiative directly contributes to technological innovation in forest harvesting systems, positioning electrified drivetrains as a viable and climate-resilient solution for future forestry operations.
The idling of forest machines reduces the energy and cost efficiency of wood-harvesting operations. This study estimated the idling times of forest machines using automatically collected machine big data. The study also examined the impact of machinery type, harvesting enterprise, seasonal variation, engine power, and machine manufacturing year on fuel consumption during idling. Furthermore, a scenario approach was used to describe the influence of idling times on the volume and price of carbon dioxide (CO2) emissions. The dataset included monthly data from 2021–2024, covering ten Finnish logging enterprises with a total of 28 harvesters and 23 forwarders.
The results showed that the proportion of idle time of harvesters and forwarders regarding the total operating hours was 14.0% and 11.1%, respectively. Significant differences were observed between harvesting enterprises, averaging from 6.7% to 18.2%. Idling proportions were also found to be around 14%-unit higher in winter months (November–February) than in summer months (May–August). Average fuel consumption during idling was 4.28 L h–1 for harvesters and 3.38 L h–1 for forwarders. Greater engine power was associated with increased fuel consumption when idling, and vice versa newer machines exhibited greater fuel consumption during idling.
Annual idle-related CO2 emissions for wood-harvesting machinery in Finland were estimated at approximately 10,700 metric tonnes (t) with costs equivalent to €48 million. Reduction in idling by 25–50% could cut emissions by approximately 2,700–5,300 t CO2 and generate additional revenue up to €55 million when the reduced idle engine hours were transferred into productive hours. If idling is reduced by 100% and all the reduced idling time was assumed to be productive working hours, CO2 emissions from harvesters and forwarders would be 30,309 t CO2 greater, whereas the annual revenue would exceed €120 million.
Based on the study, it can be concluded that understanding, as well as acknowledging, the influence of idling has a positive impact on energy and cost-efficiencies, while also ensuring more environmentally friendly wood-harvesting operations. The research strengthens the foundation for low-emission and efficient machine operations, but further studies on idling are still needed outside the forest sector as well. Therefore, we recommend that future research focus on assessing the impacts of idling in other machine segments both in Finland and globally, utilizing automatically collected machine data.
In countries with a large forestry sector, road transportation is responsible for a considerable fraction of territorial carbon dioxide emissions. At the same time, heavy road freight in rural areas is arguably one of the most challenging cases for electrification. In the project TREE: Transition to efficient electrified forestry transport, 12 fully battery electric roundwood and wood chip trucks are studied during 18 months of operation. The feasibility and success of any means of transportation must be evaluated from its role in the logistics system. One component of evaluating the electric trucks of the TREE project is a 1:1 focus study, where each truck was compared to a corresponding diesel truck, with respect to energy use and productivity. We also observed the drivers’ different experiences from the two types of trucks. The trucks were set up to perform as similar transportation missions as possible in a two-day study. On the first day, the trucks worked tightly in pairs to closely replicate the physical conditions (time of day, driving pattern etc.) during loading and driving. On the second day, the trucks carried out the same transportation missions, but the drivers were free to work as best suited the respective truck’s conditions. The comparisons show that, depending on the mission, the electric truck can in some case replace a diesel truck without much adaptation whereas in other cases, it is difficult to match the productivity of the diesel truck. More importantly, the focus study will support the understanding of how electrification might influence forestry transport operations.
Demand and price for logging residues in Sweden have increased due to the security situation in Europe and ambitions for energy independence and low‑carbon energy systems. Accurate information on the volume and spatial distribution of logging residues from final fellings is therefore highly valuable for contractors, wood fuel suppliers and end‑users (heating and combined heat‑and‑power plants). More than 15 years ago, Skogforsk developed a system for forecasting and positioning logging residue volumes at stand level using harvester production files. Implemented through the calculation module “hprCM”, the system is used by several forest companies. For accurate forecasts, it is essential that harvester operators correctly distinguish fuel‑adapted areas from non‑adapted areas in the stand. In practice, the harvester and forwarder often need to reinforce strip roads and wet areas with logging residues to prevent soil damage. These adjustments currently rely on manual input via keyboard or touchscreen in the onboard computer.
Within the Swedish Energy Agency-funded project “Improved forecasts of logging residues for precision planning”, we conducted a large‑scale evaluation of the system by comparing forecast data with measurements from heating plants for several hundred wood chip deliveries. The evaluation revealed a large variation in how contractors used the system and in how well the forecasted volumes matched the measured volumes at the plant. This analysis formed the basis for developing improvements to the system.
One of our proposed solutions is to automate the registration process by using sensor technologies in modern harvesters. We carried out a field study in central Sweden, combining machine data with high‑accuracy ground‑truth measurements. Through this study, we identified key variables for automating the registration process. In the mid‑term, if implemented, automated registration of logging residue volumes can reduce the cognitive load on operators, improve the reliability of forest fuel stock levels, and strengthen supply security.
Seepage water can occur when wood chips are stored outdoors. In relation to precipitation, the amount of seepage water is small. It contains high levels of dissolved organic matter. If such water enters surface waters, it could pose a threat to aquatic life. If the water seeps into humus-rich forest soils, microorganisms decompose the dissolved organic matter of the water quickly and thus filter them out. Unsealed pavements at storage yards don’t have a humus-rich and biologically active soil layer. Until now, the filtering effect of a storage area ground was unclear. To investigate this, wood chips were piled up on a gravel storage area and seepage water was collected and analyzed using lysimeters both above and below ground to a depth of 30 cm.
The experiment showed that the COD values (chemical oxygen demand) of the seepage water were significantly lower than in previous experiments. The values in the subsoil did not differ significantly from those on the surface. The BOD and TOC values (biological oxygen demand and total organic carbon) were also low. The reasons for the low COD values are not entirely clear. Possible influencing factors include the smaller size of the wood chip pile, dilution by precipitation water, biological degradation during the retention time in the collection container, and binding to calcium in the filter material of the lysimeters, although only a minor effect could be detected for the latter. No significant filtering effect of the subsoil was observed. There is no danger to soil organisms above the groundwater line from the seepage water because there is sufficient oxygen available. As a precautionary measure, it is recommended not to store wood chips outdoors at sites with very high groundwater levels.
This study estimated the availability of unused materials for woody biomass power generation plants registered in FIT (Feed-in Tariff) as the supply potential through the introduction of trailers with 20-t payload instead of trucks with 10-t payload which were widely used on winding road networks in mountainous areas of Japan. As a result, supply potentials of used and unused materials were estimated at 65,413,601 m3/year and 13,082,720 m3/year, whereas those availabilities were estimated at 64,322,304 m3/year and 12,864,461 m3/year, respectively. Therefore, the rate of the availabilities to the supply potentials was 98.3%. The availabilities of used and unused materials without the introduction were 54,268,941 m3/year and 10,853,788 m3/year. Thus, its introduction contributed to increase the availability. Since the demand of woody biomass power generation plants registered in FIT was estimated at 11,018,104 m3/year, the availability through the introduction met the future demands in Japan as a whole.
Bioenergy accounts for approximately 30% of Sweden’s energy supply, of which about 80% is derived from the forest sector, e.g. in the form of logging residues such as treetops and branches. In recent years, both demand and prices for logging residues have increased in Sweden, driven by societal decarbonisation efforts and ambitions to strengthen national energy independence in response to Europe’s security situation. The moisture content in logging residues is a key payment‑determining quality attribute, as lower moisture increases energy density in the material and thus its market value. Each year, logging residues worth millions of euros are forwarded from harvesting sites and stored along forest roadsides without any possibility to monitor their moisture content. For fuel suppliers, reliable information on wood fuel stock levels and expected energy content is essential for delivery planning and for fulfilling contractual obligations. Consequently, there is a need for improved understanding of the factors influencing moisture content in wood fuel and for robust methods to estimate it.
Moisture dynamics depend on storage conditions which are strongly influenced by weather‑related factors such as precipitation, temperature, humidity and wind. While large‑scale weather datasets can explain part of this variation, local microclimatic conditions and practises at the storage site are not captured in the available datasets.
Within the Swedish Energy Agency–funded project “Improved forecasts of logging residues for precision planning”, we analysed datasets of wood chip moisture content measured at power plants for thousands of deliveries. The initial aim was to develop a data-driven prediction model to estimate moisture contents and to analyse how storage conditions and supply-chain activities, such as forwarding, could explain the observed variability and provide better decision-support. However, our results showed that forecasting moisture content based solely on available weather data is challenging due to poor representation of local site conditions. It is more effective to promote favourable drying conditions through improved planning and work practices along the supply chain. Enhanced decision-support for moisture management can increase the value of this residual assortment, improve profitability throughout the supply chain and increase overall energy efficiency.
This study assesses the quality and energy potential of softwood plantation harvest residues from south-east Queensland, with the aim of supporting emerging biomass markets. The focus is on the effect of different comminution methods—chipping and grinding—and, for chipped material, the influence of screening.
Harvest residues, mainly clearfall tops and other off-cuts, were collected from two common harvesting systems: whole-tree processing at the roadside and cut-to-length processing at the stump. A total of 26 samples from 16 harvesting sites were analysed, with almost all samples originating from Southern Pine plantations. Key properties examined included moisture content, ash content, and net energy value.
Although the limited sample size prevented statistically robust comparisons, several clear trends were observed. Unscreened Bruks mobile chips produced at roadside processing sites showed higher moisture content but lower ash content compared with unscreened grindings from cut-to-length operations, while net energy values were similar. Screening of Bruks mobile chips reduced moisture content but had little effect on ash content or energy value. In contrast, screened fines contained substantially higher ash content than screened chips, despite similar moisture and energy values.
These findings highlight how harvesting system, comminution method, and screening can influence biomass quality and should be considered when supplying plantation residues to energy and bio-based markets.
Precise forest management under the increasing pressures of climate change necessitates a transition from static inventory databases to dynamic, high-resolution digital twins. Current modelling approaches frequently rely on episodic remote sensing data, such as LiDAR or aerial photogrammetry, which suffer from temporal latency and fail to reflect the immediate structural changes following harvesting or stochastic disturbance events. This presentation introduces an operational framework for the management of dynamic, single-tree-based digital twins utilising the "DigiTreeS" data standard. This architecture incorporates a robust API designed to facilitate interoperability between various forestry management systems, including silvicultural decision-support tools, wood supply chain logistics, and predictive growth modelling.
The core innovation of the DigiTreeS system is an automated interface that leverages near real-time production data from harvesters during operational processing. By integrating geolocated harvester production data, the digital twin is updated instantaneously: harvested individuals are removed from the model, while the attributes of the residual stand, including diameter at breast height (DBH), stem quality, and curvature, are validated and refined through in-situ sensor measurements. Preliminary results indicate that this hybrid approach significantly enhances the accuracy of timber volume and assortment structure predictions compared to traditional models derived solely from remote sensing.
By shifting the digital twin from a static representation to a learning ecosystem, this research provides the foundational data infrastructure for adaptive forest management, transparent carbon certification, and precision logistics. Furthermore, the single-tree granularity of the model enables advanced visualization of thinning strategies and operational planning through virtual and augmented reality (VR/AR) interfaces. This presentation delineates the technical concept and evaluates the results achieved in experimental test areas, demonstrating the utility of dynamic single-tree digital twins as a transformative tool for modern forest operations.
One of the most important parameters for the successful establishment of forest stands and forest habitat restoration programs is forest seedling quality. Some of the traditional methods for seedling quality assessment are invasive, time consuming, and labor-intensive. Hyperspectral imaging, in contrast, provides a noninvasive evaluation method, making it a new and a promising tool for seedling quality assessment. This is crucial from aspect of practical nursery production and outplanting success, but also as an background for development of new scientific methods. By capturing hundreds of continuous spectral bands, HSI enables the detection of detailed plant traits. This review focuses on the application of HSI in assessing forest seedling quality for application in nurseries and after outplanting in the frame of restoration initiatives highlighting its strong potential for evaluating key parameters. However, HSI alone cannot capture dimensional/physical traits, such as shoot height, root mass, bud dormancy, and therefore is best used in combination with complementary technologies, including LiDAR, vegetation indices, and image-based methods. This review also examines current challenges in implementation of HSI in operational nursery settings, such as illumination position sensitivity, plant-specific traits, large data volume, and high equipment cost. By integrating hyperspectral plant phenotyping and nursery practices, this review identifies future directions for HSI implementation in forest seedling production, thereby supporting more efficient nursery settings.
Climate-driven forest mortality is having an increasingly significant impact on harvesting intensity, the stability of wood flows, and the medium-term availability of timber. However, translating climate signals into operationally meaningful indicators remains challenging, as the strength of the relationship between drought and mortality depends on the spatial scale. This study evaluates the extent to which climatic water balance (CWB) can explain Scots pine (Pinus sylvestris L.) mortality at different levels of forest management in Poland: forest district (operational scale) and regional directorate (strategic scale).
Mixed-effects modelling was employed to distinguish ecological drivers from scale-dependent variability, employing long-term mortality records and high-resolution climate data. The model's predictions were expressed on a logarithmic scale of relative mortality. Drought-related mortality was found to range from approximately 6 × 10³ to 2 × 10⁵ m³ at the regional directorate level (mean standing volume: 61,755,798 m³), and from 2 × 10¹ to 7 × 10² m³ at the forest district level (mean standing volume: 224,213 m³).
The study shows that spatial aggregation strengthens the relationship between drought and mortality. Short-term harvest scheduling and resource allocation are complicated by the strong influence of local site conditions and stochastic disturbances on mortality variability at the forest district level. Local variability diminishes at broader spatial scales, allowing a coherent multi-year drought signal to emerge and impacting mid-term timber supply and processing capacity planning.
The results indicate that spatial scale is an essential component of climate risk assessment for forest operations. Identifying the appropriate level of aggregation enhances mortality forecasting, supports adaptive harvest planning, and makes forest-based value chains more resilient under increasing climate stress.
In thinning operations in rotation forestry, reasonable thinning intensity is key to ensuring favorable development of the stand in terms of sawlog yield, growth loss, and carbon sequestration. In recent years, realized thinning intensities in the Nordic countries have often violated recommended limits, and operators have voiced their desire for decision support and reporting assistance in thinnings. As a response, we present here a new concept, named R-TIM, for real-time monitoring of thinning intensity. R-TIM is based on collecting a remote sensing-based single-tree map of the stand prior to harvesting, constituting a digital twin of the forest. The twin is then kept up to date during the harvesting via matching harvested trees with their counterparts in the tree map using regular harvester measurements and crane-tip localization. Thinning intensity metrics are computed, in real time, for removed tree stock from the harvester data, and for standing stock from the digital twin. We present a case study and evaluation of R-TIM in Finland using real forest data and simulations built on top of these. In addition, we discuss future directions of development and broader implications of the R-TIM concept.
Accurate estimation of a tree’s expected leaning direction is essential for safe and controlled felling, particularly in complex crown architectures or winch-assisted motor-manual felling operations. In practice, such decisions rely primarily on visual assessment by experienced forestry workers. The extent to which this estimation reflects the tree’s actual mechanical mass distribution and the direction of its centre of gravity (CoG) remains insufficiently quantified. This study compares personal visual tree assessments with terrestrial laser scanning (TLS)-derived CoG directions and extends the analysis to operational implications in winch-assisted motor-manual felling.
54 trees were independently assessed by three experienced workers (162 assessments in total). Each worker estimated the direction in which the tree was most likely to fall. For each tree, a three-dimensional CoG was calculated using TLS-based quantitative structure models (QSM). Monte Carlo simulations were applied to consider structural and density-related uncertainties. Across all assessments, the mean absolute deviation between personal visual assessments and TLS-received CoG directions was 48° (median 35°), with extreme deviations reaching 141° (95th percentile). Differences between workers were small relative to the overall variability, indicating that uncertainty was not just limited to individual performance. Only 13% of assessments fell within ±10°, and less than half were within ±30°. This demonstrates limited precision at operationally relevant tolerances. Error patterns revealed systematic directional tendencies rather than random scatter. The forest workers frequently showed similar assessments of individual trees. Nevertheless, these assessments did not consistently align with the TLS-derived CoG directions, suggesting that visual assessment does not automatically imply mechanical accuracy. The observed deviations suggest that visually prominent features, such as stem curvature or inclination, strongly influence judgments, while less visible crown mass asymmetries that shift the CoG may be underestimated. Building on the three-dimensional CoG modelling, the study is being extended to quantify the mechanically required winch cable attachment height for each tree. This allows the geometric consequences of mass displacement to be assessed beyond directional estimation. These attachment heights will be compared with practitioner-based height selection in winch-assisted motor-manual felling, providing an operational perspective on CoG displacement.
Overall, the research indicates that personal visual tree assessment alone is insufficient to reliably capture the mechanical mass distribution of complex tree structures. The integration of laser-based CoG modelling with practitioner-based assessment establishes a framework for quantifying the boundaries of visual judgment, thereby promoting safety-oriented decision-making in forestry operations.
Efficient dispatching of timber trucks is a major determinant of cost, productivity, and supply reliability in forest transport systems. In operational contexts, where transport distances are long, pickup locations vary daily, and mills operate under strict time windows, dispatching decisions must account for dynamic constraints including queue conditions, service times, driver duty limits, and interactions among trucks competing for shared resources. Traditional truck scheduling models can optimize fleet plans for a full day, but their computational requirements limit their use for real-time decision support when conditions change during operations.
This study presents a hybrid framework that integrates optimization and machine learning (ML) to enable rapid, adaptive dispatching decisions. A fleet-level truck scheduling model is first used as a “teacher” to generate optimized daily schedules across multiple simulated scenarios representing realistic Australian forest transport conditions. These solutions are decomposed into sequential decision points, capturing system state variables such as truck locations, remaining duty time, active transport tasks, origin–destination distances and costs, queue estimates at pickup points and mills, and mill time-window constraints. These data are used to train an ML model that learns to predict the next best transport task for each truck.
Because mill destinations remain relatively stable while pickup locations change daily, the model is structured to generalize across new harvesting configurations. Results indicate that the learned dispatching policy can approximate optimization-quality decisions while operating fast enough for real-time deployment. The proposed approach demonstrates how combining optimization-based planning with data-driven learning can support intelligent, responsive dispatching systems for modern forest transport operations.
Truck weight regulations play a central role in shaping the efficiency, safety, and infrastructure impacts of timber transportation in the southern United States. Under federal law, gross vehicle weight (GVW) is limited to 80,000 lbs. (36,287kg) on interstate highways, while in Alabama, state roads permit up to 88,000 lbs. (39,916kg) through a 10% allowance. This study surveyed logging business owners across Alabama to assess their perceptions of these regulatory limits and the implications on operational performance, safety, and road use. Respondents (n=146) were asked about current weight tolerances, potential policy adjustments, safety outcomes, and infrastructure effects. Results revealed broad support for increasing allowable weights on interstate highways, with over 85% of respondents indicating that such changes would improve efficiency, primarily through enhanced route flexibility and reduced fuel and maintenance costs. A majority (87%) also perceived interstate highways as safer than state routes, citing fewer traffic stops, better engineering, and smoother traffic flow. While 97% of the respondents believed their existing trailers could safely handle increased weights, only 39% were willing to invest in equipment upgrades without regulatory certainty or financial incentive. Although concerns about infrastructure damage were expressed by a minority, most participants (78%) viewed interstate highways as capable of supporting heavier loads. These findings suggest that modest GVW increases, particularly when aligned across state and federal roadways, may enhance productivity and safety in the logging sector without imposing disproportionate infrastructure risks. However, successful implementation will require careful coordination across permitting systems, enforcement practices, and local infrastructure planning.
Uncertainty in harvesting output and road accessibility significantly affects supply chain planning and the ability to meet customers’ demand for high-quality wood products in a timely manner. Despite a long-term harvest scheduling, variations in product volumes can reduce delivery precision and increase the risk of product backorders. Weather conditions also influence landing accessibility; for example, schedules based on expected road conditions may suddenly change due to warmer weather, thawing, or heavy rainfall.
In response to these challenges, there is interest in modeling variations, such as predicting road accessibility, which is further stressed by climate change. Predicting product variations and road accessibility can support forest managers in operational planning; however, decision-making on transport prioritization must remain aligned with planning objectives to ensure that established delivery performance targets are met. Evaluating road accessibility is an important tool, but it should be complemented by decisions on transport prioritization to meet delivery goals, such as minimizing backorders and delivery variations.
To address this problem, we present a methodology that uses an AI agent for decision-making on transport prioritization, fed with data from a digital twin to monitor landing inventory levels and road accessibility. Despite operational variability, the agent’s goal is to prioritize transports to maintain adherence to multiple long-term delivery targets.
The methodology is implemented in a decision-support prototype that enables managers to track long-term planning goals on a dashboard and to evaluate how operational decisions based on predicted road accessibility and real-time landing inventories align with delivery targets. The dashboard visually helps identify landings at risk of reduced road accessibility and provides guidance on prioritizing transport. It also supports the comparison of alternative decision scenarios, helping managers determine when a new long-term harvest schedule may be needed.
The prototype is expected to be further developed and integrated into operational environments in future projects in collaboration with forestry stakeholders.
Log truck weights in the US state of Georgia are governed by state law, including limits for gross vehicle weight, single axle, and tandem axle weights. A very small percentage of loads are inspected by law enforcement, however. Some forest products mills have internal policies regarding the weight of delivered loads to reduce liability and protect mill infrastructure. The objectives of this study were to 1) evaluate the prevalence of mill weight policies, 2) document the structure of mill weight policies, and 3) analyze the variability in the weight of loads delivered to Georgia mills. We conducted a survey of Georgia mill-owning companies and obtained a sample of nearly 130,000 mill scale tickets from cooperating mills. Survey responses were received from owners of 27 Georgia mills representing 14.7 million tonnes of annual wood consumption (~33% of annual wood consumption in the state). Eighty percent of companies had instituted an internal weight policy to control gross vehicle weight (GVW) of delivered loads. The stated reasons for instituting weight policies included liability reduction, improvements to highway safety, and protection of mill infrastructure. The weight policies were effective at preventing the delivery of grossly overweight loads (>5,400 kg over legal limit). Only 45% of delivered loads were within ± 1.5 tonne of the legal GVW limit and 27% of delivered loads were at least 1.5 tonnes below the legal limit. Improved management of truck payload by logging businesses could reduce hauling costs and improve timber transportation profitability in the state.
More competitive wood supply is dependent upon increasing load sizes for timber trucks. With a background of increasing percipitation and temperatures the latest revision of Norwegian forest road standards have been adapted with improved construction and drainage specifications. However, the existing forest road network has been constructed over a time span of 60 years, throughout the evolution of truck combinations to the current 60t/24m standard. At present both 65, 68 and 74t truck-trailer combinations are being run on a test basis in Norway.
With this background, road upgrading investments require documentation of prioirities for effective transport operations. The goal of the study was therefore to map out the key challenges of todays forest road network from a transport availability perspective. The purpose is to map out the most critical factors for improving road availability and competitive roundwood transport.
The study provides a national analysis covering both lowland and mountainous topographies on the eastern- and western sides of the Scandinavian mountain range. The first part presents the results of focus group interviews mapping the key dimensions of current challenges (30 trucking companies). The second part presents a quantitiative analysis of transport messaging during 2024 and 2025 reporting i) noncompliance with road standards and i) consequent delivery delays (all Norwegian wood supply organizations).
The quantitative analysis is enabled by new functionality in the forest sector transport system (SkogData VSYS-transport). The existing transport messaging (PapiNet std) includes load-level data such as wood order/transport order identification, time stamp, origin-destination coordinates, loaded transport distance, public road standard (GVW class), wood assortment code and load volume. Since 2024 noncompliancies have been registered within the same system. The noncompliance category for forest roads includes 6 sub-categories: bearing capacity, road width, road grade, landing grade, roadside vegetation and insufficient snow removal/sanding (with explanatory free text). Noncompliances are linked to a secondary function for resulting transport order delays («TO parking»). TO delays have 6 sub-categories of underlying factors: public roads, private roads, forest roads, delay until improved bearing capacity, delay until road/turnaround is improved, and other factors (eg. receiver status).
In this presentation the STT project (STT; Forest road conditions and availability) reports seasonal distributions and durations for 33 000 non-compliances and 15 500 delivery delays during 2024 and 2025. AI-assisted interpretation of free text descriptions provided a more detailed distributions of underlying factors. These are seen in the context of risk evaluations by 30 transport companies, where seasonal transport conditions vary considerably between coastal and interior regions.
The practice of poplar forest establishment, maintenance, and harvesting varies globally, potentially inducing significant effects on cost structures and environmental impacts. In Romania, for instance, a typical life cycle involves operations such as ground clearing, stump removal, soil milling, pit drilling, and manual planting, followed by several cultivation stages, thinning, and clear-felling. While mechanized pit drilling is not a new option, modern machinery has seen significant technical and operational improvements. In this study, we utilized an extensive GNSS-annotated dataset collected from two tractor classes during pit-drilling operations. Specific events - including entering a row, drilling a pit, moving between locations, re-entering maneuvers, and relocation - were documented at one-second intervals. These data were used to extract event durations and frequencies to compare the performance of the two tractors. Given the non-normal nature of the data, bootstrapping was employed to infer the mean values and compare differences at the event level. Our findings suggested significant performance differences between the tractors, which were subsequently treated as two distinct options for cost assessment. The cost assessment integrated detailed statistics on fixed, variable, and labor costs. Using performance metrics such as event duration and speed, we conducted a cost simulation driven by plot size (1 to 20 hectares), variable plot dimensions, and four common planting schemes (5 × 4, 4 × 4, 4 × 2, and 3 × 2 m), resulting in nearly 84,000 comparison datasets. We evaluated two unit cost metrics: cost per pit and cost per hectare. Performance strongly influenced outcomes; the larger tractor class yielded higher unit costs, likely due to increased maneuvering and inter-pit travel time. Within the same tractor class, unit costs increased non-linearly with plot width, reflecting the impact of frequent entry and exit maneuvers relative to shorter on-row operations. When analyzed against plot length, costs were largely clustered by plot size, with cost differences becoming most pronounced for plots up to 5 hectares. These results provide updated performance and cost figures for poplar establishment in Romania. Furthermore, they demonstrate that the current practice of clear-felling in plots up to 5 hectares represents an economically critical threshold, beyond which cost reductions become marginal. Internationally, our findings illustrate how plot geometry and planting schemes drive costs, supporting the selection of optimal tractor classes for specific site conditions.
One of the key objectives of Estonia’s current National Energy and Climate Plan 2030 is to reduce national greenhouse gas (GHG) emissions by 80% compared with the 1990 level. In line with the European Green Deal, the Estonian government has also committed to achieving climate neutrality by 2050. Meeting these ambitious targets requires a substantial increase in the use of renewable energy sources, including forest-based biomass fuels.
This study analyses the procurement of energy wood using crawler excavators equipped with guillotine-type cutting heads. The machines investigated were the two Sany SY135 and Volvo ECR145DL. Data were collected over 4,300 productive machine hours and a total harvested volume exceeding 60,000 m³ (solid volume) during 2025. Fuel consumption was evaluated in litres per productive machine hour (l h⁻¹) and ranged from 5.8 l h⁻¹ for the Sany SY135 to 14.0 l h⁻¹ for the Volvo ECR145DL. The observed differences in fuel consumption were largely attributable to variations in machine mass, with operating weights of 14 t and 16 t, respectively.
Average productivity reached 10.2 m³ h⁻¹ for the Volvo ECR145DL and 14.0 m³ h⁻¹ for the Sany SY135, with operator experience also influencing performance. The operations were mainly conducted in the clearance of overgrown land, involving tree species such as grey alder (Alnus incana), silver birch (Betula pendula), aspen (Populus tremula), black alder (Alnus glutinosa), Norway spruce (Picea abies), and Scots pine (Pinus sylvestris).
The energy utilisation of shrub biomass requires specialised machinery adapted to shrubland conditions, enabling efficient harvesting and collection in cost-effective formats suitable for energy production. Proper handling is essential to minimise soil contamination and ensure the production of high-quality biofuels. Current harvesting technologies integrate clearing and biomass collection into a single operation, producing either baled or mulched material, each associated with distinct logistical requirements within the biomass supply chain.
This study presents a comparative analysis of mechanised shrub biomass harvesting using two different machines associated with different logistics systems: a López Garrido TBA-2300 harvester-mulcher and a Biobaler WB55 harvester-baler. The techno-economic evaluation was conducted in a natural mountain shrubland dominated by rockrose (Cistus laurifolius L.) in Soria (Spain). This species commonly occurs in association with oak, pine and holm oak forests on acidic or decalcified calcareous soils. In the Iberian Peninsula, rockrose occupies 51,377 ha as a dominant species and 680,856 ha as a secondary species (San Miguel et al., 2004; Montero et al., 2020), extending across wide areas of the central-northern, eastern and southern regions.
The López Garrido TBA 2300 harvester-mulcher, powered by a 200 hp forest-adapted tractor equipped with a reverse driving position, is driven by the power take-off and hydraulic system and has a working width of 2.3 m. Biomass is shredded by mobile hammers and conveyed by a screw and turbine into a 5 m³ front container for temporary storage. In contrast, the Biobaler WB55 performs shrub cutting, collection and baling in a single operation with a similar working width (2.3 m). The harvested biomass is compacted into cylindrical bales (Ø 1.2 m), automatically tied and discharged at the rear, facilitating handling, transport and storage.
The mechanised harvesting trial was designed to assess weight productivity (tWM/PMH, tonnes of wet matter per productive machine hour), area productivity (ha/PMH), operational costs (€/tWM) and harvesting efficiency (CE, %), defined as the ratio between collected biomass and the biomass available prior to harvesting. Operational times were recorded using the Valtra Connect telemetry system installed in the tractor, allowing remote monitoring of productive working hours, travelling speed, fuel and additive consumption, and spatial tracking of the operated areas.
Under similar stand conditions, the harvester-mulcher achieved significantly better performance than the harvester-baler. Biomass recovery reached 3.9 tWM/ha compared with 1.3 tWM/ha, while productivity was 1.4 versus 0.5 tWM/PMH. Harvesting efficiency was also higher (82% compared with 52%). These differences translated into substantially lower collection costs, with biomass harvested using the harvester-mulcher costing 48 €/tWM compared with 192 €/tWM for the baler, and productive hourly costs being 22% lower (82 €/PMH versus 105 €/PMH).
For biofuel production, the material was ground using a 30 mm screen and sieved through a 2 mm mesh to reduce fines. The resulting hog fuel was classified as class I1 according to ISO 17225-9. Rockrose represents a low-cost solid fuel characterised by low emissions of NOx, SO₂, HCl and minimal slagging, but its high ash content makes it more suitable for industrial than residential applications.
Short-rotation coppice is a modern industrial crop, and its profitability depends on the mechanisation of all management operations. Once the crop is mature, trees can be felled with a variety of machines, but shears represent the most effective solution. However, shears can cause some stump damage, raising concerns about stump mortality and re-sprouting vigor, both crucial to coppice regeneration. This study, conducted at five sites in western Slovakia, compared the mortality and resprouting vigor of stumps cut with shears and chainsaws, the latter representing the control treatment. Measurements were taken on 6 compartments across these sites. In each compartment, between 9 and 10 sample plots were established, each consisting of two parallel row segments containing 10 live stumps (20 stumps per plot). One row segment was cut with a shear-type felling head, the other with a chainsaw-type felling head. One year after cutting, we assessed stump diameter, number of live shoots, and diameter and height of the 5 tallest shoots. Data were pooled into a structured database and statistically analysed. The results of the study showed that site fertility had a very strong effect on resprouting vigor; the method of the cutting technique did not affect survival stumps, but shear-type cut stumps produced slightly more shoots, and dominant shoots on shear-cut stumps might grow more slowly due to stronger intra-stump competition, especially in less fertile sites.
Utilising biomass (tree branches and treetops) for energy purposes is a common practice in mechanised forest operations. Collecting biomass, however, may be time consuming if, for easy and effective forwarding, the harvester operator is focussed only on timber piling, but not paying attention to biomass piling. To make biomass forwarding more effective, placing it in piles is an alternative that can boost productivity in the last stage of forest operations: biomass extraction. The objective of this research was to compare the productivity and costs of cut-to-length (CTL) technology in two types of forest operations: with biomass piling (WBP) and with no biomass piling (NBP), while the timber processing in both cases was the same.
The research was carried out in North-West Poland in a pure, 96-y.o. pine (Pinus sylvestris L.) stand, in which clear cutting was applied on an area of 3.33 ha (half WBP and half NBP). A John Deere 1170E harvester and John Deere 1110E forwarder were used, with the forwarder also extracting the remaining biomass. In the WBP and NBP operations, the timber was piled next to the strip roads as usual. However, in WBP, the biomass was placed in piles next to the strip roads (there were empty spaces between each pile), while in NBP, the biomass was equally and continuously distributed along the strip roads.
The productivity (m3 PMH0-1) of timber harvesting and forwarding in WBP was 14% lower in comparison with NBP, while the biomass extraction was 34% higher in WBP than in NBP. Specifically, the productivity (m3 PMH0-1) of the harvester in WBP vs NBP was 18.75 vs 21.32; the productivity of the forwarder was 18.58 vs 21.17; and biomass extraction was 7.37 vs 5.52 (34% more in WBP).
The total costs of timber harvesting, as well as timber and biomass forwarding (weighted arithmetic mean) was 8.84 € m-3 in WBP and 7.88 € m-3 in NBP (thus WBP was 12% more expensive). However, the cost of biomass forwarding only was lower by 33% in WBP (9.19 € m-3) in comparison with NBP (12.18 € m-3).
Forest harvesting operations increasingly require real-time, non-destructive tree-level information to support precision decision-making directly within machine workflows. Although harvester-mounted Mobile Laser Scanning (MLS) systems provide detailed three-dimensional representations of tree stems, most existing processing approaches depend on full point cloud reconstruction, frame alignment, and Simultaneous Localization and Mapping (SLAM). These pipelines are computationally demanding and typically require post-processing, limiting their applicability for real-time decision support in operational environments. Consequently, sensing, measurement, and bucking optimisation are often addressed as sequential and loosely connected tasks, constraining the development of fully integrated, sensor-driven harvesting systems.
This study presents an integrated real-time framework that operates directly on raw MLS data through two-dimensional projections of individual MLS frames, thereby avoiding explicit 3D reconstruction. Each frame is independently transformed into a structured 2D representation that enables stem instance segmentation using a 2D Convolutional Neural Network (CNN). From the segmented projections, geometric attributes such as stem diameters and longitudinal profile are estimated sequentially, allowing the progressive construction of a dynamic stem representation during machine operation. These measurements are directly coupled with an optimisation module implementing an optimal bucking strategy based on predefined product specifications and operational constraints. By integrating perception, measurement, and optimisation within a unified continuous processing pipeline, the proposed framework establishes a reconstruction-free approach to real-time, sensor-driven bucking decisions and contributes to advancing automation in forest harvesting systems.
The economic sustainability of energy-oriented forest supply chains depends on the ability to process low-value biomass using robust and cost-efficient technologies. In such contexts, reducing maintenance requirements, minimizing downtime, and maximizing operational productivity are essential to maintain profitability. Conventional grapple-saw systems, although widely adopted, are primarily designed for higher-value assortments and may not represent the most efficient solution when bucking material destined for energy production. This study evaluates a radio-controlled shear prototype powered by an autonomous 89 kW engine, designed to provide a simple and robust bucking system for low-value biomass assortments. Field trials were conducted in aged chestnut coppice stands, where whole logs were bucketed at a landing into two target lengths (2.4 m and 6 m), comparing the shear prototype with a conventional grapple-saw. When bucking to 6 m lengths, the shear prototype achieved a productivity of $10.21~\mathrm{m^3\,h^{-1}}$, approximately 49% higher than the grapple-saw ($6.85 ~\mathrm{m^3\,h^{-1}}$). Bucking to shorter 2.4 m lengths reduced productivity by 29% for the shear prototype and by 53% for the grapple-saw, highlighting a greater sensitivity of the conventional system to shorter assortments. Fuel consumption per unit volume was reduced by approximately 58% at 6 m and 76% at 2.4 m with the shear prototype compared to the grapple-saw. The results demonstrate that shear-based bucking provides a technically reliable and advantageous alternative for low-value biomass, combining higher productivity with substantially lower fuel use, thereby reducing operational costs and improving the overall efficiency of energy-oriented harvesting operations.
Skogforsk has carried out reoccurring tests of harvesters with a focus on value recovery (including e.g. damages, measuring accuracy, optimization efficiency) and StanForD implementations since the mid 1990’s. A new test with seven different final felling harvesters was carried out in the autumn of 2025 in one single location north of Uppsala. The test was carried out in a Norway spruce stand with an average stem size around 0.9 m3sub where each machine harvested at least 220 stems and 34 of these stems were manually control measured. The test also included, for the first time, an evaluation of the precision of stem coordinates from harvesters equipped with high precision GNSS systems.
All active manufacturing members in the StanForD working group were invited to participate. The following harvesters participated:
Base machine Harvester head Controller
Komatsu 951 C144 MY23 MaxiXT
Rottne H21 EGS706 Dasa Forester H70
Ponsse Scorpio Giant New Ponsse H7 Opti 5G
EcoLog 590G Ec661 Lf Dasa Forester H70
Logset 12H Logset TH80 TOC-MD
JohnDeere 1470H H425 Timbermatic H
Tigercat 1165 Log Max 5000V Log Mate 510
It is possible to analyze the historical development of several key figures as this type of test has now been carried out five times with the same basic methodology. The presentation will thus include comparisons between the harvesters in the present test as well as general comparisons over time.
The presentation will include results on length, diameter and volume accuracy for the different harvesters where manual control measurements are used in order to compare with the values generated by the harvester. Results on the performance of the bucking optimization will also be presented including the effect of the measuring errors. All harvesters today have the possibility to use two different optimization methods and results for both methods are thus included in the study. The overall trend is that the measuring accuracy has improved since the previous test (e.g. standard deviation for diameter measurements has decreased from 2,3 mm to 1,4 mm since the test 2006) and that the bucking optimization algorithms are performing as expected (≈99% of theoretical maximum).
The test also included a thorough analysis of the implementation of the StanForD standard including what data is available from each control system and what StanForD version that is supported. Three of the harvesters had implemented the latest StanForD version (4.1) which enhance the reported production data, especially when it comes to positioning information such as the log positions. Another important aspect was to test and verify that the flexibility in controlling the bucking optimization works as expected, e.g. making it possible to change the mix of assortments or species harvested at a specific site.
Three harvesters reported the true position of the stem and most machines could report tracking data describing machine movements. All systems support registering “stem codes” that makes it possible to identity e.g. retention trees and wind-throws.
Other aspects such as precision of stem positioning, processing damages and sawing speed will be covered in other presentations at the conference.
Efficient bucking systems and careful timber handling are key factors in maximizing timber value. To evaluate the performance and long-term advancement of such systems, Skogforsk has conducted a series of large-scale wood value trials. The most recent study, conducted in 2025, compared seven leading harvester systems under operational conditions. Data were collected in a spruce-dominated final felling stand in central Sweden, which resulted in several complementary datasets used to study dimension measurement accuracy, bucking performance, positioning, timber handling, and system functionality according to StanForD 2010.
The present analysis examines how cutting dynamics influence the occurrence of bucking splits, which are caused by the gravitational forces acting on the freely hanging part of the log during the cut. Cutting time was determined from high-speed video recordings of cuts across a representative range of log diameters, and a comparison with cutting speed data from previous wood value trials will be presented. The video material was further used to describe the progression of the cut with regard to saw bar angle. Measurements of bucking splits were performed on the bucked logs from 210 stems. Discs were manually cut from the log ends and struck until breakage, allowing classification of the break as either natural or the result of a bucking split based on characteristic patterns.
A logistic regression model was developed to relate cutting dynamics to the likelihood of bucking splits. The model incorporates cutting time, saw bar angle, and log geometry to mechanistically capture the interplay between stress development and cut progression, which determines whether the cut is completed before split-inducing stress levels are reached. Preliminary analyses indicate that log geometry is the dominant factor explaining the occurrence of bucking splits, while aspects of the cutting process also show statistically significant effects.
This study analyzes and compares the productivity and economic efficiency of pull-through delimbers and processors in the processing of whole trees from partially mechanized timber harvesting in Baden-Württemberg, Germany.In forest-rich and topographically challenging regions such as the Black Forest, fully mechanized harvesting systems often reach operational limits. As a result, adapted semi-mechanized approaches are frequently applied, typically relying on the pre-concentrationof timber. Within such systems, the pull-through delimber represents a practical alternative to processor-based processing. However, its performance and operational potential have so far received little systematic investigation.
The study combines qualitative and quantitative methods to evaluate operational workflows and machine performance. Expert interviews were conducted following the Grounded Theory approach to gain insights into practical applications, operational constraints, and decision-making processes. In addition, time studies were carried out to determine work productivity and cost parameters. These data were used to compare the pull-through delimber system with the processor-based processingsystem regarding productivity, cost efficiency, and operational suitability.
The results show that the processor-based system achieved higher and more consistent performance in terms of pure processing productivity. In contrast, the pull-through delimber system demonstrated considerable operational flexibility and adaptability. The pull-through delimber proved to be suitable for small to medium scale operations, under steep terrain conditions, structurally complex stands, and the production of treelength assortments. Furthermore, it was capable of handlinglarger stem diameters and could be deployed rapidly during salvage logging operations following disturbance events.
From an economic perspective, the analysis indicates that the pull-through delimber system can achieve competitive or even lower costs per cubic meter under favorable operating conditions. Cost advantages arise primarilyfrom the machine’s high mobility, lower fixed cost risks during idle periods, and the potential to generate additional revenue from chipped biomass. Considering the established markets for treelength and wood-chip products in southern Germany, the resultshighlight a significant operational potential for pull-through delimber systems in semi-mechanized harvesting systems.
Two-stage trucking systems in New Zealand forest harvesting operations have become increasingly common, with their adoption rising from 8% to 24% since 2016. Despite this growth, productivity modelling research remains largely reliant on elemental time studies, which typically generate small datasets covering only several days of work. This study presents an automated, data-driven framework for constructing a digital twin of forest road networks and applying it to two stage trucking operations to model truck productivity over extended periods.
Controller Area Network bus and GNSS position data were collected from a two-stage truck operating in the Hawke’s Bay region of New Zealand over a 230-day period across two harvesting sites. Latitude, longitude, altitude, and truck pitch data were filtered, resampled, and projected to a metric coordinate system. A raster-based skeletonisation approach was applied to buffered GNSS traces to reconstruct road centreline geometry. The extracted network was segmented into road sections and attributed with geometric properties including gradient, curvature, and length.
The segmented road network enabled each GNSS point to be mapped to a corresponding segment, allowing precise identification of truck position within the network. A segment-based state logic integrating road geometry and CAN data was developed to classify operational states (loading, driving loaded, unloading, driving unloaded, and idle). This allowed the capture of individual cycles without having to manual delineate the data depending on which landings the truck loaded or unloaded from. Segment-level metrics—including speed profiles, tonnage carted, and cycle times—were derived and linked to network topology to model productivity.
Results demonstrate that the digital twin road network and segment-based state logic enables productivity modelling without reliance on pre-existing GIS datasets or manual elemental time studies. The proposed method supports automated cycle detection, road topology performance benchmarking, and scenario testing for two-stage transport planning optimisation. The method could also be integrated into a live dashboard for real-time analytics and tracking of trucks during their operations.
Reliable and up-to-date forest road information is essential for sustainable forest operations, affecting technical road classification, timber transport accessibility, and maintenance planning. However, conventional road inventories are often incomplete, outdated, and costly to maintain due to their reliance on manual surveys and expert assessments. Over the past four years, we have developed RoadSens, a mobile proximal sensing and analytics platform designed as a technical solution for comprehensive forest road digitalization, enabling multiple operational applications from a single data acquisition workflow. This contribution synthesizes the development, methodological advances, and validation of the RoadSens platform.
RoadSens as a flexible to use vehicle attachment, integrating stereo vision spatial mapping, deep learning–based perception, and precise geo referencing through the fusion of post processed GNSS and stereo derived odometry. The platform provides a detailed geometric description of forest roads, including longitudinal slope, horizontal and vertical curvature, and cross sectional parameters such as road width, centerline position, sidefall slopes, and ditch characteristics. Validation across several forest roads in Norway demonstrates that the platform is capable of delivering consistent, spatially continuous geometric information suitable for operational use.
This geometric foundation enables multiple downstream applications. First, RoadSens supports automated technical road classification according to defined road standards, using geometry based criteria such as slope, curvature, and width. Case study results indicate that road width is often the dominant constraint in meeting technical standards, underlining the value of systematic and up to date geometric data. Second, RoadSens provides essential input for operational planning, including the assessment of accessibility and restrictions for timber trucks, by delivering road parameters relevant for vehicle performance and safety. Third, the platform has been extended towards maintenance planning through automated detection and mapping of surface deterioration features, which are aggregated into intuitive, segment level maintenance indicators.
Across all applications, the main current bottleneck is not the sensing or geo referencing concept, but the performance of object detection models used for road surface interpretation. While technically feasible, detection of irregular and heterogeneous deterioration features remains challenging. Importantly, this limitation is primarily data driven and can be systematically addressed through larger, more diverse, and higher quality training datasets, making the approach well suited for continuous improvement as additional data become available.
Overall, RoadSens demonstrates the potential of mobile sensing as a scalable and cost effective alternative to manual forest road surveys, providing a unified data basis for technical classification, operational planning, and maintenance decision support.
Forest cable yarders are a highly effective timber extraction system, particularly in steep mountainous terrains. Operating these systems demands high expertise, as does the preliminary planning and ground layout of the lines. Accurate field tracing and post-installation verification are critical to ensure technical feasibility (e.g., maintaining ground clearance for suspended loads) and operational safety regarding tensions and payload capacities. Traditionally, in Italy these operations are carried out directly by forest operators, except in the most complex cases where surveying relies on professional technicians conducting field visits. However, involving external technicians incurs additional costs and logistical delays for forestry companies. Conversely, when operators conduct the surveys without specialized tools, formal safety and documentation requirements are often not fully met. Therefore, there is a clear need to simplify this phase, and digital instruments offer highly functional solutions. While the digitalization of the forestry sector has led to a rapid proliferation of mobile applications, there remains a distinct lack of tools specifically tailored for the fast, simple, and immediate surveying of cable yarder lines. This study presents the development of a novel application for smartphones and tablets—named "Teleferika"—distinguished by its innovative, iterative design approach driven directly by the practical needs of forest operators. The development process began with a targeted questionnaire administered to experienced operators and forestry instructors. This user-centered approach directly informed the developers about the core functionalities to be implemented in the app. Primarily, the app allows operators to easily record fundamental line characteristics (such as anchor points and intermediate supports) during the layout and installation phases. The app was equipped with an intuitive graphical interface featuring real-time cartography, effectively functioning as a navigation system along the theoretical line. The collected spatial data can be exported in standard formats (.shp, .kml, etc.) and transmitted via email or messaging platforms to off-site technicians. They can then generate formal site documentation and perform rigorous payload verifications using specialized software like Seilaplan. Moreover, the initial operator survey revealed a strong demand for a modern alternative to the traditional compass for field tracing, an operation the app was expected to perform effectively and improve. In this sense, initial field testing alongside operators highlighted the inadequate accuracy of standard smartphone GNSS receivers. To resolve this, the system was integrated with low-cost RTK-GPS antennas, successfully refining positioning accuracy to approximately 1 meter. Throughout the development process, the application underwent multiple iterations based on continuous operator feedback. Finally, to ensure true practical viability, comprehensive usability tests were conducted. These trials not only identified the system's strengths and limitations but also provided valuable insights into the operators' reception of new technologies and their overall readiness for the digitalization of the forestry sector. In this contribution the development process and the results of both the functional and usability tests will be shown in detail.
AI models are becoming increasingly important in forestry operations, enabling robotic systems to automatically detect and classify forest features such as stems, logs, rocks, and roads from vehicle-mounted images. This automation enhances operational efficiency, reduces human error, and supports sustainable management through real-time monitoring and analysis. However, the development of such models is constrained by the limited availability of data, which is frequently incomplete or partially labeled, thereby limiting model generalization and adaptability.
To address these challenges, we propose an incremental learning strategy for object detection using the transformer-based detection model DETR, which allows the number of classes to be expanded over successive training phases. By fine-tuning DETR on new forestry classes while retaining knowledge of previously learned categories, our approach enables continuous adaptation to evolving datasets without requiring full retraining. The DETR model leverages the capabilities of state-of-the-art foundation models by using a fixed backbone, while transfer learning is applied to the detection head. This strategy allows for rapid adaptation to new datasets with minimal labeled data, maintaining high detection performance across multiple forestry classes.
Overall, this approach demonstrates the potential of combining transformer-based detection models with incremental learning to create flexible, efficient, and scalable solutions for automated forestry feature detection.
The forestry sector is undergoing a rapid digital transformation, driven by the demand for accurate, timely, and actionable data. Across public, private, and industrial operations, the challenge remains the same: bridging the gap between field data collection and the decision-making processes that guide sustainable forest management. This presentation highlights how G.A. Logix has developed modular digital solutions that streamline the entire chain—from field capture to operational and strategic decisions.
Our applications—MobileLogix, SyncLogix, and ForestLogix—enable operators, contractors, and managers to collect geospatial and operational data directly in the field, synchronize it across teams and machines, and analyze it in formats compatible with widely used GIS platforms such as Esri and QGIS. An AI layer allows natural-language queries and automatic creation of tailored reports (harvest progress, compliance summaries, cost dashboards), with human-in-the-loop validation. In addition, textual information and reports can be exported in any format required by organizations, ensuring flexibility for reporting, integration, and long-term records. By ensuring accuracy and eliminating delays, these tools empower organizations to optimize resources, improve transparency, and meet evolving sustainability requirements.
Examples from active operations demonstrate how digital workflows reduce errors, accelerate planning, and enhance collaboration between field crews and management. The presentation also explores the role of real-time communication and rugged field-ready hardware in enabling continuous operations under challenging conditions.
By bridging the gap between the forest and the office, G.A. Logix contributes to a more efficient, transparent, and sustainable future for forest operations worldwide. Attendees will gain insights into the practical benefits of digital transformation and how modular solutions can support the evolution of forestry practices in diverse contexts.
In Finland, the forest sector currently needs and will continue to need new workers. With demographic development, the age groups are becoming smaller and more urbanized, which means that young people may not have any contact with the forest sector before applying for education. Previous studies have emphasized that students’ perceptions of different educational paths and occupations can be easily influenced by prejudices. For some young people, the forest sector still appears conservative, traditional, and masculine. The forest sector is still highly gendered. Additionally, according to the previous studies, raising awareness of the forest sector is essential for young people to be able to apply for training in the field. Strengthening the attractiveness of the forest sector is important in preventing future labor shortages. Study counselling is a free-of-charge support included in education in Finland that helps students plan their studies, choose their education, get to know the working life, and plan their careers. The purpose of study guidance is to promote justice, equality, and inclusion. The criteria for good guidance state that it is multidisciplinary cooperation that supports the development of career-planning skills.
This study examined guidance counsellors' attitudes and knowledge regarding the forest sector. In addition, the study was conducted to determine how guidance counsellors perceive young people's interest in the forest sector. The data were collected through interviews in the autumn of 2025. In total, 49 guidance counsellors participated in the interviews, 23 of whom worked in general upper secondary schools and 26 in comprehensive schools.
Based on the preliminary results, the interviewees were very familiar with the educational opportunities in the forest sector. However, awareness of working life came to the fore, which should be given attention in the future. In addition to telling students about their studies, sharing concrete career stories, work tasks, and working weeks should also provide young people with information about what it means to work in the forest industry. The guidance counsellors emphasized that communication about the forest sector should be based on encounters between people. Despite this, almost half (47 %) of schools represented in the study had received a visit in the previous 12 months from a forest-sector organization, educational institution, forestry student, or other representatives to provide information about education and working life in the forest sector. Furthermore, 61 % of the general upper secondary guidance counsellors and 65 % of the comprehensive school’s guidance counsellors interviewed perceived the forest sector positively. Only two counsellors (4.1%) expressed a negative attitude towards the forest sector. The guidance counsellors perceived forests as highly or very significant, both at the individual level and for Finnish society.
Keywords: forest labor, forest education, guidance counsellor, working life
Forest operations are a cornerstone of the forest-wood sector, mobilizing tens of millions of cubic meters of wood each year. These resources serve multiple societal needs: timber for construction and furniture, industrial wood for paper, cardboard, and panels, and energy wood for heating and biomass. Sustainable forest management ensures that forests continue to provide these resources while preserving ecosystems.
Over the last 25 years, forestry has become increasingly technical and complex. Modern machinery now integrates digital technologies, GPS systems, and advanced harvesting tools. Scientific research has improved our understanding of how machinery impacts soils and forest environments, while regulations and certifications have strengthened sustainability standards. Market demands are also evolving, with growing opportunities in industrial wood and bioenergy, which require operators to adapt continuously.
Despite these advances, the sector faces a serious challenge: a shortage of young, trained professionals. Even the most advanced machinery and innovative techniques cannot reach their full potential if there are no skilled people to operate them. Young professionals are crucial not only to handle technical equipment but also to apply best practices, respect environmental standards, and innovate in response to societal expectations. Without them, efficiency, safety, and sustainability are at risk.
CEFOR (Common Digital Knowledge Base for Forest Operations) addresses this challenge directly. It provides a structured, practical, and accessible digital platform covering eight key thematic areas, from planning and site preparation to harvesting, transportation, and safety management. By presenting processes chronologically, CEFOR makes complex operations easier to understand and apply in real-life scenarios.
Developed with forestry professionals, CEFOR ensures that all content is both technically accurate and pedagogically relevant. Its goals are to train newcomers in modern forestry techniques, improve operational performance, and support sustainable, responsible forest management in line with environmental and climate goals. Case studies from real worksites help learners see the diversity of challenges and career paths in forestry.
CEFOR also serves as a tool to attract new talent. By showcasing the technical nature of modern forestry, the skills required, and the use of advanced machinery, the platform aims to inspire young people to consider careers in this essential sector. Partnerships with training centers, professional organizations, and cooperatives ensure wide accessibility and integration into education and professional development programs.
In short, CEFOR is more than a digital resource—it is a strategic solution to two critical challenges: the increasing technical demands of forestry and the urgent need to attract and train the next generation of professionals. Without a skilled workforce, even the most advanced equipment cannot be used effectively. By providing practical, structured, and accessible knowledge, CEFOR ensures that forest operations remain efficient, safe, and sustainable, while securing a workforce capable of meeting the challenges of tomorrow.
The aim of this contribution is to present the analysis of the professional interests and motivations of young people regarding forestry careers. An online survey (N = 162) was conducted in May 2025. The study identifies the profile of a typical forestry candidate as an individual from a rural area, men or women, who prioritizes job security and personal satisfaction over high financial reward. The findings confirm that while forestry students show the strongest interest in jobs in forestry, intrinsic motivation and a personal relationship with nature are the primary drivers across the board. The key results indicate that practical field experience and communication with forestry professionals significantly increase career interest. Furthermore, the emerging generation of foresters demonstrates an "ecologically realistic" perspective, favoring sustainable management and biocentric values over traditional paradigms. The study suggests that to combat the declining workforce, forestry communication should leverage these intrinsic values and increase practical engagement during university studies. Respondents assigned the greatest importance to statements reflecting biocentric and ecocentric values, emphasizing the intrinsic value of forests, ecosystem impact considerations, and their role in environmental balance and human well-being.
The need for safe, efficient and standardized training methods has promoted the use of virtual reality simulators as an alternative tool to conventional training, especially in high-risk operational industries. These simulators allow technical skills to be developed while reducing operating costs, physical and environmental risks, and machinery wear, and facilitate the controlled and safe repetition of tasks. In this study, the evaluation of the learning process and the acquisition of technical skills were analyzed through individual and group learning curves, with execution time as the main performance indicator for skill acquisition. Ten trainees with no previous experience in forestry machinery voluntarily participated in basic scenarios on a Komatsu simulator for harvesters (gripping and felling) and forwarders (loading and unloading). The aim was to develop basic operational skills, promote familiarization and adaptation, and teach control commands. The training consisted of a total of ten repetitions per trainee distributed across the six scenarios of varying complexity and duration. Exponential learning models were used to model the performance evolution, as they best fitted to the data, demonstrating that completion time-based learning was non-linear. The results showed a rapid and steep initial improvement followed by a phase of deceleration and progressive stabilization. Scenarios with greater complexity or those with longer execution times presented curves with slower stabilization and greater interindividual variability, suggesting grater cognitive and psychomotor demands, while scenarios with shorter execution times showed faster adaptation with less dispersion. All six scenarios showed a common pattern of individual and group learning with a progressive decrease in time through practice repetitions. These findings support the integration of virtual reality technologies, especially simulators, into training programs and the use of learning curves as a tool to evaluate performance improvements.
New Zealand’s forest harvesting workforce faces persistent retention challenges. The lack of clear forestry career paths, poor employment conditions, and limited awareness of opportunities have contributed to retention rates of only 30% after three years in the forestry industry (MPI, 2019; Workforce Insights, 2022). In a highly mechanised sector where operator capability is central to productivity and safety, low retention undermines both workforce sustainability and the return on training investment. This paper reports interim findings from a longitudinal evaluation of a work-based operator apprenticeship scheme that has achieved markedly higher retention than the wider industry.
The programme is delivered through an employer-integrated apprenticeship aligned to NZQA training requirements (NZ Certificate in Forest Harvesting). It combines structured block-course delivery, theory completion support, and practical on-the-job development, supplemented by simulator-based training and exposure to downstream supply chain requirements. The evaluation design follows successive cohorts over time, using repeated semi-structured interviews at key milestones (programme entry, end of theory blocks, completion of practical requirements, and post-completion follow-up). Interview recordings are analysed using template analysis to identify mechanisms that support skill development, job fit, and commitment to the industry, alongside barriers that reduce perceived value. Evaluation findings are triangulated with programme tracking data on completion and retention and with employer feedback.
Across five cohorts to date, programme tracking indicates completion/continuation rates of approximately 77% (30 of 39 participants) and industry retention of approximately 85% (33 of 39 participants), including evidence that some non-completers remain in the industry and some withdrawals re-enter the programme after re-employment. Participant accounts suggest that perceived value extends beyond technical skill acquisition to include clearer pathways and credentials, structured support to complete theory, increased understanding of “why” operational practices exist (including customer requirements) and strengthened social connection through cohort learning in an otherwise isolated industry. The least valued element is classroom-based “book work,” pointing to ongoing tensions in how theoretical components are delivered for a practically oriented workforce.
These interim results indicate that a workplace-embedded apprenticeship can function simultaneously as (i) a skills development pathway, (ii) an induction and socialisation mechanism, and (iii) a recruitment/selection filter that improves job–person fit—together contributing to substantially higher retention than sector benchmarks. Implications are discussed for designing operator development systems that strengthen both workforce capability and workforce stability in mechanised harvesting.
Receiver-reported Global Navigation Satellite System (GNSS) accuracy metrics (e.g., horizontal accuracy estimates derived from solution covariance) are sometimes used as proxies for positional reliability in forestry applications. However, their predictive validity under canopy and obstruction remains insufficiently quantified. This study evaluates the relationship between receiver-reported accuracy and empirically observed positioning error across four static environment classes representative of forestry sector conditions.
Static occupations were conducted on pre-surveyed control points in forest interior, forest edge, tree-adjacent, and building-adjacent environments. Horizontal error was quantified using Distance Root Mean Squared (DRMS), the average value for stand conditions reaching 2.3 meters. Error distributions were strongly non-normal in all environments (Shapiro–Wilk p < 0.001). Correlation between reported accuracy and DRMS was highly environment-dependent. In dense forest interior, linear predictive power was negligible (R² = 0.002; Pearson r = -0.05; Spearman ρ = 0.14). At forest edge, explanatory power remained minimal (R² = 0.005; Pearson r = 0.07; Spearman ρ = -0.44). In contrast, minor associations were observed in building-adjacent (R² = 0.10; Pearson r = 0.32; Spearman ρ = 0.51) and tree-adjacent conditions (R² = 0.19; Pearson r = 0.44), indicating partial covariance tracking under certain obstruction geometries.
Despite these differences in correlation structure, receiver-reported accuracy systematically underestimated empirical error in all environments. Median DRMS exceeded reported accuracy by factors ranging from approximately 2.6 (forest edge/tree-adjacent) to 8.0 (forest interior/building-adjacent). Dispersion inflation under severe obstruction was not proportionally reflected in reported accuracy metrics.
Dynamic measurements across 13 operational environment classes further demonstrated horizontal degradation relative to open-sky conditions and consistent vertical error inflation. Dynamic horizontal receiver-reported accuracy ranged from 273 mm in unobstructed locations to 758 mm in forested environments with vertical obstructions, while vertical accuracy varied between 397 mm and 1333 mm across the same conditions. These values highlight the impact of vegetation, vertical structures, and terrain slope on the GNSS solution's reported accuracy, and suggest that even when absolute positioning errors differ, reported precision can serve as a relative indicator of environmental complexity affecting receiver performance.
These findings indicate that receiver-reported GNSS accuracy exhibits environment-dependent predictive validity but consistently underestimates true positional error magnitude in obstructed forest conditions. Reported accuracy metrics should therefore be interpreted cautiously and not as absolute values. Given the generally weak or negligible linear correlations in highly obstructed environments, these metrics cannot be reliably corrected or rescaled to match true error; simple multiplicative adjustments will not produce accurate positioning estimates, they may however find a use as relative or qualitative metrics.
Modern Swedish forestry has contributed to a substantial decline in ground lichen, a key winter forage resource for semi-domesticated reindeer, while demand for timber and bioenergy continues to increase. This study develops and tests a generic, easily implemented GIS-based screening method to identify priority areas for lichen-adapted young stand management that can simultaneously support biomass utilization and reindeer husbandry in northern Sweden.
We conducted a case study on productive forest land owned by Sveaskog, overlapping the Ubmeje Tjällddie and Vapsten reindeer herding communities. The analysis was restricted to productive forest land, as only these areas are subject to regular management operations. Stand data included 32,347 stands (309,476 ha), of which 20,200 stands (237,372 ha) were classified as productive forest land. Young stands were defined as those with a need for pre-commercial thinning, initial thinning, or complementary thinning (management classes 21, 31, 32), while stands treated with fertilizer were excluded due to known negative effects on ground lichen.
We combined corporate stand data with spatial datasets on lichen coverage (10 m raster, continuous % cover), soil moisture (2 m, categorical), and management history (polygon attributes). Ground lichen potential was quantified by aggregating the lichen raster to stand level using a hierarchical threshold system inspired by the Swedish National Forest Inventory, but with expanded classes tailored to reindeer grazing needs. First, we calculated the proportion of pixels with ≥10% lichen cover and assigned stands to five lichen cover classes, recognizing that <10% cover is too low to be of grazing relevance. Second, we calculated the proportion of pixels with ≥25% lichen cover and grouped stands into four additional classes to highlight areas with higher forage quality. Third, we recorded the presence/absence of pixels with >50% cover, treating these as “hotspot” indicators due to their low frequency but high value for grazing.
This multi-threshold approach generates stand-level indicators describing both the extent and intensity of lichen occurrence, enabling screening of a wide combination of stands for lichen-adapted management. The resulting indicators can be used to rank and map priority stands for young stand treatments that increase light availability and improve accessibility, such as boom-corridor thinning and geometric thinning systems. These methods have shown potential to increase mechanized harvesting productivity in dense young stands, offering opportunities for additional biomass recovery while improving conditions for reindeer.
Ongoing work compares the mapped priority stands with an independently developed list of stands prioritized through a collaborative planning process between Sveaskog and the reindeer herding communities. This comparison will be used to validate the method’s relevance for operational co-planning at both stand and landscape scales. The proposed framework is replicable, relies on commonly available stand registers and remote sensing products, and can be scaled to other reindeer herding districts. It therefore offers a practical decision-support tool for integrating lichen-oriented, adaptive young stand management into forest operations while contributing to a growing bioeconomy.
Forestry operations require the creation and maintenance of internal networks called extraction trails (cloisonnements), which are essential for machine movement and silvicultural work.
Currently, marking and updating these networks in the field remains costly, time-consuming, and sometimes imprecise.
The SCAN project aims to modernize and optimize the management of these infrastructures by using remote sensing technologies, digital mapping, and machine guidance systems.
The main objectives are to:
• Reduce operational costs,
• Improve the precision of silvicultural operations,
• Limit the impact of interventions on soils,
• Improve working conditions for field teams,
• Ensure better continuity and traceability of data between operations.
The project is based on the systematic digitization of forest plots before machine operations take place.
This approach makes it possible to:
• Identify existing extraction trails or determine optimal routes,
• Prepare operations in advance through data analysis,
• Provide operators with digital materials that can be directly used in the field,
• Reduce manual marking and optimize machine movements.
The generated data are then stored and reused to improve long-term management and monitoring of forestry operations.
To address the diversity of forest conditions (terrain, stands, accessibility), the project aims to :
Test multiple technologies for integration into the reference data collection process
• High-density LiDAR: produces a detailed 3D representation of the terrain, capable of detecting fine structures and historical traces.
• Very high-density LiDAR (drone): provides detailed 3D terrain acquisition on demand, independent of temporal constraints.
• Drone photogrammetry: generates recent orthophotos, particularly useful in hard-to-access areas.
Additionally, a mobile or terrestrial LiDAR will be used as a reference for high-precision measurements and data validation.
Test guidance technologies for specialized machinery
• Guidance system (GNSS / RTK / UHF): enables precise vehicle guidance, including in areas without network coverage, using radio base stations.
Initial work shows that LiDAR data can :
• Reveal old or barely visible trails,
• Identify topographic irregularities,
• Generate digital terrain models and analyses (slope, shading, micro-topography).
Using these data in GIS software facilitates :
• The detection and routing of extraction trails,
• The production of operational maps and tools for field use,
• A significant reduction in the time required for field marking.
Machine guidance trials also confirm the potential of these tools, enabling more precise and consistent work, even in varied forest conditions or in areas without network coverage.
By combining remote sensing, geographic information systems, algorithmic processing, and machine guidance, the SCAN project proposes a renewed approach to forestry operations. This methodology aims to improve operational efficiency, reduce costs, preserve soils, and enhance working conditions in the field.
The labor shortage is becoming a growing problem in the European forestry sector. As current data collection techniques rely on field labor, data collection for the purposes of forest road design is more difficult than in the past. For that reason, there is a need to identify more efficient field data collection method. The previous method of field data survey in forest road design process was based on the use of classical survey methods (theodolite and level) and modern survey methods (total station). The mentioned methods provide sufficient accuracy during survey but are time-consuming and required a large number of auxiliary workers. The quantity of collected field data that classical and modern field survey methods enable us to do are incomparable to the quantity of spatial data collected by LiDAR sensors. LiDAR survey could eliminate the problem of labor shortage and the and reduce the time required to collect field data. The aim of the research is to determine the applicability of collecting field data using LiDAR sensors and the influence of vegetation on the results of field measurements in the process of designing and developing the main project for the reconstruction of a forest road. The research was conducted regular forest stand in moderately steep terrain of Republic of Croatia, the main tree species in the research area are: Fagus sylvatica L. and Quercus petraea Liebl.. Two different LiDAR survey method were tested: using an unmanned aerial vehicle (ALS) and airplane (ALSDGU). Based on the collected field data, two versions of the forest road reconstruction main/implementation designs were developed. Total length of reconstructed roads was 44+93.18 hm. Based on the main projects of the researched forest road, the amounts of earthworks required for its reconstruction were calculated. The research determined the level of influence of vegetation on the accuracy of the tested measuring equipment, which provided new insights and guidelines for choosing the most suitable method of collecting field data when designing forest roads.
The presence of standard-compliant forest road networks is essential for the sustainable management of forest ecosystems and the efficient execution of forestry operations. Enhancing the resilience of these roads to environmental impacts, preventing water erosion, and extending their service life depend on positioning engineering structures at the optimal topographic locations along the route. Identifying the locations of these structures using conventional ground-based surveying techniques is a labor-intensive, time-consuming, and costly process. This study aims to determine the optimal locations for the engineering structures required on forest roads using unmanned aerial vehicle (UAV) photogrammetry. Within the scope of the research, UAV flights were conducted in selected study areas to acquire high-resolution aerial photographs. These images were processed using photogrammetric software to generate digital elevation models (DEM) and orthomosaics. Based on the generated data, hydrological and geomorphological analyses were performed in a geographic information systems (GIS) environment to identify the locations of the drainage and retaining structures. The points obtained from the analyses were validated in the field, and the reliability of UAV technology as a decision support tool in forest road infrastructure planning was evaluated. The findings indicate that UAV photogrammetry provides data in a shorter time, at a lower cost, and with higher spatial accuracy compared to traditional methods. The study concludes that UAV technology is an effective decision support tool that saves time and labor in the process of identifying engineering structures.
Modern cut to length (CTL) harvesters increasingly incorporate positioning information as part of their standard data output. This information ranges from machine traces that continuously log machine movement, to base machine positioning during felling, to the positioning of stumps and, ultimately, each processed log. While high accuracy GNSS receivers with coordinate corrections can theoretically provide centimetre scale accuracy for base machine positioning, stump and log positioning require additional knowledge of boom tip location and head orientation—factors that introduce further uncertainty. Understanding the performance of these systems is important when developing practical use cases.
In late 2025, the Wood Value Trial 2025 was conducted north of Uppsala, Sweden, with the aim to assess the performance of modern CTL harvesters across several key categories, including measurement precision (length and diameter), bucking performance, incidence of timber damage, and the capability and accuracy of stem and log positioning. Seven manufacturers participated in the trial, Log Max, Ponsse, Logset, John Deere, Ecolog, Rottne and Komatsu. All harvesters provided base machine positioning during felling, three included functionality for stump positioning, and one supported log-positioning. Stump positioning were evaluated by comparing recorded stump coordinates with control measurements collected using a handheld GNSS receiver with centimetre scale accuracy. The same comparison method was applied to evaluate log center positioning. For the stump positioning a sample of 30 stumps were evaluated for each manufacturer while the evaluation of the log positions consisted of 22 logs.
Results showed average stump positioning errors ranging from 0.16 to 1.38 m, with standard deviations between 0.11 and 0.73 m. For log positioning, the average error was 0.42 m, with standard deviations of 0.45 m and 0.26 m along the principal axes.
As a demonstration of a practical application using positioning, the Vinnova funded DigForeTrace and Mista Digital Forest projects used log positioning to enable identification of individual logs by forwarders, allowing stem specific information gathered by the harvester to be connected to individual logs carried out to roadside locations. As the identities of logs were known so too were the harvester measurements and combined with industry gate data, logs could be matched to a reasonable subset of candidates using length and diameter attributes. This passive tracing method produced lists of suggested matches for each log, recovering 67–76% of logs with an average of 3–33 suggestions per log. Because each candidate’s unique harvester identifier is retained, the full suite of harvester recorded positioning data can be used to visualize where each log was cut and where each stem originally grew showcasing a use case for precise positioning information.
Weather events such as tornados and hurricanes can significantly impact forest stands and create operational challenges for harvesting and utilization. Rapid salvage following events is critical in the southeastern US due to the onset of insect damage and wood discoloration due to blue stain fungus. In September 2024, Hurricane Helene impacted the region as a Category 4 hurricane, with recorded wind speeds of up to 225 kph. The storm caused an estimated 2.6 billion USD impact to the forestry sector in the state of Georgia alone. The urgency of salvage operations makes it difficult to study harvesting productivity, cost, and value recovery resulting in a significant need to better understand effective forest management strategies post-disaster. This study aimed to evaluate the use of cut-to-length harvesting systems in these unique conditions. Specific objectives were to assess individual machine and system productivity, system cost, and the effect of differing tree damage classifications on harvester productivity. We selected two loblolly pine (Pinus taeda) stands in the lower Coastal Plain physiographic region of Georgia, USA, that sustained significant damage from the hurricane. We conducted time studies of a Ponsse Ergo harvester and Elephant forwarder. Data collection consisted of an elemental time study for the three main functions: felling-processing, forwarding, and loading. Time study data combined with stem volumes from onboard computers and load weights from mill scale tickets were used to estimate individual machine and system productivity. Results indicate that overall system productivity was 14.19 m3 SMH-1, onboard truck cost was 16.03 USD m-3 and cut and haul costs were 24.03 USD m-3 which is competitive with published rates for the region. This study provides insight into the feasibility of using CTL harvesting systems in challenging salvage logging conditions following storm events and will aid forest managers and logging contractors in understanding productivity and cost expectations.
Maintaining the competitiveness of the forest sector requires a holistic understanding of interactions across the forest-to-value chain, particularly within harvesting and transport operations where inefficiencies can propagate system-wide effects. This work presents a conceptual framework based on systems dynamics and systems-based thinking to investigate structural behaviour, feedback mechanisms, and bottlenecks in forest value chains. The approach integrates causal loop diagrams, stock-and-flow modelling, and scenario simulation to represent nonlinear relationships among forest resources, harvesting capacity, logistics performance, operational costs, and supply reliability.
The proposed framework is designed to be combined with artificial intelligence and optimization methods, enabling data-driven parameterization, predictive analysis, and evaluation of alternative decision strategies. Machine learning techniques support estimation of uncertain variables and forecasting of system responses, while optimization models assess harvesting allocation, fleet deployment, and transport planning under operational constraints.
Preliminary results from case studies in Australia illustrate the applicability of the framework and demonstrate its potential to identify leverage points, inefficiencies, and unintended consequences within forest operations. These early findings highlight how integrated systems modelling can provide insights not captured through conventional analytical approaches. The study positions the framework as a decision-support tool to assist industry stakeholders in diagnosing system limitations and designing interventions that improve productivity, resilience, and long-term competitiveness across the forest sector.
Selecting a timber harvesting system is a context-dependent decision that requires substantial expertise to balance economic performance, environmental impacts, operator safety, and technical feasibility. In practice, system choice is influenced by stand and site conditions (e.g., slope, extraction distances, soil conditions), machine characteristics and operational constraints, and trade-offs between sustainability dimensions are inevitable. This work presents the concept and architecture of a decision support system (DSS) that supports harvesting planners in selecting the most sustainable harvesting system – i.e., machine combinations for felling/processing and extraction – for a given stand.
The DSS follows a multi-criteria decision analysis (MCDA) framework. Sustainability is represented by nine core criteria operationalized through 15 indicators covering the three pillars (economy, environment, social/work safety). The workflow comprises seven steps: (1) User Input: scenario definition (stand, site, operation) and criteria weighting, (2) configuration of potential machine chains, (3) feasibility screening, (4) calculation and harmonization of indicator values using an evidence-based data layer, (5) transformation of indicators into normalized utilities, (6) weighted aggregation into overall scores and rankings, and (7) reporting of machine chain ranks and trade-off visualizations.
The MCDA framework is underpinned by a data layer that provides reference values and parameter-dependent ranges for indicator estimation, while feasibility screening uses these operating envelopes to constrain system recommendations under the specified conditions. Using this foundation, a combination of data-driven and process-based models is employed to derive indicator values from user-defined stand and operational inputs. These indicators form the basis for multi-criteria evaluation and optimization, allowing machine chain alternatives to be assessed and ranked according to trade-offs between economic, environmental, and social objectives. The relative importance of these objectives can be adjusted through user-defined criteria weightings, allowing the system to reflect different stakeholder priorities and decision contexts. The overall aim is to generate context-specific machine chain recommendations that balance sustainability dimensions in timber harvesting operations.
Our contribution highlights the modular system design (data layer, rule layer, formal model and user interface) and discusses open research needs, in particular systematic evidence expansion, harmonization standards, and the derivation of a minimal dataset to guide future field data collection and model calibration.
Traditional forest measurement techniques which are carried out using calipers and diameter tapes are accurate but they require a lot of manual work in large scale operations. The recent advancements in mobile LiDAR technology which can be installed in consumer-grade devices like smartphones and tablets have introduced new possibilities of quick and remote tree measurement. ForestScanner and Arboreal are two major tree measurement applications based on the in-built LiDAR sensor in Apple devices which are used to approximate the diameter at breast height (DBH) using point cloud-based circle reconstruction techniques. The mathematical comparison of such tree measurement applications in similar field operation environments is still inadequate.
This study compares ForestScanner and Arboreal in the measurement of DBH in natural mixed forest stands at Muratdere Forest Enterprise Chief, Bursa Province, northwestern Turkey. Distribution of natural stands of Scots pine (Pinus sylvestris L.) and Uludag fir (Abies nordmanniana subsp. bornmuelleriana Mattf.) is between the elevations of 871 to 1578 m. Three methods were used to measure 500 trees in 20 sample plots: (1) a traditional diameter tape as a reference, (2) the ForestScanner application and (3) the Arboreal application. To eliminate any variations in hardware, the two mobile LiDAR applications were operated on a similar iPad device with a LiDAR sensor, and the differences observed were due to the software algorithms.
Pearson correlation coefficient, R squared, MAE, RMSE, bias and MAPE will be used to evaluate accuracy. Paired t-test analysis will be used to compare the methods. We also compare the method efficiency by recording the total measurement time per sample plot using each method. Our previous research in this area showed that ForestScanner is highly accurate in DBH estimation (r = 0.990, RMSE = 1.49 cm, MAPE = 2.82%) and 65% reduction in time compared to the traditional measurement. The current study extends this work by introducing Arboreal and directly comparing both tools under identical conditions. Particular attention will be given to how application-specific algorithms influence accuracy across tree species, diameter classes, and topographic conditions. The research aims to offer practical guidance to the forestry professionals who would like to integrate mobile LiDAR tools in their measurement process.
Manual handling in forestry and nursery work involves repetitive low postures, trunk bending, and short-distance load carriage, all of which can elevate the risk of musculoskeletal disorders (MSDs). Passive exoskeletons are sometimes suggested as a practical way to reduce postural exposure and strain, but their effectiveness varies depending on the job. Their effectiveness depends heavily on the activity required movements (i.e., the task pattern) and the weight of the load being handled.
So far, either the kinematics of movement or the ergonomic risks associated with work performance and specific tasks are examined. In this laboratory study, we compared the two methods simultaneously by quantifying the influence of different passive exoskeleton models on kinematic exposure metrics and ergonomic risk indicators during two simulated manual task sets.
The first, Task Set A, simulated tree planting and included low-load repetitive handling, repeated trunk flexion and kneeling, as well as upright walking over 20 meters while placing small items (0.5 kg every 2 meters; 1 kg every 4 meters). Task Set B (Moderate-Load Handling and Carriage) simulated plant seeding and included repeated lifting of loads with deep squatting, short-distance walking, and lifting/carrying loads of 5, 10, and 15 kg over 4 meters.
Participants performed the protocol both with and without an exoskeleton (two models for Task Set A and three models for Task Set B). Whole-body motion was digitally recorded at 60 Hz samples using Inertial Movement Units-based motion capture and converted into 3D ergonomic joint angles and movement acceleration.
To provide a comparable assessment, we analyzed several kinematic exposure outcomes, including: (i) time spent in trunk flexion/extension, (ii) trunk flexion/extension dose (time × depth in degrees), (iii) relative flexion/extension between lower extremity segments and pelvis, (iv) upper arm positions and flexion/extension dose, and (v) trunk sway, pelvis movement, and knee position as a stability-related proxy during walking/carrying. We contextualized kinematic exposures using ergonomic risk evaluation frameworks, including ISO 11226 (angle–time postural loading) and EN 1005-4 (repetition characteristics). We complemented these with time-based distributions consistent with OWAS/RULA logic for identifying high-risk postures. Condition effects were summarized as within-participant differences versus the no-exoskeleton condition, with confidence intervals.
The results show that the effects of passive exoskeletons vary across the equipment × task combinations. In the low-load pattern, which is dominated by bending and kneeling, the main potential benefits are reducing the movement dose and improving walking and activity stability. These are closely related to body strain and performance in repetitive tree nursery activities. In the moderate-load pattern, 'effectiveness' is better judged by whether the exoskeleton allows efficient squatting and stable carrying without increasing demands for upper-arm elevation or trunk sway/rotation, rather than by reducing trunk bending alone. In practice, these findings support the selection of exoskeleton models based on dominant task patterns and expected load ranges. They also emphasize the need for field testing on uneven terrain and with full protective equipment and working tools before operational implementation.
improving productivity and operational efficiency. However, despite increasing levels of mechanization, transport-related activities such as log extraction, uphill hauling, load stabilization, and auxiliary water transport continue to expose workers to significant ergonomic and safety risks. Challenging terrain conditions, whole-body vibration, repetitive movements, awkward postures, and time pressure contribute to musculoskeletal disorders and long-term occupational health problems in forestry operations.
This study proposes an expert-based ergonomic risk evaluation framework for mechanized forestry transport activities. Rather than relying solely on accident statistics or large-scale datasets, the research emphasizes structured expert judgment and systematic risk scoring to identify and prioritize critical ergonomic hazards in field operations.
In the first stage, a panel of forestry engineers, machine operators, and occupational health and safety specialists was consulted to identify key ergonomic risk factors. These factors included posture load, repetition frequency, manual force exertion, vibration exposure, slope gradient, task duration, and environmental constraints. A structured evaluation form was developed based on these parameters.
In the second stage, ergonomic risks were assessed using the Hazard Rating Numbering System (HRNS). Each transport task was evaluated in terms of severity, exposure frequency, and probability. Risk scores were calculated and classified into risk categories (low, medium, high, and very high). This approach enabled systematic prioritization of transport-related activities under mechanized forestry conditions.
The findings indicate that tasks performed on steep slopes, repetitive hose handling in water transport, and manual-assisted log stabilization present the highest ergonomic risk levels. Whole-body vibration and prolonged static postures were identified as critical contributing factors. Expert consensus emphasized the need for ergonomic redesign of operator cabins, improved vibration damping systems, better task rotation planning, and slope-adaptive operational planning.
The proposed framework provides a practical and implementable decision-support tool for forestry managers without requiring complex data infrastructure. From a forestry mechanization perspective, integrating structured expert judgment into risk assessment contributes to safer operational planning, reduction of musculoskeletal disorders, and sustainable workforce management.
This study highlights the importance of proactive ergonomic evaluation in mechanized forestry systems and demonstrates how expert-based assessment can support safety-oriented mechanization strategies in challenging field environments.
Overseas studies and emerging data from within Aotearoa New Zealand indicate that those working in forest environments experience a significantly elevated risk of suicide. Yet the limited research conducted within global forest industries has primarily focused on psychosocial conditions and their health impacts, under-recognizing how structural and cultural forces interact to shape workers’ responses to stress and distress.
This presentation offers an explanatory account of how stress and wellbeing are constructed by logging machine operators, drawing on constructivist grounded theory to centre the lived experiences of 27 participants. The analysis demonstrates that wellbeing arises from the dynamic interplay between an operator’s habitus, the embodied dispositions, values, and cultural logics shaped through the logging lifeworld, and the structural determinants of the psychosocial conditions they face. This challenges the prevailing assumption that mental health outcomes in high-risk industries can be understood primarily as responses to environmental stressors, showing instead that they are culturally mediated and shaped by how operators navigate production pressures, hierarchy, identity, and control.
This reframing responds directly to the conference theme of enabling resilience through innovation by demonstrating why effective suicide-prevention and wellbeing initiatives must address both structural determinants and the cultural logics that shape coping. It highlights how the same psychosocial environment can produce markedly different wellbeing outcomes depending on the operator’s habitus and their position within the crew and contracting hierarchy.
This perspective provides a foundation for co-designing interventions that are both culturally grounded and structurally informed. A brief case example, drawn from a separate study that developed a video-based operator training system, illustrates how centring operator habitus and lived experience can inform more effective, context-aligned intervention design. Together, these insights offer a reframed understanding of suicide risk in forestry and point toward innovative, lived-experience-led approaches to prevention in high-risk work settings.
To ensure the continued functionality and intended purpose of forest road infrastructure, regular maintenance interventions are required. Maintenance operations employ technologies with differing levels of mechanization, which subsequently influence task difficulty and workers’ exposure to occupational noise. The study examined a technological system representative of practices in small private forestry enterprises in Slovenia, comprising three machines: a tractor (FENDT Farmer 309 LS Turbomatik) equipped with boom-mounted trimmer and grading attachment (KAHLBACHER), a truck (MAN 414 F2000), and a roller (BOMAG BW‑120 AD‑4).
Measurements of heart rate and noise exposure were performed over a total of 4.8 hours on three workers across individual maintenance phases. The results indicated that the physiological workload—measured as the percentage of Heart Rate Reserve (HRR)—reached 33.6% during non‑productive time, which is 8% higher than during productive tasks. Within productive phases, the highest HRR value occurred during manual road cleaning following trimming (56.8%), whereas the lowest was recorded during grading (19.6%). Noise exposure (LAeq) during non‑productive periods was 81.9 dB(A), exceeding productive‑time exposure by 3.2 dB(A). Among productive tasks, the highest noise level was recorded during rolling with a roller without a protective cabin (88.4 dB(A)), and the lowest during manual road cleaning (70.5 dB(A)).
The findings suggest that, when considered as an integrated process, forest road maintenance does not exceed established thresholds for physiological workload and occupational noise exposure (HRR = 40%; lower exposure limit for LAeq = 80 dB(A)). However, these limits are exceeded at certain workstations and during specific operational phases, necessitating the implementation of preventive measures to mitigate overexertion and overexposure. The study further indicates that with increasing levels of mechanization, excessive workloads tend to shift from productive to non‑productive periods, highlighting the need for active monitoring and management of activities conducted during non‑productive time.
The layout configuration of machinery in forest products manufacturing facilities plays a critical role in occupational health and safety, operational efficiency, and production sustainability. In sawmills, timber processing plants, panel production units, and surface treatment lines, machine placement is generally designed according to production flow and space constraints. However, ergonomic risk factors such as awkward postures, repetitive movements, excessive material handling distances, limited visibility, noise exposure, and vibration are often insufficiently considered during layout planning. This may lead to work-related musculoskeletal disorders, increased accident risk, productivity losses, and long-term operational inefficiencies.
This study proposes an expert-based Multi-Criteria Decision-Making (MCDM) model to optimize ergonomic machine layout alternatives in forest products facilities. In the first phase, relevant ergonomic and operational criteria are identified through a comprehensive literature review and field observations. The proposed criteria framework includes: (i) operator reach distance, (ii) material handling and transportation distance, (iii) visibility and control field suitability, (iv) noise and vibration exposure level, (v) emergency accessibility, and (vi) maintenance accessibility. These criteria aim to integrate both safety-oriented and production-oriented considerations into a unified decision structure.
In the second phase, expert evaluations are collected from professionals in forestry engineering, occupational health and safety, production management, and ergonomics (minimum 10-15 experts). Criteria weights are determined using either subjective approaches such as AHP or Fuzzy AHP, or objective weighting techniques such as Entropy or CRITIC, allowing a comparative assessment of weighting strategies. Subsequently, alternative layout scenarios such as linear flow layout, U-shaped layout, and cellular layout are developed for a representative forest products production system. These alternatives are ranked using TOPSIS or Fuzzy TOPSIS to determine the most ergonomically and operationally suitable configuration.
The proposed model provides a structured decision-support framework that simultaneously minimizes ergonomic risk and supports operational efficiency. A sensitivity analysis is conducted to evaluate the robustness of ranking results against changes in criteria weights.
Unlike conventional forestry mechanization studies that primarily focus on productivity and machine performance, this research integrates ergonomic risk assessment and occupational safety into layout optimization. The findings contribute to the development of sustainable and human-centered mechanization strategies in forest products manufacturing. The proposed approach is practical, adaptable to different facility types, and aligned with the objectives of the FORMEC symposium in promoting advanced and responsible forest mechanization solutions.
Precision forestry seeks to improve the sustainability and efficiency of forest management by applying new technologies to characterise individual trees. Accurate tree-level characterisation, typically through remote sensing, facilitates optimising timber yield and minimising waste. To that end, optimal bucking methods can be applied given accurate enough estimates of tree features. Among these features, stem shape and branching patterns are critical for stem bucking, as they determine the wood quality and the dimension and amount of products obtained. Yet, deriving this information from 3D point clouds remains challenging due to noise, non-circular cross-sections, and the difficulty of separating stems from branches.
In this study, we developed a novel stem shape characterisation and bucking method for tree point clouds capable of accounting for noisy data and irregular or curved trunks. Furthermore, we integrated tree branching information to estimate branch insertion points, find co-dominant or forked trunks, and gauge crown size. To do so, we leveraged graph theory to segment tree branches and isolate stem points. We then fit robust splines recursively across and along the stem, capturing curvature and irregular growth patterns through a recursive procedure. The resulting stem geometry and branching metrics served as inputs to the optimal bucking algorithm. We validated the proposed method on simulated point clouds, as well as LiDAR data acquired using a Mobile Laser Scanner (MLS) in Outokumpu, Finland. Our method successfully integrates branching patterns into bucking solutions, advancing tree-level decision-making. Moreover, it provides valuable insights into tree morphology for forest resources and biodiversity inventorying.
In Sweden, half of the forestland is owned by approximately 300,000 individual private forest owners. In total, they spend about 11 million work hours per year on forest management activities in their forests. Each to fulfill their own personal goals with forest ownership. Typical activities are planting, pre-commercial thinning and undergrowth clearance. Contractors are often hired for logging, meaning that forest owners also are customers of forestry services and need to evaluate who to contract and be able to assess the services provided. Today, this is often done subjectively and not based on objective information. This not only risks poorer target achievement for the procured services, but also generally gives an incomplete picture of service quality. Many of today's private forest owners lack forestry education and experience but are motivated to take advantage of most of the forest's values and protect high cultural and environmental values. One way to make it easier for small-scale forest owners to make good decisions and achieve high goal fulfillment is to innovate and implement new user-friendly tools such as digital decision support. The objective of this study was thus to identify and quantify the needs for digital decision support that can improve forest owners' purchasing skills when procuring forest services and facilitate the implementation and evaluation of forestry activities they carry out themselves. Data on forest owners’ current perceptions and use of digital decision support was collected through a questionnaire that was sent out to 1495 randomly selected non-industrial private owners in Sweden. The response rate was about 25% (n=366). The results show that about one third of the responding forest owners today use digital decision support for one or several purposes. The most common use was when assessing the need for forestry management measures on the forest property, and as a support when contracting service providers. About 40% of the users used digital decision support when following up on actions taken or as support when carrying out forestry work themselves. Almost as many used it for selecting the best management option based on their goals. Most users experience that digital decision support saves them time compared to other approaches, and that it contributes to better decisions and increased quality. Among non-users, it was more common for them to feel that they lacked the necessary skills to use digital decision support than that their use would be limited by a lack of technical conditions. Moreover, many non-users did not believe that digital decision support would make it easier for them to adapt their forestry to their goals. The study concludes that there is great potential to increase the use of digital decision support among small-scale forest owners, but that both educational and information efforts are needed to increase knowledge and raise awareness of the existence and benefits of these tools. By highlighting the needs of different types of forest owners, this study can contribute to the development of practically applicable digital support systems that can simplify and rationalize the forest owner’s work.
The use of the cut-to-length method as a wood harvesting system creates a demanding operating environment for harvester operators, especially in the Nordics. In addition to the harvesting method itself, high productivity demands and environmental conditions such as terrain, weather, vibration and cabin tilting increase the demands of the operator’s work. This environment places mental, physical, and cognitive strain on the operator, making the work holistically challenging. The workload of harvester operators can be reduced through various operator assistance systems. Currently, harvesters are already equipped with some assistance systems such as boom-tip control, automated bucking system, a reversing camera, and GPS positioning and maps. These systems support the operator’s decision-making and machine control. In the future, operator assistance systems will increasingly utilize forest machine data produced by various sensors, reducing unnecessary tasks and workload for the operator.
The objectives of this study were to determine what kinds of operator assistance systems will be needed in the future, how information provided by these systems should be presented, and what types of support operators require in their work. In addition, the study examined which assistance systems and advance information harvester operators currently use and how they receive feedback on their work. The interviews included 50 harvester operators across Finland and were semi‑structured, incorporating a two‑step Likert scale, with additional open‑ended questions added to complement and deepen the information obtained from the Likert‑scale items. The two‑step Likert scale was adapted from the Importance–Performance Analysis approach and analysed within this framework, enabling a combined assessment of how important each factor is to operators and how well it currently performs.
According to the results, harvester operators wish for assistance systems to support controlling thinning intensity, reporting harvest quality, and identifying tree characteristics and quality. These systems rely on data collected by mobile LiDAR scanners, which capture precise 3D measurements of the forest stand and make it possible to automatically identify tree attributes relevant to thinning and quality assessment. Operators were generally satisfied with the current method of presenting information on a traditional computer display, but they also expressed interest in new presentation methods such as AR‑glasses and head‑up display (HUD) systems. Among the existing assistance systems, GPS positioning and maps were considered the most essential. Boom-tip control was also regarded as useful, although operators felt that the system still requires further development. Receiving advance information about legally protected forest nature sites, buffer zones around water systems, and roadside landing areas was considered particularly important. In addition, based on the interviews, operators expressed a desire to receive regular monthly feedback on their work.
This study provides new, up‑to‑date insights into the needs of harvester operators regarding the development of operator assistance systems, thereby supporting the future development of forest machines in collaboration with manufacturers, contractors, operators, and researchers.
Working conditions in forests are changing mainly due to the current state of forests as a result of climate change. More deadwood and pre-damaged wood are being felled, and there is more deadwood, especially in the crowns of the deciduous trees to be felled. As the risk increases, so does the potential danger of accidents. The results of the KWF accident statistics (https://unfallzahlen.kwf-online.de) show that there is an increased risk of accidents when felling trees manually with a chainsaw within a 5 m radius of the tree. For this reason, mechanical felling wedges are required to fell the tree manually from a safe distance with reduced risk.
The KWF e.V. is aware of the current state of knowledge in this area through previous research work and exchanges with partners in the field.
A basic data analysis was carried out as part of a bachelor's and master's thesis by Mark-Fabian Franz in 2020 and 2023. The topics covered were the required lifting forces and distances for different tree lean angles and for different hinge (height and width). One result was the creation and development of an online calculator to determine the required lifting force for a hanging tree. (https://hubkraftkalkulator.kwf-online.de/)
Furthermore, findings on the changing lifting forces and distances with changing dimensions of the hinge (height and width) were recorded.
The resulting knowledge about the interactions between tree species and tree condition as well as lifting force and distance can be used to optimize the selection of tools, reduce misjudgments and thus significantly increase occupational safety.
The development and application of the measurement methodology for lifting force measurement also took place in this work. Current literature on further studies reviewed and analyzed within this work. Furthermore, practical partners, manufacturer contacts and interviews with practical partners as well as literature from past decades are available to take into account the development of break bar geometries.
However, further necessary measurements and comparisons with practice are still lacking. The determination of lifting force has been published internationally on several occasions. It is necessary to investigate which measurement methods were used in each case. Furthermore, the differences between the various measurement methods used in terms of results should be explained.
A meta-analysis is currently being prepared on this topic in order to compare the measurements taken. The research question is: ‘How can the current state of knowledge about lifting forces and distances be further developed and the measurement methods improved?’
Another point is the consideration of the lifting force required for tree felling with a lifting cylinder compared to the lifting force required for tree felling using a technical felling wedge. The hypothesis here is that there is a slight difference in the lifting force required for tree felling using a lifting cylinder compared to that required using a felling wedge. However, this will be irrelevant in practice due to the necessary safety margin.
Digitalizing the forest supply chain is essential for building resilient value chains that can handle increasing regulatory and environmental complexity. However, implementation is currently hindered by fragmented systems and breaks in communication between harvesting, logistics, and processing. The lack of integration, which is often caused by proprietary data formats, undermines the digital continuity required for transparent operations and compliance with stringent mandates such as the EUDR and the Digital Product Passport (DPP).
To address these challenges, this study proposes an innovative approach to ensure interoperability using the international StanForD standard. The solution involves integrating robust Logscom colour-spray modules directly into harvester aggregates or the forwarder's stake grid. These modules automatically apply alphanumeric identification codes to the cross-cut section of the timber and creating a persistent physical-digital link. This mechanical application is designed for the harsh environments of forestry operations, ensuring resilience against weather and handling.
By linking machine-independent data, including georeferenced harvest locations, dimensions and species, to each unique log, the system enables seamless traceability. Upon arrival at the sawmill, the markings are captured via automated optical recognition, linking the physical material back to its digital twin. To guarantee maximum data integrity and proof of origin, selected datasets are secured via EBSI blockchain notarization. This end-to-end integration eliminates manual documentation errors, strengthens the resilience of the chain of custody and provides a scalable foundation for a sustainable, verified forest-based bioeconomy.
In Japan, woody biomass power generation is increasing due to the Feed-in Tariff (FIT) scheme. However, woody biomass power generation suffers from low energy efficiency. Therefore, improving energy efficiency through combined heat and power (CHP) generation is being considered. Preferential treatment under the FIT program for small-scale facilities (generating less than 2 MW) has begun, and the introduction of gasification CHP systems, suitable for small-scale applications, is advancing. However, gasification imposes stringent requirements on fuel, leading to difficulties in procuring it and soaring fuel prices that cause business failures. Fuel chips for gasification require minimal fines and a uniform, relatively larger particle size. Large chippers designed for paper mill chips can produce a large proportion of relatively large chips. However, not only do these chips fuel applications compete with paper applications, but the production scale also far exceeds the demand of small-scale gasification CHP plants, leading to price increases. Conversely, medium- and small-sized chippers, which produce quantities suitable for small-scale gasification CHP, tend to generate more fine particles and struggle to produce relatively large chips. This fact necessitates screening to select chips suitable for fuel production, further contributing to fuel price increases. Therefore, a medium-sized drum-type chipper was modified to produce chips suitable for gasification CHP. Preliminary test results showed that chips produced by a commercially available medium-sized drum-type chipper met the ISO P31s specification. However, compared with the chip requirements specified by various gasification CHP boiler manufacturers, the chips contained too much fine material and lacked sufficient larger chips. We therefore conducted development and testing to reduce fine material and enable the production of larger chips. To reduce fine particles, the number of blades was reduced from 6 to 3. To produce larger-sized chips, the rotor speed and feed rate were optimized. Using a chipper equipped with a standard drum and a chipper equipped with a prototype drum featuring 3 blades, and two types of screens, we conducted chipping operations on sets of 10 Japanese red pine logs with a butt diameter of less than 30 cm and a length of 2 m. We observed the operations and collected samples of the produced chips. Each condition was tested three times. After returning to the laboratory, the samples were dried in the oven, weighed to determine water content, and then sieved. The particle size analysis showed that reducing the number of blades significantly decreased fine particles and increased the proportion of larger chips; however, this still fell slightly short of the gasification CHP boiler manufacturers' requirements.
The Birds Directive protects wild bird species, their habitats, and prohibits deliberate disturbance during the nesting season. Forest operations—particularly timber harvesting—during the nesting season may have negative impacts on bird habitats, breeding success, and populations. The recent decision of the EU Court of Justice creates a need to specify and clarify national regulations regarding logging practices during the breeding season also in Finland.
In Finland, the nesting season typically extends from April to July, although its exact timing varies by region and species. Limiting logging, forest management, and regeneration activities during this period would have substantial implications for forest management practices, the forest industry, and the availability of wood for industrial processing and energy use.
Wood harvesting in Finland is seasonal, with a greater emphasis on the winter season. Additional restrictions during the nesting season would further intensify this seasonality, increasing the demand for forest machines and their operators during a short winter window. Timber storage needs would rise, affecting an increase in costs of harvesting, logistics and storage. Concurrently, there would be an increase in winter transport volumes, thereby amplifying seasonal pressure on timber trucks and labour availability.
We examined the impacts of harvesting restrictions during bird nesting-season on wood supply, forestry operations, and the forest industry in Finland using predefined scenarios. These scenarios were compared with a baseline to assess cost impacts, labour and machine demands, and effects on production volumes and value added in the wood-processing industry.
Scenario analysis of the most extreme change scenario showed that eliminating all summer logging would increase the unit timber harvesting and storage costs (incl. logging, trucking and storage costs) by around 6% and reduce annual wood harvesting by approximately 25%, resulting in substantial economic impacts.
In an alternative scenario where 50% of the harvesting potential during the nesting season were shifted to harvest to winter (January–March), the unit harvesting and storage costs would increase by around 17%. This shift would require an additional 3,723 operators (including harvester, forwarder, and timber truck operators) and 2,383 additional machines and vehicles during winter.
The least disruptive scenario involved reallocating harvesting of broadleaved-dominated sites and spruce-dominated bogs to winter. This option increased harvesting and storage costs by only 2.7% compared to baseline scenario. This scenario would align with national recommendations from the forest industry and the Finnish FSC standard, as well as with the current legislative proposal in Finland.
It should be noted that not all the scenarios studied are feasible, or their implementation would require significant changes in wood procurement practices, business operations, and the availability and stability of the workforce. The most challenging factors mentioned are the limited availability of skilled labour for short-term seasonal work and the difficulty of running a profitable and sustainable harvesting business in such a seasonal sector. Additionally, a critical factor is the impact of increased logistics costs on stumpage prices and the availability of wood, as well as on the competitiveness of the forestry industry.
The positive effects of soil preparation on the survival and growth of planted seedlings have been widely known for a long time. However, recent studies have highlighted the importance of developing new technical solutions for soil preparation to mitigate its harmful environmental impacts. For example, particular interest has fallen upon methods that reduce the intensity of soil disturbance yet still provide favorable growing conditions for seedlings.
A novel and less intensive drill-based soil preparation device was recently developed in Finland. The drill significantly reduces soil disturbance compared to conventional methods, such as disc trenching and mounding. However, the performance of the drill and its effects on forest regeneration success have thus far remained unexamined.
In this study, we established five experimental sites in eastern Finland to evaluate the feasibility of drill-based soil preparation for planting conifer seedlings. The aim of the study was also to provide practical knowledge for the further development of the device. First, we assessed the quality of the planting spots (N = 4120) created by a crawler excavator fitted with the drill-based device. Secondly, we evaluated the early performance of Norway spruce (Picea abies (L.) Karst.) and Scots pine (Pinus sylvestris L.) seedlings that were planted throughout the growing season following soil preparation.
Based on our feasibility study results, the drill primarily resulted in good (61%) and satisfactory (34%) planting spot quality. We found that the presence of stumps, roots, logging residues, and stones decreased the spot quality. Two years after planting, Norway spruce seedlings were, on average, more viable than those of Scots pine. Site fertility and seedling vitality were significant predictors of Norway spruce height. However, further research is needed on the factors that influence seedling survival and growth when using a drill-based soil preparation method.
To mitigate wildfire risk in the western United States, forest fuel treatments such as mechanical thinning and prescribed burning have been widely implemented in fire-prone areas. While these treatments play a critical role in reducing wildfire hazards, they also present both opportunities and challenges. A substantial amount of biomass generated from fuel treatments offers new opportunities for bioenergy development. However, without effective downstream utilization, the resulting residues may lead to issues related to piling, transportation, and secondary combustion. Existing forest biomass supply chain studies have primarily focused on logging residues and sawmill residues, with limited attention paid to biomass derived from fuel treatments as a feedstock source. Moreover, spatially explicit assessments of biomass availability aligned with priority investment landscapes remain scarce. This study aims to develop a novel biomass supply chain framework that incorporates fuel treatment products into upstream biomass supply and to evaluate their supply potential and system costs under different utilization scenarios. The proposed framework integrates multiple methodological approaches, including forest management simulations based on the FVS-BioSum model, GIS-based spatial suitability analysis, and a comparative application of mathematical programming and machine learning techniques for supply chain optimization. The results are expected to provide quantitative decision support for the Wildfire Crisis Strategy and regional bioeconomy development. This study contributes to the coordinated advancement of sustainable forest management and the energy transition.
Growing interest in logging residues in Sweden, resulting from geopolitical developments and ambitions for energy independence, has intensified the need for precision planning of wood fuel supply. More than 15 years ago, Skogforsk developed a system that uses harvester production files to forecast volume and location of logging residues at stand level. Implemented through the calculation module “hprCM”, the system is used by several forest companies. Accurate forecasts require that harvester operators correctly distinguish fuel-adapted areas from non-adapted areas. In practice, harvester and forwarders frequently use logging residues to reinforce strip roads and wet areas to prevent soil damage, and such adjustments must be manually registered via the keyboard or touchscreen in the onboard computer.
As part of the Swedish Energy Agency-funded project “Improved forecasts of logging residues for precision planning”, we performed a large‑scale evaluation of the forecasting system. The analysis revealed substantial variability in how contractors used the system and how well the forecasted volumes matched the measured volumes at the heating plants. To better understand the underlying causes, semi-structured interviews were conducted with six experienced harvester operators. The interviews focused on planning, instructions, working methods, system use, and feedback routines connected to logging residues. The results identified several categories of usability issues related to system functionality, instructions, follow-up procedures, and attitudes toward logging residues.
Based on these findings, Skogforsk proposed a set of potential solutions and subsequently organized a workshop to anchor ideas and discuss the proposed measures. The workshop brought together representatives from forestry and energy companies, IT developers, and machine manufacturers, which together identified concrete, implementable solutions which aim to increase the accuracy of precision planning of wood fuel supply. A new follow‑up assessed the progress of implementation, highlighting pathways to improve short‑term forecast accuracy and long‑term processes for working with logging residues.
Information about the trafficability of gravel roads is of paramount importance for sustainable, cost-effective, and year-round delivery of timber for industrial use. For heavy vehicles like a timber truck, it is essential that the roads are in a trafficable condition when they are being driven on. The trafficability of forest roads varies depending on the season and weather conditions. Many roads become impassable during the spring thaw and following torrential rain. However, there are no models which could predict the bearing capacity of gravel roads in real-time and for the near future, i.e., the next few days.
The aim of this study was to develop and evaluate a simple dynamic bearing capacity model for gravel roads based on weather and soil type data, i.e., inputs which are readily widely available, and thus could be easily upscaled to encompass eventually a whole country. A further aim was to test the sufficiency of weather and soil type data and to evaluate the potential of calibration measurements of roads to improve the predictions.
The road data consisted of ten approximately 100-m long gravel road segments in eastern Finland. The road segments represented typical forest road conditions regarding subsoils and road structures. Data on air temperature and rainfall were recorded by nearby weather stations on hourly basis. Furthermore, bearing capacities of the road segments were sampled using a light-weight deflectometer at 20 m intervals during the 2.5-year data collection period.
The role of temperature, rainfall, soil type, and their interactions were investigated using a linear mixed-effects model. For temperature, we used a 72-hour moving average capped at the freezing point. For rainfall, we used a 48-hour moving average. The optimal number of calibration visits was estimated using a cross-validation approach in which the model was tested using one segment at a time and trained using the other nine segments.
Rainfall, temperature, and soil type alone explained about 19% of the total variance based on the linear mixed effects model. When road segment and specific location are also included, the full model can explain 48% of the variance in bearing capacity. In general, the predicted dynamical behaviour followed the trend of the bearing capacity measurements for all road segments. However, the predictions tend to have less variation than the ground measurements, especially the minima and maxima tend to be over- and underpredicted, respectively. Two calibration measurements improved the prediction accuracy by approximately 15%. After two field visits the benefit of additional visits was marginal.
The relatively simple prediction model can be further improved with additional explanatory variables, such as features extracted from LiDAR data. The calibration is independent from the used prediction model, and the field visit can be done any time. However, to get unbiased estimates when the bearing capacity is at its lowest, more measurements are needed in poor conditions.
One-third of forestland ownerships in Minnesota are comprised of privately-owned, family forests. The majority (65%) of those family forests are found on less than 8-hectare parcels and the owners of which commonly have alternative land stewardship objectives that differ from traditional forest management goals. Timber harvesting as a management objective is universally low in priority for small family forest owners, yet it is often still required to achieve many of the landowner’s other stewardship objectives. A logging business owner makes various decisions based on economic factors when considering a timber sale purchase; but they may also consider other factors particular to their business characteristics. All these considerations become increasingly important as the size of a timber sale decreases and may become barriers to implementing a timber harvest on small family forests.
Northern Minnesota logging business owners were surveyed to collect individual business-specific data regarding factors affecting their decisions to purchase small-sized timber sales, barriers to harvesting timber on small tracts, and the general characteristics of their business. Results suggest that some logging businesses may be more readily adaptable to harvesting on small family forests. A logging business whose annual production was <3,625 cubic meters and used a chainsaw as a primary felling method was the most likely to purchase a timber sale of 725 cubic meters and less. While they generally target purchasing higher volume timber sales, larger producing businesses annually complete more of these small sales than the smaller logging businesses. The most important factor for a larger logging business’ consideration to purchase a small timber sale is whether the sale is adjacent to another larger and available timber sale. If the family forest tract is not adjacent to another timber sale, the lowest volume producers appear to be the most likely to purchase the sale.
Norway has a forest road network of 50 000 km, where a single road typically has 4-5 owners. The current road network has been progressively extended since the 1960s. As a result of climate change and transition to heavier timber trucks, maintenance practices are becoming increasingly important.
Over the last decades there has been increased focus on cooperation between forest owners and road owner associations. This led to the development of road maintenance associations to coordinate tasks and maintenance between road associations across administrative borders. The activities for the maintenance associations drive task planning, cost distribution and accounting through a web-based platform (https://vvfweb.no/).
The VVM project (Road Maintenance module) aims to develop the next generation of web-based solutions for road- and road maintenance associations, to support a gradual upgrading of forest road standards. The current accounting platform has been in operation since 2008 and includes a historical database of maintenance activities and cost categories. The database spans 17 maintenance associations with 11 000 maintenance tasks across 458 roads and 1471 forest owners. The current management practices have been mapped at three stakeholder levels, and improvement potentials are identified.
One of the project goals is to ensure the dissemination of knowledge on effective summer maintenance and their costs. This presentation focuses therefore on the analysis of task-specific cost development from 2008 to 2024 from one region of southeastern Norway. The results support road owners in making realistic plans for maintenance tasks and budgets. It also documents the effects of cooperation through road maintenance associations road maintenance cost levels.
This study examines the barriers to entry for third-party log truck drivers in the southern United States and situates those barriers within the broader structure of the log trucking workforce. Using a mixed-methods design, a 23-item survey was administered in person at 23 fuel stations, yielding 573 completed responses from company, independent, and third-party drivers in Alabama, Louisiana, Mississippi, South Carolina, and Texas. The sample was predominantly male, with an average age of 48 years, and highly experienced, averaging roughly 15 years in log trucking and 18.6 years of total trucking experience. Quantitative data on demographics, experience, operational challenges, and perceived entry difficulty were analyzed using descriptive statistics, Kruskal–Wallis tests, chi-square tests, and logistic regression, while open-ended responses on barriers and suggested improvements were coded thematically in MAXQDA. Results show an aging and experienced workforce with relatively few young drivers and a modest share of independent operators compared to previous studies. Across multiple questions, three barriers consistently emerged as central to entry into log trucking: experience requirements, meeting insurance qualifications, and high initial investment costs. Insurance and experience were especially important for company and third-party drivers, while startup and repair costs were most critical for independent owner-operators. Although nearly half of respondents believed logging has more barriers to entry than other trucking sectors, about 80% indicated they did not want to leave log trucking and roughly seven in ten were likely to recommend the job, suggesting strong attachment to the sector despite persistent challenges. Only 9% of non–third-party drivers would consider working for a third-party contractor, and mainly only if pay and benefits were clearly better. The findings identify unattractive financial returns and stringent experience expectations as the biggest barriers specific to third-party log truck positions, indicating that improving recruitment and entry will require easing experience and insurance constraints while strengthening compensation and cost support for small operators.