Assessing The Efficiency of Mechanized Hardwood Thinning Operations Using Machine Time Study Data and Lidar/ai Technologies
Mechanical thinning of stands to yield high-value logs represents a significant investment, requiring careful planning to manage costs effectively while achieving desired silvicultural objectives. In Australia, an increasing number of forest companies are transitioning from pulp to solid wood plantation regimes, highlighting the importance of understanding the financial implications of thinning operations and identifying the key factors influencing the efficiency of mechanized equipment, particularly in hardwood stands. This study utilised machine productivity data to assess mechanised thinning operations in Eucalyptus nitens stands, pinpointing the key factors affecting productivity and costs. Additionally, sensing technology, including mobile LiDAR systems, has been employed to evaluate the quality features of the remaining trees, such as Diameter at Breast Height (DBH), height, branchiness, tree form, and stem damage.
The objectives of this study were as follows:
1. Quantify the productivity and costs associated with mechanized thinning operations in Eucalyptus nitens stands, identifying the key factors that significantly impact productivity and costs.
2. Develop a methodology to capture tree quality features in thinned stands using sensor data.
3. Provide data and methods aimed at enhancing the efficiency of hardwood plantation thinning operations.
4. Test optimization and artificial intelligence (AI) algorithms to improve the effectiveness and efficiency of tree selection decisions during thinning operations.
The research trial was conducted in Northwest Tasmania, where a single thinning regime and intensity were implemented at age 9, with the remaining trees allowed to grow until approximately 25 years of age. Mechanized thinning operations were carried out using a wheeled harvester/processor.
The results of the machine study revealed that over 85% of the variability in thinning productivity can be attributed to stem volume. Additional variability was explained by the location of the removed trees during thinning (outrows or outer rows). In contrast to the plot with marked trees, the Diameter at Breast Height (DBH) and stem volume of trees selected by the machine operator were consistently higher, irrespective of the location of the removed trees (whether in adjacent or outer rows), except for trees removed from outrows.
This study confirms the invaluable role of LiDAR technology in informed decision-making during thinning operations. LiDAR point cloud data collected before thinning empower harvest planners to make well-informed decisions regarding the optimal location of thinning roads, tree selection during thinning, and alternative thinning strategies before field implementation. Real-time or pre-thinning LiDAR data also aids harvester operators in making optimal tree selection decisions, incorporating variables like tree density for removal decisions and other factors such as basal area, tree form, and branchiness.