SAHA: Supervised Autonomous HArvester for selective forest thinning

TL;DR

SAHA demonstrates the feasibility of autonomous selective forest thinning using a compact 4.5-ton harvester equipped with a perception payload, learning-based traversability, and a motion-primitive local planner. The system integrates LiDAR-Inertial Odometry, probabilistic traversability mapping, and an adaptive hydraulic arm controller to achieve autonomous navigation and targeted tree grasping in real forests, validated through extensive field trials in northern Europe. Key contributions include hardware-tuned perception and control for under-canopy navigation, a MinkUNet-based traversability classifier trained on DigiForests, and a robust, locally-planned autonomy stack that operates with supervisor oversight. The work provides a practical blueprint for scalable, safer autonomous forest thinning and highlights directions for multimodal sensing, global path preferences, and close-range perception to enhance reliability and deployment in diverse forest environments.

Abstract

Forestry plays a vital role in our society, creating significant ecological, economic, and recreational value. Efficient forest management involves labor-intensive and complex operations. One essential task for maintaining forest health and productivity is selective thinning, which requires skilled operators to remove specific trees to create optimal growing conditions for the remaining ones. In this work, we present a solution based on a small-scale robotic harvester (SAHA) designed for executing this task with supervised autonomy. We build on a 4.5-ton harvester platform and implement key hardware modifications for perception and automatic control. We implement learning- and model-based approaches for precise control of hydraulic actuators, accurate navigation through cluttered environments, robust state estimation, and reliable semantic estimation of terrain traversability. Integrating state-of-the-art techniques in perception, planning, and control, our robotic harvester can autonomously navigate forest environments and reach targeted trees for selective thinning. We present experimental results from extensive field trials over kilometer-long autonomous missions in northern European forests, demonstrating the harvester's ability to operate in real forests. We analyze the performance and provide the lessons learned for advancing robotic forest management.

Figures (27)

• Figure 1: The developed autonomous harvester system has been deployed in real-world experiments across various terrains and seasons. It can navigate along service trails (top left) and traverse cluttered forests (top right) during different seasons (bottom left), and reach and grasp selected trees for thinning (bottom right).
• Figure 2: Overview of the SAHA system. It localizes itself initially in a prior map and uses onboard odometry for position tracking. A traversability estimation module identifies navigable terrain, guiding path planning by the motion primitive-based planner responsible for collision avoidance during navigation. Driving, arm control, and chassis balancing controllers manage autonomous movement and stability. A state machine coordinates transitions between navigation and cutting operations.
• Figure 3: The depicted SAHA robot is a versatile and autonomous forest logging machine. Its lightweight platform compared to other harvesting machines simplifies forest logistics and helps reduce environmental impact through decreased soil compaction.
• Figure 4: Multiple proprioceptive sensors are installed on the SAHA arm. The red boxes mark the position of IMU sensors, and the cyan box marks the wire draw encoder.
• Figure 5: The integrated control module mounted on the left front support cylinder of SAHA.