Automating A Mini Crawling Tractor: Designing An Autonomous Controller For The “Pfanzelt Moritz”
Remote controlled mini crawling tractors have become an important component of modern forestry operations, thanks to their efficiency, ease of use, and low environmental impact. These machines are often used in thinning, log skidding, or site preparation, while more recent innovations have demonstrated their utility in more complex tasks such as planting and felling. For small forest owners, mini tractors can play a crucial role, as the use of conventional large-scale forestry equipment is uneconomical, rendering small-scale forestry infeasible.
This presentation outlines our mechanical, electrical, and software design for an autonomous control system tailored for the Moritz mini crawling tractor. Our proposed system integrates a central GPU-powered real-time edge computer running ROS2, which collects data from an extensive sensor suite. A 3D LiDAR system will enable precise self-localization and mapping, allowing the machine to detect trees and obstacles in its environment and track its position and velocity with high accuracy. Ground-oriented stereo cameras mounted at the front of the system support local navigation and obstacle avoidance. Additionally, these cameras will enable future image-based AI algorithms to intelligently detect and classify objects in the machine’s environment, further improving its decision-making and autonomy. Additionally, a GNSS system will ensure precise positioning under GPS coverage, while IMU’s improve localization and monitor for potentially hazardous conditions such as tilting or sudden accelerations. Finally, a communication module will enable the device to be programmed and controlled remotely, while also relaying data back to a central system.
Our aim is to significantly reduce the need for human intervention in forestry operations, increasing efficiency and safety, while reducing costs. Future directions of this work will explore how to effectively utilize this platform’s sensing and computing capabilities for fully autonomous forestry operations, such as sowing, mulching, or soil preparation within designated areas.