Using Bucket Force/Torque Feedback For Control Of An Automated Excavator

This paper presents design and implementation details of a fuzzy-behavior-based controller for an automated excavator. The robotic excavator is equipped with sensors, actuators and an embedded control system. The excavator controller uses neural-networks to interpret forces and torques gathered during the excavation process from bucket-mounted force sensors. Implementation details card experimental results using the fuzzy-behavior-based excavation controller are given for several excavation tasks. These excavation tasks demonstrate some of the complex skills that are essential for an autonomous mining-excavator controller.