TOWARD AN ADAPTIVE CONTROL MODEL FOR ROBOTIC BACKHOE EXCAVATION

Abstract

Research in robotic excavation has been focused mainly on path control to let a machine search and adapt trajectories automatically. Approaches to detect and handle underground obstacles such as rocks or utility lines are also critical for robotizing the operation. A newly established research facility to study robotic backhoe excavation using a real-size hydraulically powered and computer-controlled manipulator is introduced. A hierarchical planning and control model for robotic excavation is presented. The control system for this robotic backhoe is based on multiple sensors for force and position measurements and an approach for the detection and recognition of underground obstacles is discussed. Experimental data are used to analyze the force and acceleration patterns while the bucket hits an obstacle. Finally, a decision model for obstacle handling strategy derivation is introduced.