Summary This paper describes a system for the automated planning, control and inspection of robotic edge deburring. The task planner uses an object-oriented distributed artificial intelligence approach along with computer-aided design models to plan a collision-free path. The path is adjusted online to maintain the desired chamfer depth of cut using a high-bandwidth active end effector. The control system incorporates: a) sensor fusion of force and vision data, b) parameter adaptive predictive control, and c) learning control. Task completion is verified through in-process inspection of the chamfered edges. Simulation as well as experimental results are presented.