Dual-Sensor Based Robotic Deburring

Abstract

New results quantifying the performance of a vision and force sensor based robotic deburring system at high feedrates and on stainless steel parts are presented. The performance with stainless steel is similar to that achieved with mild steel due to adaptation of the process model. Increasing the feedrate from 25 mm/s to 50 mm/s did not greatly affect deburring performance and could be used to reduce cycle time. The dual-sensor approach is shown to be 14 times more reliable than a single sensor solution. The overall accuracy of the system is ±30 μm for straight edged parts, and ±60 μm for 2-d edged parts with geometry similar to those tested.