A Virtual Mechanism Approach for Exploiting Functional Redundancy in Finishing Operations

We propose a new approach to programming by the demonstration of finishing operations. Such operations can be carried out industrial robots in multiple ways because an robot is typically functionally redundant with respect task. In proposed system, human expert demonstrates operation, and demonstrated motion recorded Cartesian space. The robot’s kinematic model augmented virtual mechanism, which defined according applied tool. This way, expanded additional degrees freedom that exploited compute optimal joint space without altering essential aspects task execution as expert. Finishing operations, such polishing grinding, occur contact treated workpiece. Since information about point position needed control during we have developed novel for accurate estimation points using measured forces torques. Finally, iterative learning refine compensate inaccurate calibration different dynamics demonstrator. method was verified on real grinding tasks. Note Practitioners —This work motivated need automation contemporary robots. Existing approaches are both too complex time-consuming flexible small-scale production, often requires frequent deployment applications. Our based enables also users who not specialists programming. Programming especially useful teaching it transfer knowledge skills providing lengthy descriptions or manual coding. Besides desired availability machine tool carry operation. provide several practical examples tools how integrate them into our approach. Another feature system user demonstrations transferred between combinations tools.