Behavior-based control of multi-robot assembly/palletizing systems

A decentralized palletizing multi-robot system was developed using fuzzy behavior-based control. The system consists of a conveyor, two robots and a vision system. The system´s role is to perform a palletizing task using parts arriving on the conveyor. Two control algorithms were compared: a simple algorithm and a behavior-based algorithm combined with fuzzy logic. Thirty experiments were conducted while changing the control algorithm, parts arrival frequency, and conveyor velocity. Performance measures included number of parts taken. number of pallets, robots utilization and work cell utilization. Experimental results indicated better performance for the simple algorithm. However. the fuzzy behavior-based algorithm equalizes the robots utilization. Equal utilization is a major factor in multi-robots systems, and contributes in decreasing mean time between failure (MTBF). We expect that for a heavier loaded system the fuzzy behavior-based algorithm will yield higher performances.