Automating short-term insertion of parts for heterogeneous robots using a control basis approach

Abstract

Robotics technology is quickly evolving and demanding robots to perform more actions and with greater complexity. Modular construction tasks are well suited for heterogenous robots. In this paper we studied if modularizing a distributed multi-agent architecture with a control modular framework was a viable approach to generate complex robotic behavior with short term autonomy. To this end we tasked a heterogeneous robot team to perform an assembly task using force sensing. We also studied which robots would be better suited for a given task. Our experimental results concluded that the modular control approach was viable to generate short- term autonomous complex robotic behavior and our analysis characterized the advantages of each robot for the assembly task.