Multi-Robot Path Planning with Collision Avoidance

Integrating robotics into manufacturing tasks is now a decades-old practice with ongoing advances making robotics faster, cheaper, and more accurate on an almost daily basis. Individually, robots have become highly effective at performing tasks in isolation. The goal of this research is to advance automatic path planning for multiple robot agents to intelligently and cooperatively accomplish manufacturing tasks in close proximity. These robots can range from simple robot architectures to several 6-DOF articulated arms on mobile bases to be used for spray coating, pressure washing, media blasting, and sanding. These applications are low-volume, high-mix manufacturing environments where task variability renders human programming impractical. While addressing the possibility of collisions for multi-agents, practical manufacturing constraints also need to be considered. In additive manufacturing, for example, it is important that each raster be completed by a single agent to prevent undesirable tool retracts that will affect print quality. The methodology for addressing this research includes the development of optimization models that simultaneously incorporate both manufacturing process constraints and the manipulator's kinematics and collision constraints. This project proposes to develop coordination planning techniques for N overlapping robot architectures composed of 3 or more revolute and/or prismatic joints.