Simultaneous dual-arm motion planning for minimizing operation time

Abstract

Dual-arm robots are expected to perform work in a dynamic environment. One of the most basic tasks that a dual-arm robot does is pick-and-place work. However, this work is more complicated when there are several objects in the robot's workspace. Additionally, it is likely to take a long time to finish the work as the number of objects increases. Therefore, we propose a method using a combination of two approaches to achieve efficient pick-and-place performance by a dual-arm robot to minimize its operation time. First, we use mixed integer linear programming (MILP) for the pick-and-place work to determine which arm should move an object and in which order these objects should be moved while considering the dual-arm robot's operation range. Second, we plan the path using the rapidly exploring random tree so that the arms do not collide, enabling the robot to perform efficient pick-and-place work based on the MILP planning solution. The effectiveness of the proposed method is confirmed by simulations and experiments using an actual dual-arm robot.