Dynamic Collision and Deadlock Avoidance for Multiple Robotic Manipulators

A flexible operation of multiple robotic manipulators operating in a dynamic environment requires online trajectory planning to ensure collision-free trajectories. In this work, we propose real-time capable motion control algorithm, based on nonlinear model predictive control, which accounts for static and obstacles. The proposed algorithm is realized distributed scheme, where each robot optimizes its own with respect the related objective constraints. We novel approach collision avoidance between manipulators, predicted movement neighboring robots. Additionally, method reliably detect resolve deadlocks occurring setup manipulators. validate our pick place scenarios involving common workspace realistic simulation set up Gazebo. robots are controlled using Robot Operating System. Our scales 4 computes path simultaneous 94% all investigated cases without deadlock detection 100 % resolution algorithm. contrast, conventional planners, such as PRM, PRM*, CHOMP RRT-Connect, successfully plan at most 54% hindering their application setups