Methods for Optimizing the Building of Collaborative Robot Routes in a Dynamic Environment

Abstract

Modern collaborative robots (cobots), which work alongside humans in a shared workspace, must plan trajectories that simultaneously meet the requirements of safety, smooth motion, timely responsiveness to dynamic obstacles (people, other robots, moving objects), energy efficiency, and task accuracy. The dynamic environment imposes additional constraints on the planner: incomplete information about future obstacle trajectories, uncertainty in sensor measurements, rigid kinematic/dynamic constraints of the robot, and the need for fast real-time response. Because of this, trajectory optimization in a collaborative context usually requires a combination of: predictive models of human/obstacle behavior, rapid local corrections, and global optimality criteria — which is why the comparison of methods should focus not only on “trajectory quality” but also on computational costs, safety guarantees, and integration with human behavior prediction.