Optimization-based Task and Trajectory Planning for Robot Manipulators

We introduce optimization-based algorithms that address the problem of robot task scheduling and trajectory planning. Following a two-layer hierarchical control structure, we first decouple the task scheduling from the trajectory planning by introducing two separate optimization problems, a discrete one for task scheduling and a continuous optimization problem for trajectory planning. In a further step, we integrate both planning layers into a monolithic layer in the form of a mixed-integer recursive optimization problem. The algorithms are implemented and validated using Robot Operating System (ROS) on an experimental setup with a robot manipulator performing pick-and-place tasks.