We present a hybrid differential dynamic programming algorithm for planning and closed-loop execution of manipulation primitives with frictional contact switches. Planning and control of these primitives is challenging as they are hybrid, under-actuated, and stochastic. We address this by planning a trajectory over a finite horizon, considering a small number of contact switches, and generating a stabilizing controller. We evaluate the performance and computational requirements of our framework in ablations studies for two primitives: planar pushing and pivoting. We can plan pose-to-pose trajectories from most configurations with only a couple (one to two) hybrid switches and in reasonable time (one to five seconds). We further demonstrate that our controller stabilizes these hybrid trajectories on a real pushing system.
For more information: http://mcube.mit.edu/pdfs/2020_ICRA_H...