Multi-object Grasping in the Plane

We consider a novel problem where multiple rigid convex polygonal objects rest in randomly placed positions and orientations on planar surface visible from an overhead camera. The objective is to efficiently grasp transport all into bin using multi-object push-grasps, are pushed together facilitate grasping. provide necessary conditions for frictionless push-grasps apply these filter inadmissible grasps planner. find that our planner 19 times faster than Mujoco simulator baseline. also propose picking algorithm uses both single- pick objects. In physical grasping experiments comparing performance with single-object baseline, we the system achieves 13.6% higher success 59.9% faster, 212 PPH 340 PPH. See https://sites.google.com/view/multi-object-grasping videos code.