Manipulation Planning with Probabilistic Roadmaps

Abstract

This paper deals with motion planning for robots manipulating movable objects among obstacles. We propose a general manipulation planning approach capable of addressing continuous sets for modeling both the possible grasps and the stable placements of the movable object, rather than discrete sets generally assumed by the previous approaches. The proposed algorithm relies on a topological property that characterizes the existence of solutions in the subspace of configurations where the robot grasps the object placed at a stable position. It allows us to devise a manipulation planner that captures in a probabilistic roadmap the connectivity of sub-dimensional manifolds of the composite configuration space. Experiments conducted with the planner in simulated environments demonstrate its efficacy to solve complex manipulation problems.

A general technique for fast comprehensive multi-root planning on graphs by coloring vertices and deferring edges

“…[1]) may appear better suited for finding multiple diverse connections. For example, they have been used for efficiently solving the single-object manipulation problem [10].…”

A hierarchical approach to manipulation with diverse actions

“…The re-grasping problem [14], [15] requires multiple manipulations for a single object. The framework proposed by Siméon et al (2004) breaks the problem into finding transits for the robot moving alone and rigid-transfers for the robot and object.…”

“…The framework proposed by Siméon et al (2004) breaks the problem into finding transits for the robot moving alone and rigid-transfers for the robot and object. They note that any switch between transit and rigid-transfer must occur in the part of the configuration space where the object is sitting stably and the robot is grasping the object, which we will refer to as P G. They prove the "reduction property" that any two points in the same connected component of P G can be connected by a finite number of transits and rigid-transfers.…”

Manipulation with Multiple Action Types

“…The re-grasping problem [12,16], especially as framed by Siméon et al, is an example of planning with two manipulation actions. Siméon et al [16] take a hierarchical approach to the problem, first finding a high-level sequence of transits (motions for the robot alone) and rigid-transfers (motions in which the robot rigidly grasps an object) and then planning each in the robot's configuration space.…”

“…Siméon et al [16] take a hierarchical approach to the problem, first finding a high-level sequence of transits (motions for the robot alone) and rigid-transfers (motions in which the robot rigidly grasps an object) and then planning each in the robot's configuration space. Unfortunately, their method relies on the grasped object being able to move instantaneously in any direction, which does not hold for non-prehensile manipulation.…”