Sensor-based roadmaps for motion planning for articulated robots inunknown environments: some experiments with an eye-in-hand system

摘要

We present a real implemented “eye-in-hand” test-bedsystem for sensor-based collision-free motion planning for articulatedrobot arms. The system consists of a PUMA 560 with a triangulation basedarea-scan laser range finder (the eye) mounted on its wrist. Theframework for our planning approach was presented in Yu and Gupta(1998). It is inspired by motion planning research and incrementallybuilds a roadmap that represents the connectivity of the freeconfiguration space, as it senses the physical environment. We presentsome experimental results with our sensor-based planner running on thisreal test-bed. The robot is started in completely unknown and clutteredenvironments. Typically, the planner is able to reach (planning as itsenses) the goal configuration in about 7-25 scans (depending on thescene complexity), while avoiding collisions with the obstaclesthroughout