Improving Robot Grasping Plans with Affordance

The quality of grasping is supported by leveraging the concept of affordance in this paper to improve grasping operations. When introducing flexibility in handling different objects in a dynamically changing environment, uncertainty in grasping areas and operations may cause problems. We have studied the relationship between the objects, their associated operations, and the effect of performing such operations. With such information, the robot arm may be able to limit some improper grasping operations and angles in order to introduce better plans more efficiently. We use OpenRAVE as the simulation environment to carry out the plans generated. The experiments show that better plans were generated with affordance.