Robotic Manipulation and Motion Planning

Every day robotic devices move, transform, or otherwise manipulate objects for many different purposes and using a high variety of approaches.

Manipulation can take many forms. Humans and robots can manipulate objects by carrying, holding, squeezing, pulling, pushing, throwing, and dropping them, and they can do so by using a single arm and hand or both arms and hands; they can even take advantage of the environment while performing their actions. It is common to distinguish between prehensile and non-prehensile manipulation. We consider both types of manipulation on various kinds of objects and take into account possible shape variations of the manipulated objects.

In prehensile manipulation the aim is to hold, or grasp, an object and thus constrain its motion. We study the analysis, existence, and synthesis of immobilizing grasps, which completely restrain an object, and caging grasps, which only prevent an object from escaping the fingers.

In non-prehensile manipulation the goal is to set an object in a desired motion by applying a resultant force. We focus on the role of actions such as pushing, squeezing, and dropping in the design of effective solutions to common high-level industrial tasks such as part orienting, or feeding,  and sorting.

We also study the motion planning problem. Motion planning aims to determine a collision-free motion for a moving entity in an environment with obstacles from an initial placement to a goal placement. Motion planning is important to manipulation but also plays a fundamental role in many other application areas, ranging from crowd simulation to drug design. 

Highlighted Papers 
Fatemeh Panahi, Mansoor Davoodi, A. Frank van der Stappen, Orienting parts with shape variation, IEEE Transactions on Automation Science and Engineering, to appear (2016)
Elon Rimon, A. Frank van der Stappen, Immobilizing 2d serial chains in form closure grasps, IEEE Transactions on Robotics 28(1), pp. 32-43 (2012)
Jae-Sook Cheong, Heinrich Kruger, A. Frank van der Stappen, Output-sensitive computation of force-closure grasps of a semi-algebraic object, IEEE Transactions on Automation Science and Engineering 8(3), pp. 495-505 (2011)
Mostafa Vahedi, A. Frank van der Stappen, Caging polygons with two and three fingers, International Journal of Robotics Research 27(11/12), pp. 1308-1324 (2008)

Our group has a long-standing collaboration with the group of Professor Ken Goldberg at the University of California, Berkeley.