Development of a generic decomposition framework for affordance-based grasping and design optimization

Project Leaders: Robert Haschke, Stefan Menzel, Heiko Wersing
PhD Student: Daniel Dornbusch

The relation of local and global structural properties is a key issue in complex problem domains like visual perception and motor planning. Methods that are capable of identifying decompositions of a large problem into more manageable local elements are an important approach in such domains to achieve structural robustness and generalization.

The target of this project is the development of a generic decomposition framework for affordance-based grasping using anthropomorphic robot hands and shape design optimization.

Unlike existing grasp selection systems, a solution is aimed for, which does not depend on an a-priori known 3D shape of the object. Instead it uses a decomposition of the object view into local, grasping-relevant shape primitives, whose optimal grasp type and approach direction are known or learned beforehand. Based on this decomposition a list of possible grasps can be generated and ordered correspondingly to the anticipated overall grasp quality.

In the field of design optimization and design conception it is also very advantageous to extract local knowledge on shape primitives in the sense of how they influence the overall design quality. On this basis, new design conceptions can be generated by combining different desired local performance contributions and selecting the associated shape primitives accordingly.

Both application areas share the idea to associate local shape features (i.e. object parts) with functional data (e.g. position of fingertips / hand posture, relative distance object-hand, or aerodynamical flow). In this way an optimization process (e.g. grasp selection), which incorporates previously acquired experience, can be boosted with respect to an overall cost function.