DEXMAN— Improving robot’s DEXterous MANipulability by learning stiffness-based human motor skills and visuo-tactile exploration

DEXMAN is an international collaborative research project hold by three partners: Universität Hamburg, Universität Bielefeld, and South China University of Technology, together founded by DFG (Germany) and NSFC (China). The goal of DEXMAN is to develop learning and control approaches based on multimodal data to improve the ability of robot dexterous and compliant manipulation.

More specifically, we will focus on the following three key technologies to achieve robust dexterous grasping and in-hand manipulation on robotic arm/hand system:

• A framework of augmented motion primitives embedding perception in formation for human skill extraction and generalization to new tasks;
• Reconstructing and tracking an unknown object by exploiting interactive manipulation and multi-modal feedback; and
• Multiple sensor fusion based adaptive grasping and manipulation control framework enhanced by human motor skill learning.

Project overview at DFG (GEPRIS): click here