Grant: Discovery Grants Program - Individual

Dexterous object exploration, grasping and manipulation is a challenging area of robotics where multiple fingers cooperate to carry out the task. Controlling a human-like dexterous robotic hand poses major design challenges for the tactile sensors that are the first line of contact, for processing the multisensory data, detecting object geometry, slippage and contact pressure, and then for the tactile and kinaesthetic feedback-based coordinated control of the finger movements and applied forces.    The long-term objective of the proposed research is to develop a new generation of biomimetic tactile sensor-based intelligent control algorithms for human-like multi-finger robot hands able to dexterously explore, grasp, and in-hand manipulate 3D objects.    The short-term objectives are (1) development of novel biomimetic tactile sensors able to provide coherent multisensory feedback for the real-time control of dexterous robot hands, (2) development of new neuro-fuzzy reactive behavior control algorithms for concurrent position and force control of hand fingers during object exploration, and (3) development of more efficient task-driven intelligent control algorithms for dexterous object grasping and in-hand manipulation.     The novel intelligent tactile based hand control techniques to be developed will have a significant impact on an emerging range of robotics applications requiring advanced dexterous object exploration, grasping, and manipulation capabilities such as: (i) autonomous or remotely operated robotic handling of materials in hazardous environments, high risk zones and security operations, or difficult to reach undersea and outer space environments, (ii) tele-medicine, (iii) a new generation of humanoid robots for elder care assistance and human-robot interaction.