The Birmingham Hand
Academic rationale for hand research at Birmingham
1) The hand is scientifically and culturally iconic. The human hand is intimately intertwined with human evolution and cognition. A robotic hand and associated algorithms to control grasping and manipulation, that can achieve the dexterity of a human, is an important, but still far from attained, goal of automation engineering. Such a hand would have numerous applications, represent a pinnacle in mechatronic technology and AI, and form a major milestone in the development of real-world intelligent machines.
2) There are currently no satisfactory humanoid or non-humanoid dexterous multi-finger robot hands available, for research or industrial applications. Existing hands are:
i) Prohibitively expensive, typically £50,000 - £100,000 per hand.
ii) Delicate and fragile – coming nowhere near the robustness of industrial grippers.
iii) Weak – payloads of a few hundred grams, in contrast to human hands which can lift order 100kg.
iv) Sorely lacking in sensory capabilities, especially tactile sensing which plays a critical role in human grasping.
v) Lacking satisfactory control algorithms for autonomous grasping and manipulation under uncertainty.
vi) Lacking suitable input and sensory feedback devices (and knowledge of best practice and human-factors issues in using them) for human controlled tele-operative grasping.
vii) Insufficiently informed by an understanding of human and animal cognitive processes for grasping and manipulation.
3) Research in these areas is highly interdisciplinary, and would bring together a wide range of expertise from many schools across the University of Birmingham campus, as well as building on major university investments in cross-school centres and initiatives:
i) Mech. Eng. School – outstanding manufacturing facilities, and expertise in MEMS sensors (inc. tactile).
ii) CS school - autonomous grasping and manipulation algorithms (three major FP7 grants).
iii) EECE - expertise in bilateral control for human-robot interaction in tele-operated grasping (recent defence funding).
iv) Pyschology – leading centre of expertise in human grasping and manipulation.
v) Chem. Eng. – expertise in the tribology of fingertips and their interaction with objects and materials.
vii) New centres including CNCR (centre for Computational Neuroscience and Cognitive Robotics) and new HCI (centre for Human-Computer Interaction).
- Aaron Dollar, Yale, USA – novel hand mechanisms.
- Marco Gabiccini, Antonio Bicchi, U of Pisa, Italy – haptics and dexterous robotic manipulation.
- Vincent Hayward, Université Pierre et Marie Curie, France – haptics and tactile sensing devices.
- Peter Kyberd, Inst. Biomedical Engineering, U of New Brunswick, Canada – prosthetic grasping.
Guests from industry and non –university labs:
- Christoph Borst, DLR German Aerospace Research Centre – robotic hands and robotic manipulation.
- John Heindel, Schunk Intec – robotic hands.
- Richard Greenhill /Armando De La Rosa, Shadow Robot Company – robotic hands.
University of Birmingham delegates:
- School of Computer Science – Intelligent Robotics Lab: Aaron Sloman,Ales Leonardis, Marek Kopicki, Richard Dearden, Jeremy Wyatt, Sebastian Zurek, Chris Burbridge, Claudio Zito, Rustam Stolkin.
- School of Psychology: Alan Wing, Dietmar Heinke, Satoshi Endo, Hoi Fei Kwok, Max DiLuca
- School of Mechanical Engineering: Duc Pham, Mozafar Saadat, Mike Ward, David Cheneler
- School of Electrical Electronic and Computer Engineering: Bob Stone
- School of Chemical Engineering: Mike Adams
- School of Biosciences: Jackie Chappell
- Institute of Advanced Studies: Malcolm Press, Sue Gilligan