INTERNATIONAL JOURNAL OF PRECISION ENGINEERING AND MANUFACTURING Vol. 12, No. 6, pp. 1095-1103 DECEMBER 2011 / 1095 DOI: 10.1007/s12541-011-0146-0 1. Introduction In recent decades, many myoelectric hands have been developed to imitate the grasping capabilities of the human hand. The first-to-market myoelectric hands, such as the MyoBock hand 1 and ProControl hand, 2 have a good reliability and robustness. However, their mechanism of two rigid fingers and a rigid thumb linked in opposition by a lever and simultaneously actuated by a single motor, is unable to adapt to different object sizes and shapes. More recently, a new myoelectric hand, the i-LIMB hand, 3 became commercially available. This hand allows the four fingers and thumb to stop independently when they come into contact with an object. Yet, the fingers and thumb only provide a fixed curling trajectory and the abduction and adduction of the thumb must be performed manually by the user. In the field of myoelectric hand design, weight, size, power consumption, and cosmetic appearance are all important factors. However, to improve the grasping capability, a ‘fully-actuated’ approach requires a large number of actuators for high degrees of freedom. This conflicts with other priorities of myoelectric hand design. Thus, to overcome these limitations, a new design approach, called ‘underactuation’ has been introduced that can maintain a small number of actuators, while increasing the degrees of freedom. This facilitates ‘adaptive grasping’, where the fingers and thumb adapt to the shape of a grasped object to increase the contact points between the hand and the object. 4 As a result, underactuated hands can be designed to be light-weight and small in size, while retaining the grasping capabilities of a fully-actuated hand. Until now, a variety of underactuated mechanisms have been proposed, including a compression spring, 4-6 pulley, 6,7 whiffle-tree, 8 adaptive linkage, 9 tendon-routing mechanism, 7,10 and many others. 14 In particular, the spring mechanism has several advantages over other mechanisms. They are simple and small, and can facilitate adaptive grasping, including shape adaptation between fingers or phalanges. For example, Dechev et al. 4 developed an underactuated hand with a compression spring mechanism that allows the four fingers and thumb to flex inward independently to conform to the shape of an object. Meanwhile, Carrozza et al. 5 proposed a shape adaptation mechanism among fingers and phalanges, where each finger includes cables and compression springs, and each cable is fixed to a corresponding phalanx through each spring. When one link contacts an object, the compression of the corresponding spring allows the other links to continue bending. A linear slider with three pulleys drives all the cables for the fingers, and three pairs of cables Myoelectric Hand Prosthesis with Novel Adaptive Grasping and Self-locking Jun-Uk Chu 1 , Dong-Hyun Jeong 2 , Inchan Y oun 1, *, Kuiwon Choi 1,#, * and Yun-Jung Lee 3 1 Biomedical Research Institute, Korea Institute of Science and Technology, 39-1, Hawolgok-dong, Seongbuk-gu, Seoul, Korea, 136-791 2 Robotics R&D Group, Daewoo Shipbuilding & Marine Engineering Co. Ltd., 7-1, Dangsan-dong 1-ga, Yeongdeungpo-gu, Seoul, Korea, 150-041 3 School of Electronics Engineering, Kyungpook National University, 1370, Sankyuk-dong, Book-gu, Daegu, Korea, 702-701 # Corresponding Author / E-mail: choi@kist.re.kr, TEL: +82-2-958-5921, FAX: +82-2-958-5909 * These authors contributed equally to this paper as corresponding author KEYWORDS: Hand prosthesis, Underactuation, Adaptive grasping, Self-locking, EMG pattern recognition This paper presents a novel hand prosthesis controlled by electromyography (EMG) signals. Using an underactuated approach, the hand allows four fingers to flex independently with one actuator, plus the fingers and phalanges provide adaptation with respect to the shape of an object. An innovative self-lock is also embedded in the metacarpophalangeal joint to prevent back-driving when external forces act on the fingers. The thumb is designed to perform flexion/extension and abduction/adduction with one actuator, while the wrist is driven in a differential manner using two actuators. As a result, the hand has eighteen degrees of freedom with only four actuators. Furthermore, it is controlled in real-time by the previously proposed EMG pattern recognition method. 13 Ten kinds of hand motions are classified from four channel EMG signals with a high accuracy and corresponding motion commands are generated for hand control. Experimental results demonstrate the effectiveness of the hand design and validity of the EMG-based hand control. Manuscript received: April 1, 2011 / Accepted: May 5, 2011 © KSPE and Springer 2011