Robot Hands and Sensing

Abstract

It is not an overstatement that the history of robot hands is equal to that of 40-year-old manipulators. Initial hands were based on the simple ON/OFF switching system, with pneumatic or hydraulic power used as a source. The hands could realize their primary purpose of firmly grasping an object, however, it was impossible to achieve the dexterous manipulating motion which was another important function. Since the latter half of 1970's, the dexterous functions have been in demand for robot hands such as the functions of an inspection robot for the power plant. This caused research institutes in many countries to start development projects on multi-fingered hands. The Okada-Hand, Salisbury-Hand, and Utah/MIT-Hand are particularly well-known among multi-fingered hands developed through such projects. In parallel with this research and development, theoretical research activities progressed for stable grasping, fingertip force analysis, and grasping force control. Theoretical studies of hands reached a peak in both quality and quantity in the latter of 1980's. However, the experimental studies using actual multifingered hands were far behind the theoretical studies. Based on the reconsideration of the importance of experimental validation, experimental works have been done to verify the theory of stable grasping or manipulation with actual hands. From another perspective, researches have recently been started in an attempt to use fingers not only as actuator for grasping, but also as an active sensor for recognizing the external world with tactile motion. This causes the research field of hand to spread. Based on the current status of hands researches, this special issue will compile conventional works and provide an outlook for future hands. The editor of this issue will be very pleased if this material can provide with any useful information to hand researchers. Finally, the editor wishes to express his sincere thanks to the contributors.