"This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5)."
Robot hands are usually simple, with just two or three fingers, perhaps a single actuator, and most often no sensors at all. These simple hands are also very specific in their function, such as picking up a specific part. Research on more general robot hands usually focuses on complex hands, often resembling human hands. This project is developing simple hands with general capabilities. The approach is inspired by a variety of simple hands, such as a prosthetic hook, which have proven generality when controlled by a human, yet have never demonstrated great generality when controlled by an autonomous robot. In particular the project is developing hands that can blindly capture objects among clutter, and testing these hands both in a factory automation application and in a home assistive robotics application.
Results from this study will be broadly applicable. Every advance in hand design enables new applications, so development of new principles broadly advancing the generality of hands will be useful. Specific cases are advancing the nation's manufacturing workforce productivity, and enabling the elderly to live independently. Results will be disseminated by scholarly publication of new principles, analysis, and experimental results, as well as distribution of analytical software, planning and control software, and hand designs.
