NSF/NASA Workshop on Autonomous Mobile Manipulation (AMM)

Executive Summary Research in Autonomous Mobile Manipulation (AMM) aims to create robotic agents capable of performing physical work in unstructured and open environments. Such technology will impact a variety of application areas with significant economical, societal, and scientific importance. Among them are assistive and elderly care, planetary exploration, manufacturing, and supply chain management. Many countries have initiated well-funded and focused research programs in Autonomous Mobile Manipulation or closely related areas. In the United States, such a coordinated research initiative is still lacking. This workshop gathered some of the leading researchers in robotics, computer vision, and related fields, to devise a set of recommendations towards the initiation of such a program. These recommendations include: The specification of a general research program that builds on the existing strengths of research and technology in the United States and positions the academic community competitively , relative to comparable initiatives in other countries. The creation of center-level funding opportunities for multidisciplinary initiatives in AMM. This is necessary to address the broad range of research problems that arise in this area in an integrated manner. Furthermore, it is consistent with initiatives in other countries. A proposal for three concentrated research thrusts within the NSF Robotics and Computer Vision programs. These thrusts focus on specific research topics that are viewed as elementary building blocks for an initiative in AMM. They include a program for the development of basic skills (in particular dexterous manipulation skills), a program for the development of integrated, multi-modal sensor strategies, and a third program for the development of in-tegrative architectures to facilitate the robust and flexible operation of robotic agents in the context of AMM.