Roderic A. Grupen

Professor, Computer Science
Director, Laboratory for Perceptual Robotics
Autonomous sensorimotor systems, embedded control, developmental dynamics, multifingered hands, walking machines, animate vision

Background: Ph.D. Computer Science, University of Utah (1988); M.S. Mechanical Engineering, Pennsylvania State University (1984); B.A. Physics, Franklin and Marshall College (1980); B.S. Mechanical Engineering, Washington University (1980). Professor Grupen joined UMass Amherst Department of Computer Science faculty in 1988.

Research: Professor Grupen conducts research which integrates computing, operations research, and control theory as a means of modeling intelligent systems -- natural and artificial. A central focus of this work is computationally tractable controllers for complex, highly redundant systems. Grupen and his students have contributed techniques for motion control and collision avoidance, methods for coordinating multiple robots, methods for dextrous multifingered grasps, and techniques for formulating stable, aperiodic walking gaits. Most recently, this paradigm has been extended to express visuomotor control tasks, to scheduling flexible manufacturing systems, and to control for intelligent vehicles and highway systems. With Professor Andy Barto, Grupen is studying a means of learning sensorimotor behavior. The approach relies on structuring the search for behavior by expressing the intrinsic dynamics of physical processes and acquiring policies for robot control tasks using on-line reinforcement learning. In conjunction with colleagues in Computer Science and Developmental Psychology, we have used this framework as the basis of a computational model of sensorimotor development in humans and machines. Work is underway which seeks to explain the dynamics of early childhood development in terms of a stochastic search and subsequent policy formation in a well-defined space of reflexive, sensorimotor transformations. This study includes an empirical study on the implications of this model in the acquisition of spatial representations from vision, audition, and prioprioception in robots and human neonates.

Activities and awards: Professor Grupen is on the Editorial Board of the Robotics and Autonomous Systems Journal and Artificial Intelligence for Engineering Design, Analysis and Manufacturing (AI EDAM), served is a member of the editorial board for the Autonomous Robots Journal Special Issue on Learning in Autonomous Robots (ARJ-LAR), is a member of the Executive Committee for the IEEE Technical Committee on Assembly and Task Planning (TC-ATP), and was on the program committee for Autonomous Agents '98, and the 1998 IEEE International Conference on Robotics and Automation (ICRA'98). He received the UMass Amherst NSM Outstanding Teaching Award in 1999.

Personal Home Page | CS Faculty Directory | CS Home Page