MIT research in telerobotics

Ongoing MIT research in telerobotics (vehicles capable of some autonomous sensing and manipulating, having some remote supervisory control by people) and teleoperation (vehicles for sensing and manipulating which are fully controlled remotely by people) is discussed. The current efforts mix human and artificial intelligence/control. The idea of adjustable impedance at either end of pure master-slave teleoperation, and simultaneous coordinated control of teleoperator/telerobotic systems which have more than six degrees of freedom (e.g., a combined vehicle and arm, each with five or six DOF) are discussed. A new cable-controlled parallel link arm which offers many advantages over conventional arms for space is briefly described. Predictor displays to compensate for time delay in teleoperator loops, the use of state estimation to help human control decisions in space, and ongoing research in supervisory command language are covered. Finally, efforts to build a human flyable real-time dynamic computer-graphic telerobot simulator are described. These projects represent most, but not all, of the telerobotics research in our laboratory, supported by JPL, NASA Ames and NOAA.