Here you will find highlights of our activities on research and training in the general area of Robotic Mechanical Systems. The group is composed of research associates, postdoctoral fellows, graduate students and undergraduate interns. The focus of our research is the design and control of robotic mechanical systems at large as well as their mechanical subsystems, such as drives and sensors. On the former, you will find the Dual-wheel Transmission (DWT), the Pan-tilt drive (PTD) and the C-drive. The DWT is intended as a drive for wheeled robots; with an original layout of two identical planetary gear trains; motion is transmitted from two identical motors to two identical wheels, to produce the two disparate functions of drive and steer, the two motors operating in a cooperative way. As an undergraduate project, we have produced a rapid prototype of a dual-wheel robot by means of Lego Mindstorms elements. The PTD is a similar mechanism, that transmits torque and motion from two identical motors to one link turning about two axes intersecting at right angles, one for panning, the other for tilting. The C-drive operates under a similar philosophy: a mechanism transmitting motion from two identical motors to one collar that is capable of turning about one axis and translating in a direction parallel to the axis. Two C-drives are currently used to drive a paralel SCARA system, intended for fast pick-and-place operations, what is known as a Schönflies-motion generator. The system is dubbed the Peppermill-carrier, by virtue of the way the robot gripper is manipulated: the same principle as that used to operate long peppermills in Italian restaurants, hence the robot nickname.

On sensors, one main activity is the innovative design of biaxial accelerometers using MEMS technology and isotropic accelerometer strapdowns, intended for the estimation of angular acceleration. Cncurrently with these sensors we are also developing algorithms for the (real-time) estimation of angular velocity and attitude of rigid bodies in free motion.

Jorge Angeles, RMSLab Director