MICHAEL ERDMANN
Professor, Computer Science and Robotics
www

I am interested in Robotics and in Computational Molecular Biology, with shape sensing as a unifying theme. My research draws on tools from geometry, mechanics, and stochastic processes.

On the robotics side, I wish to make robots act purposefully and successfully despite the inevitability of noisy sensors, imprecise actions, and mistaken world models. I have worked on grasping strategies, friction representations, and automatic planning to overcome uncertainty. Recently, I created a two-palm robot. The robot consisted of two manipulator arms cooperating to manipulate objects without the need for full kinematic constraint. The arms "programmed themselves", that is, they invoked an automatic planner to find sequences of motions for reorienting objects in their palms. The planner built a geometric graph based on a critical event analysis of the underlying mechanics.

At present, my students and I are developing sensing strategies to acquire the shape of unknown objects concurrently during manipulation. Amazingly, it is possible for a robot to determine the shape of an unknown object from tactile information without requiring that the object be held firmly; the object may slip and slide in the hands. Such strategies are useful for robots manipulating novel objects in unstructured environments. One application we are considering is desktop robotics.

On the molecular side, I am interested in protein shape and motion. Currently I am collaborating with researchers in the Department of Biological Sciences to determine protein homology from sparse NMR data. Our approach draws heavily on techniques from Computational Geometry.