The goal of SToMP is to reduce the complexity of sensing and planning by using powerful mathematical tools developed in the field of topology. Topology is the science of abstraction: the process of casting away irrelevant details to focus on the essence of a problem. The SToMP project brings together researchers from eight institutions with a variety of backgrounds in mathematics, computer science, robotics, and sensors.
The focus of the research at Carnegie Mellon is to study the interplay of sensing and action required for autonomous systems to perform robustly in unknown environments. One motivating goal is to understand how systems can recognize changes in their operating environments, infer their capabilities in these new environments, and recover from unexpected events.
“As an example, imagine a car that drives into a ditch,” Erdmann said. “Ideally the car should be able to infer the shape of the ditch by observing strategically placed sensors in its chassis, wheels, and suspension, then drive itself out of the ditch.” The overall project is directed by researchers at the University of Illinois and includes experts from Bell Labs/Lucent, Arizona State University, Rochester University, Melbourne University, the University of Pennsylvania and the University of Chicago.