Ross A. Knepper
My research addresses the algorithmic aspects of autonomous robot systems and human-robot systems. Looking at humans and robots together as one system allows us to optimize its overall performance. Human beings are experts at distributed computation — we call it teamwork. Human algorithms and protocols for teamwork have been honed by culture and evolution alike. Robots recognize the wisdom of human computation methods by learning to understand human signals and conventions, such as natural language and gesture, for accomplishing activities together. My research combines insights from psychology, ethnography, and linguistics with formal representations from topology, differential geometry, and others.
A consequence of this approach is that people can program robots without the need for special training. These technologies enable autonomous robots to safely act as peers to humans in activities like navigation, manipulation, and assembly, all in environments structured for people. By working together, people and robots will increase productivity, adaptability, and worker and customer safety, thus helping all sectors of the economy to grow.