Paul Rybski, a systems scientist at the Carnegie Mellon University Robotics Institute, discusses his research and contributions to the development of robots that can determine their own internal “state” as well as that of other nearby robots Thursday, Feb. 23 in a Lawrence University Science Hall Colloquium.
A 1995 Lawrence graduate, Rybski presents, “Robust State Estimation for Intelligent Physically-Embodied Systems” at 4:30 p.m. in Science Hall Room 102 The lecture is free and open to the public.
Due to limited on-board computational power and imprecise sensing systems, researchers are working on novel artificial intelligence techniques by which robots that operate in natural real-world settings can perceive and interact with humans as well as other robots.
Rybski will discuss new research developments in the field of robotics and intelligent sensing which include: spatial reasoning techniques for small robots with limited on-board sensing that allow them to explore and build maps of their environments; algorithms that allow multiple robots to share information about their world and develop consistent world models in the face of sensor and communications errors; and recent results on an algorithm for visual object recognition that “learns” objects by observing how they are used by people.
Rybski, whose research interests include robust high-level environment modeling for sensor-poor robotic systems and distributed control of robot teams, started his robotics research at Lawrence as part of his senior honors project. He was a mathematics/computer science major at Lawrence with an interdisciplinary emphasis in cognitive science and earned a master’s and doctorate degree in computer science at the University of Minnesota.
After completing his Ph.D., he accepted a post-doctoral fellowship at CMU’s Robotics Institute in 2003 and was appointed to the faculty there as a systems scientist last July.