Abstract

In the area of multi-robot systems, an understanding of collaboration has been developed and is showing great promise in real-world applications including self-driving cars, industrial automation, precision agriculture, and disaster response. However, existing multi-robot methods still leave much to be desired. In this talk, I will argue for innovation in several areas of multi-robot coordination that are often seen in real-world applications: (1) intermittent monitoring of slowly-evolving spatiotemporal processes on large scales; (2) robots that interact asymmetrically due to imperfect sensors and interference; (3) robots that collaborate as an informed response to a team objective and the environment; and (4) robots that interact with human experts. Towards this vision, I will first discuss methods we have developed that generate temporal deployment plans for multi-robot teams that balance the cost of deployment with the quality of information gathered about some evolving process of interest. In particular, I will outline a combinatorial optimization approach to this “intermittent deployment” problem, yielding greedy solutions with bounded suboptimality. Next, I will detail recent results in coordinated motion control for robots interacting with sensors exhibiting some limited field of view (FOV), along with experimental results in outdoor environments. Finally, I will outline planning problems we have developed that generate temporal interaction structures for multi-robot teams along with trajectories that complement the efforts of humans experts in a search and rescue domain. Throughout the talk I will show examples of the projects at the Coordination at Scale Lab (CAS Lab) at Virginia Tech that motivate the above methods, and conclude by discussing future directions in the area of multi-robot systems.

Biography

Ryan K. Williams received the B.S. degree in computer engineering from Virginia Polytechnic Institute and State University in 2005, and the Ph.D. degree from the University of Southern California in 2014. He is currently an Assistant Professor in the Electrical and Computer Engineering Department at Virginia Tech where he runs the Laboratory for Coordination at Scale (CAS Lab). His current research interests include control, cooperation, and intelligence in distributed multi-agent systems, topological methods in cooperative phenomena, and distributed algorithms for optimization, estimation, inference, and learning. Williams is a Viterbi Fellowship recipient, has been awarded the NSF CISE Research Initiation Initiative grant for young investigators, is a Junior Faculty Award Recipient at Virginia Tech, is a best multi-robot paper finalist at the 2017 IEEE International Conference on Robotics and Automation, and has been featured by various news outlets, including the L.A. Times.