Date
Event Location
2219EB

CANVAS Seminar

Dr. Changjoo Nam

Postdoctoral Fellow
Robotics Institute at Carnegie Mellon University

Tuesday, March 13, 2018

10:00 a.m. – 11:00 a.m.

2219 EB

“Toward Fluid Coordination of Multiple Robots and Humans in Navigation Tasks”

Abstract:

Multi-robot systems are advantageous to complete distributed tasks collectively through coordinated performance. To maintain optimality of the coordination in dynamically changing environments, the robots need to communicate to share new information arising from dynamic circumstances and recompute the optimal solution accordingly at run-time. These run-time efforts incur additional expenses and overhead which degrade the team performance, especially in the environments where the communication cost is prohibitively large (e.g., urban areas with severe radio interference or underwater). In systems where humans are included in the loop for supervisory control, unnecessary interventions applied by humans owing to low trust in the robots may also incur additional inter-robot communication and computation to comply with the supervisory commands.

In this talk, I will present my work on reducing run-time efforts attribute to dealing with dynamically arising events, especially in navigation tasks, such as delays occurring in transportation systems or commands from a human supervising autonomous navigation of robots. First, I will introduce a set of algorithms that utilize upfront computation to analyze available information regarding navigation, thereby avoiding global communication and centralized computation at run-time. Second, I will present a recent work on modeling and estimating human trust in multiple robots performing emergent swarm behaviors for cooperative navigation. The estimates can be used to reduce unnecessary interventions from humans, which may require additional expense to follow the commands. Finally, I will conclude with my research direction toward developing a future transportation system that considers optimized navigation and different human roles included in the system.

Bio:
Changjoo Nam is a Postdoctoral Fellow in the Robotics Institute at Carnegie Mellon University, where he investigates human-robot interaction and multi-robot planning focusing on robot navigation tasks. He received his Ph.D. degree in Computer Science from Texas A&M University in 2016 and an M.S. and B.S. degree in Electrical and Computer Engineering from Korea University in 2011 and 2009, respectively. His graduate studies include optimizing collective performance of multiple robots and simultaneous localization and mapping (SLAM). Before joining Texas A&M, he was a research scientist at Korea Institute of Science and Technology where he worked on outdoor localization of multiple robots. His research interests include multi-robot coordination and planning, human-robot interaction, and autonomous vehicle navigation.

Faculty Host:  Dr. Linos Jacovides (linos@comcast.net)

Michigan State University is committed to providing equal opportunity for participation in all programs, services and activities.  Accommodations for persons with disabilities may be requested by contacting Dr. Cagri Ulusoy (Ulusoy@msu.edu)