Graduate Seminar: Developing Robots for Dynamic, Safety-Critical Environments

Abstract

Robots are entering everyday, human-centered spaces, such as hospitals, search and rescue zones, and homes. They have great potential to assist and support people, but in these environments, people's task load often fluctuates rapidly and they regularly need to manage multiple tasks at once. Thus, for robots to be successful, they must be aware of changing environmental and task contexts, as well as users' changing preferences and abilities. The goal of my research is to address challenges in integrating robots into real-world settings by analyzing relationships between workload and teaming dynamics, developing new methods for robots to adapt to changing contexts, and creating robotic systems specifically designed for dynamic environments.

In this talk, I will discuss my work developing robots that are contextualized for dynamic environments, specifically emergency departments. I will describe my work exploring how factors such as workload and task context affect human-robot teaming strategies, including in proximate human-robot teaming and shared control. Finally, I will discuss my recent work developing methods for robots to adapt to people as their workload and context changes. My work will enable robots to autonomously adapt to people in real-world environments so they can provide better support and ultimately allow people to focus on critical tasks with lower workload.

Speaker Biography

Sachiko Matsumoto is a final-year Ph.D. student in Computer Science and Engineering at UC San Diego. Her research interests lie at the intersection of robotics, artificial intelligence, and health technology. She works on developing robotic systems that can support people in dynamic, real-world environments, such as emergency departments. She has published papers in top venues in her field, including HRI and THRI. Prior to the Ph.D., she received Bachelor's degrees in Computer Science and Physics at UC Berkeley.