LCSR Seminar: Xuesu Xiao

Description

Xuesu Xiao, incoming assistant professor in the Department of Computer Science at George Mason University, will give a talk titled "Deployable Robots that Learn" for the Laboratory for Computational Sensing and Robotics.

Please attend the event using the Zoom link, available online.

Xiao will join George Mason University starting Fall 2022. Currently, he is a roboticist on The Everyday Robot Project at X, The Moonshot Factory, and a research affiliate in the Department of Computer Science at The University of Texas at Austin. Xiao's research focuses on field robotics, motion planning, and machine learning. He develops highly capable and intelligent mobile robots that are robustly deployable in the real world with minimal human supervision.

Abstract:

While many robots are currently deployable in factories, warehouses, and homes, their autonomous deployment requires either the deployment environments to be highly controlled or the deployment to only entail executing one single preprogrammed task. These deployable robots do not learn to address changes and to improve performance. For uncontrolled environments and for novel tasks, current robots must seek help from highly skilled robot operators for teleoperated (not autonomous) deployment.

In this talk, I will present two approaches to removing these limitations by learning to enable autonomous deployment in the context of mobile robot navigation, a common core capability for deployable robots: (1) Adaptive Planner Parameter Learning utilizes existing motion planners, fine-tunes these systems using simple interactions with non-expert users before autonomous deployment, adapts to different deployment environments, and produces robust autonomous navigation; (2) Learning from Hallucination enables agile navigation in highly-constrained deployment environments by exploring in a completely safe training environment and creating synthetic obstacle configurations to learn from. Building on robust autonomous navigation, I will discuss my vision toward a hardened, reliable, and resilient robot fleet which is also task-efficient and continually learns from each other and from humans.