Uploaded May 2022 | Updated September 2026, 10 hours ago
Lockheed Martin Robotics Seminar: Deployable Robots that Learn
Xuesu Xiao
Research Affiliate, Roboticist
University of Texas at Austin, The Everyday Robot Project at X
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 three 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 fine-tunes existing motion planners by learning from simple interactions with non-expert users before autonomous deployment and adapts to different deployment scenarios; (2) Learning Inverse Kinodynamics allows robots to learn from in-situ vehicle-terrain interactions during deployment and accurately navigate at high speeds on unstrucured off-road terrain; (3) Learning from Hallucination enables agile navigation in highly-constrained deployment environments by reflecting on previous deployment experiences 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.
For more information on the Maryland Robotics Center see:
https://robotics.umd.edu
Lockheed Martin Robotics Seminar: Deployable Robots that Learn
Xuesu Xiao
Research Affiliate, Roboticist
University of Texas at Austin, The Everyday Robot Project at X
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 three 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 fine-tunes existing motion planners by learning from simple interactions with non-expert users before autonomous deployment and adapts to different deployment scenarios; (2) Learning Inverse Kinodynamics allows robots to learn from in-situ vehicle-terrain interactions during deployment and accurately navigate at high speeds on unstrucured off-road terrain; (3) Learning from Hallucination enables agile navigation in highly-constrained deployment environments by reflecting on previous deployment experiences 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.
For more information on the Maryland Robotics Center see:
https://robotics.umd.edu
