Stanford Seminar - Deploying Autonomous Service Mobile Robots, and Keeping Them Autonomous

Explore the challenges and solutions in deploying autonomous service mobile robots in unstructured human environments through this Stanford seminar. Delve into three key challenges: ensuring robustness to environmental changes, anticipating and overcoming failures, and efficiently adapting to user needs. Learn about probabilistic perception models, introspective perception, and the use of representation learning and program synthesis to enhance robot autonomy. Discover how these research advances enable the deployment of wheeled and legged autonomous mobile robots on a campus scale at UT Austin, performing tasks that require robust mobility both indoors and outdoors. Gain insights from Joydeep Biswas of UT Austin as he discusses cutting-edge techniques in autonomous robotics, including visual representation learning for preference-aware planning and physics-informed program synthesis for action selection policies.