The Autonomous Systems Lab, directed by Professor Stephen L. Smith, develops algorithms for the planning, control, and coordination of autonomous systems and robots. Our research combines optimization, control theory, and machine learning, with an emphasis on systems that operate reliably in the real world, from a single robot navigating an unknown environment to large teams coordinating at scale. The lab is part of the Department of Electrical and Computer Engineering at the University of Waterloo.
Research
Nonprehensile manipulation and interactive navigation. Robots that push, nudge, and rearrange objects to make their way through cluttered spaces — learning the physics of contact so they can plan actions that go beyond simply avoiding obstacles.
Field robotics in harsh environments. Autonomous navigation in difficult real-world settings, including surface vessels that plan a route through fields of broken sea ice.
Learning-based planning and control. Bringing modern machine learning — diffusion policies, learning-based predictive control, reservoir computing — together with control theory, so robots learn from data while keeping guarantees on safety and performance.
Informative path planning and persistent monitoring. Planning routes that gather the most useful information and keep watch over changing environments, from covering unknown spaces to environmental and ocean sensing.
Human-aware and social robot navigation. Robots that move and work among people — learning human preferences, deciding when to lead or follow, and navigating in ways that are legible and predictable to those nearby.
Multi-robot coordination and optimization. Task allocation, routing, coverage, and large-scale fleet and motion-planning optimization for teams of robots and autonomous vehicles.