Ioannis Karamouzas
University of California-Riverside, CA·Computer Science·Early-career
How can computers simulate the way humans and robots move, both realistically and fast enough to be useful? Animation tools are quick but often physically inaccurate.
- Fields
Computer Science
Engineering
Biomedical Engineering
- Topics
musculoskeletal simulation
musculotendon dynamics
biomechanics
- Grants
- 2 known grants, through Dec 2029
What they work on
- Build a constraint-based simulation framework that treats muscles, tendons, skeleton and environmental contacts as one coupled implicit system, for stable, physiologically realistic simulation of contact-rich human motion.
- Train neuromuscular controllers with reinforcement learning so that simulated characters produce realistic muscle activation patterns instead of implausible joint-driven motion.
- Validate the musculotendon simulator against in-vivo biomechanical and physiological data, and benchmark its accuracy and speed against existing simulation tools.
- Learn the navigation decisions humans make in different interaction scenarios from public human trajectory datasets, and add them to local planners for human-like robot decision making.
- Develop a reinforcement learning framework and new human-robot interaction datasets for multi-robot navigation in human-populated spaces, tested in simulation and on real robots.
More about the research
Biomechanics models are accurate but slow and specialized. Robots, meanwhile, still lack people's instinct for when to yield and when to move decisively in a crowd. This research combines physics-based simulation, motion planning and reinforcement learning to model and produce realistic, human-like movement and navigation.
Why it matters
Fast, physiologically accurate movement simulation could give sports, health and rehabilitation scientists tools to study complex real-world motions and possible injury risks. It could also improve training simulators, robotics and games. Human-like multi-robot navigation could let cleaning, delivery and warehouse robots work safely and smoothly among people in homes and workplaces.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
Closest labs
Saurabh Gupta
University of Illinois Urbana-Champaign, IL·Computer Science·Early-career
Design predictive models of people, places and objects that a robot needs for decisions, such as which side an oncoming person will pass on or where household items are likely to be kept.
Kris Hauser
University of Illinois Urbana-Champaign, IL·Computer Science
Develop graph neural networks (GNNs), models that treat soil as connected regions, to predict how sand or soil responds to robot actions and to plan digging that forms target shapes, updated from camera images and force sensing.
Jaime Fernández Fisac
Princeton University, NJ·Engineering·Early-career
Develop a framework based on dynamic game theory in which a robot guarantees safety against the worst case within an Operational Design Domain (ODD), the conditions designers or regulators say it must handle, while modeling how agents gain information over time.
The school
University of California-Riverside: 21 labs on PI Pages, $232M a year on research, 328 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.