Kris Hauser
University of Illinois Urbana-Champaign, IL·Computer Science
Robots that work closely with the physical world, whether digging in soil or examining a person's eye, must predict what will happen when they act. They also need to stay safe and reliable when things do not go as expected.
- Fields
Computer Science
Engineering
Biomedical Engineering
- Topics
robotics
robot manipulation
granular materials
- Grants
- 2 known grants, through Aug 2027
What they work on
- 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.
- Scale GNN terrain models to large sites like construction zones by dividing geometric maps into local GNNs, and build multi-material GNNs that let excavator robots detect hidden roots or rocks by feel.
- Build a robot-mounted eye exam instrument that extends robot-aligned optical coherence tomography (OCT) with added modes for imaging the cornea, lens and retina and measuring eyeglass prescriptions.
- Create active safety systems for close-contact robotic eye exams, combining real-time human tracking, human motion prediction and robot motion planning, then evaluate them on head and eye phantoms and human subjects.
More about the research
Two projects share this theme of perception-driven robot control. One teaches robots to model and shape granular materials like sand and soil. The other builds a robot that performs eye exams on patients without touching them.
Why it matters
Robots that predict how soil behaves could reliably dig pits and channels and level foundations on construction sites with sandy, rocky or root-filled ground. Safe, no-touch robotic eye exams could widen access to preventive eye care, lower costs, and reduce infection risk for patients and medical staff.
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.
Ioannis Karamouzas
University of California-Riverside, CA·Computer Science·Early-career
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.
Momotaz Begum
University of New Hampshire, NH·Computer Science
Develop learning-from-demonstration (LfD) algorithms that learn a task's safety features from human video demonstrations and filter out demonstrations that would teach the robot unsafe behavior.
The school
University of Illinois Urbana-Champaign: 41 labs on PI Pages, $907M a year on research, 870 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.