Dan Negrut
University of Wisconsin-Madison, WI·Engineering
Robots built for dangerous or remote places, such as planetary rovers, autonomous construction machines or search-and-rescue robots, cannot be safely designed by trial and error. Computer simulation offers a cheaper, safer path.
- Fields
Engineering
Computer Science
- Topics
robotics
robot simulation
digital twin
- Grants
- 2 known grants, through Oct 2028
What they work on
- Develop the mathematical foundations for differentiable robot simulators, which produce gradients so model parameters can be tuned automatically to match real-world robot behavior.
- Prototype expressive simulators that combine physics equations with components learned from data, and evaluate how well they close the sim-to-real gap in robotic applications.
- Create open training materials for the Chrono simulator, including flipped versions of three courses, self-paced Jupyter notebooks, and new digital-twin models and documentation.
- Train yearly cohorts of California State University students in lunar robotics simulation through online instruction, a NASA Jet Propulsion Laboratory visit and a UW–Madison summer school with a robot competition.
- Provide asynchronous training to thousands of Chrono users through materials on NSF's ACCESS ecosystem, supported by a pilot large language model expert agent.
More about the research
A 'digital twin' is a physics-based virtual copy of a robot and its surroundings, used to test designs before hardware exists. Simulation only helps, though, if virtual robots behave like real ones, and if engineers know how to use these tools well. One project tackles the accuracy of the simulators. The other trains many people to use them, centered on the open-source Chrono simulator.
Why it matters
More accurate robot simulators could cut the time, cost and risk of designing robots for space exploration, construction, search and rescue, and industries such as automotive and aerospace. Training a larger workforce in simulation and digital twins, from university students to engineers at NASA and Department of Energy labs, could help more teams test robots virtually before building expensive or risky hardware.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
Closest labs
Daniele Panozzo
New York University, NY·Computer Science
Develop dPolyFEM, an open-source modular framework for differentiable simulation, which computes how simulation results change with inputs, for elastic objects with large deformations, contact and friction.
Claire Le Goues
Carnegie Mellon University, PA·Computer Science
Develop techniques to infer behavior models with physical attributes from robotic system code, APIs and configuration files, and check those models against component and system properties to detect faults that appear when parts are combined.
Sunita Chandrasekaran
University of Delaware, DE·Computer Science
Build a sustainable, scalable pipeline of Research Software Engineers to support social, behavioral and economic sciences and coastal sciences research, in partnership with Howard University, Delaware State University and Lincoln University.
The school
University of Wisconsin-Madison: 91 labs on PI Pages, $1.9B a year on research, 854 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.