Lightwheel and Synopsys bring CAE-grade contact physics into robot simulation
lukas_m_ziegler · x · 2026-11-06
Most robot simulators get the scene to look right but stay rough on contact: the real forces when a gripper presses, inserts, or holds something — which is the whole game for industrial tasks.
At a joint R&D showcase at GTC Berlin, Lightwheel AI and Synopsys translated decades of CAE engineering simulation accuracy (material, contact, and load physics used to design cars, aircraft, and electronics) into embodied simulation, so robot training and evaluation can sit on measurable physics instead of approximations.
The proof case was an RJ45 cable insertion task — the kind of fiddly, force-sensitive job that usually falls apart in sim — recalibrated against real-world measurements.
More from Embodied
- I watched a humanoid robot clean my apartment: remote-piloted, $30/hour, data-hungry — MIT Tech Review AI · 2026-11-06
- Harvard-led team's MFM memory system lets frozen-weight agents keep learning in Minecraft — du_yilun · 2026-11-06
- RAX lets coding agents unscrew a real jar with dexterous hands: 6 chained skills, 140° rotation — du_yilun · 2026-11-06
- HeLM: explicit language memory gives VLA robots +12.8 points on non-Markovian manipulation — SeoulNationalUniv · 2026-11-06
- 3 Years and Tens of Thousands of Lines of Code vs. One Prompt: AI Scene Generation's Leap — liliyu_lili · 2026-11-06
- UWaterloo 3D prints Invar36 flexures with directly mounted interferometer optics — jwt0625 · 2026-11-06