Shanghai AI Lab: SIM1 & GAUGE on physics-aligned world models
青稞AI · wechat · 2026-09-01
Shanghai AI Lab is hosting a livestream on physics-consistent world models, featuring Zhou Yunsong and Wang Shuai. The discussion centers on two papers, SIM1 and GAUGE:
- SIM1 (Physics-Aligned Simulator): A physics-aligned data engine for deformable object manipulation (e.g., folding clothes, cable routing). By building metric-consistent digital twins and elastic dynamics models, it converts few real demos into large-scale synthetic data. Experiments show pure synthetic data training matches real data baselines at a 1:15 ratio, with 90% zero-shot success rate.
- GAUGE: A measurement-grounded benchmark for physical fidelity. Covering 22 controlled tasks, evaluations reveal significant deviations in current physics engines during impulse contact and soft body deformation. Video world models often generate physically incorrect trajectories despite visual plausibility.
Together, they form a "Generate (SIM1) - Verify (GAUGE)" loop to bridge the sim-to-real gap in robotics learning.
More from Embodied
- Peking University's SimForcing distills simulation motion priors into robot world models — PekingUniversity · 2026-10-16
- NVIDIA and Mercedes-Benz Expand Partnership on Next-Gen Automated Driving — nvidia · 2026-10-16
- ~25,000 Humanoid Robots Shipped in H1 2026, Over 95% from Chinese Vendors — ingliguori · 2026-10-16
- openpilot brings stoplight-stopping autonomy to a 2019 Corolla — k7agar · 2026-10-16
- Hironobu Fujiyoshi publishes deck: from AI agents to Agentic Robotics — HirokatuKataoka · 2026-10-16
- CoDA co-evolves simulation and robot policies via shared failure memory, lifting OOD success from 5.3% to 27.6% — PLAN-Lab · 2026-10-16