Isaac Lab

Headlines · 2

1. 5d
   SO-101 Bench adds a digital-twin benchmark for spatial robot manipulation

   Truman Hickok’s paper introduces SO-101 Bench, an Isaac Lab digital twin of a physical SO-101 setup for language-conditioned spatial manipulation. It includes four tasks, 138 objects across seen, unseen-instance and unseen-class splits, 916 evaluation episodes and 3,629 simulated demonstrations; the digital twin reproduced the physical arm’s policy ranking with Spearman rho of 1.00, although simulation overestimated success. Across nine VLAs and world-action models, the best learned policies scored nearly 61%, while an LLM agent combining code generation with a learned grasping policy reached 83.2%.

   Hugging Face · Papers · 🔥 0
2. Oct 8
   NVIDIA publishes five-step workflow for SimReady robotics assets using frontier AI models

   An NVIDIA technical blog post walks through preparing CAD assets for robotics simulation using SimReady Foundation specifications, NVIDIA Omniverse libraries and agentic NVIDIA skills. The workflow has five steps: STEP-file conversion, appearance validation, physics configuration, SimReady validation, and a two-arm pick-and-place task with an ABB Robotics YuMi robot in NVIDIA Isaac Sim. GPT-6 Astra is cited as one example frontier model that interprets reference materials and calls Omniverse tools to write Python code; NVIDIA also lists its simulation assumptions on mass, inertia and collision geometry, and notes results may vary by model and prompt.

   NVIDIA Technical Blog · 🔥 0

Experience and discussion from the community

Share my Isaac Lab experienceAsk about Isaac Lab