B-spline Policy Speeds Up Robot Manipulation
Haoyu_Xiong_ · x · 2026-07-15
- This thread introduces B-spline Policy (BSP), co-led by Haoyu and xshenhan, aiming to improve action representation in robot manipulation.
- While traditional robot manipulation primarily uses "success rate" as the main metric, this work argues that speed is equally important. They use continuous B-spline curves to parameterize actions, rather than predicting discrete, fixed-frequency action chunks.
- Combined with system design, BSP enables low-cost robotic arms to perform faster manipulation. The author highlights that the robot motions in the video are noticeably faster and smoother than baseline methods, all in real-time.
- This is a research/engineering breakthrough focused on robot manipulation, highlighting improvements in both action representation and execution speed.
Related event: B-spline Policy proposes faster, smoother robot manipulation(5 posts)→
More from Embodied
- NVIDIA open-sources FLORA, a robotics data engine ~125x cheaper than GPT-6 Astra — nvidia · 2026-10-27
- Embodied AI startups need a different operational discipline, investor says — ___Mufasaa · 2026-10-27
- π0.5 study: raising demos from 10 to 438 lifts Place Ball in Bowl success from 7% to 83% — stingning · 2026-10-27
- GPT-6 Astra controls dexterous hands zero-shot at just 5.2% success in 500 trials — stingning · 2026-10-27
- HandsOn benchmark: 66 verified tasks and 18,000 human demos for dexterous manipulation — stingning · 2026-10-27
- DexPort: a handheld interface with a real robot hand for dexterous-hand data collection — pabbeel · 2026-10-27