BSP Makes Robotic Arms Faster and Smoother
Haoyu_Xiong_ · x · 2026-07-16
This thread introduces a novel robotic control method called B-spline Policy (BSP).
Core Concept
- Traditional robotic control typically outputs fixed-frequency, discrete action chunks. BSP instead parameterizes actions using continuous B-spline curves.
- The goal isn't just to improve the success rate, but to simultaneously optimize for speed and motion smoothness.
Results & Significance
- Combined with system design, BSP enables low-cost robotic arms to perform faster manipulations.
- The author notes that the robot's movements in the video are visibly faster and smoother than the baseline, with all demonstrations running in real-time.
- This research was co-led by @xshenhan, and the author promises more details in upcoming threads.
Related event: B-spline Policy proposes faster, smoother robot manipulation(5 posts)→
More from Embodied
- Bedrock Robotics raises $600M as its fully autonomous excavators do paid work on live job sites — Sethwinterroth · 2026-10-27
- AW-0 hits 80.8% on RoboTwin tasks, beating strongest baseline by 16 points — trevordarrell · 2026-10-27
- AW-0's code stream fixes world model prediction: picks mango where baseline grabs apple — trevordarrell · 2026-10-27
- AW-0 runs at 230ms on a single RTX 5090, 2.4x faster than autoregressive code planning — trevordarrell · 2026-10-27
- AW-0 debuts: one generative world action model plans in code, predicts, and controls robots — trevordarrell · 2026-10-27
- Skyfall AI's SKY-001 world model turns sketches into CAD builds in 3.9s, vs 199s for GPT-6 Astra — rohanpaul_ai · 2026-10-27