🔒 Interview Brief: DexHand Sensing × Data Flywheel
Private interview brief for dexterous hand sensing integration, data collection, consistency, and commercialization. Available on request.
Available on request
Blog
EN
中文
A lightweight space for publishing my ideas. Add a new HTML page, then add a new post card below.
🤖 DexHand Sensing System
Sensor stack, multi-rate loops, time sync, calibration, and a learning-ready data schema for dexterous hands.
🧩 DexHand Sensing System Design (Deployment × Data × Self-Recovery)
A deployable sensing design for dexterous hands: multi-rate loops + time sync + two-mode data (Deploy vs Research), using A (Teacher→Student distillation) + C (RM/QC flywheel) to get stable deployment and high-resolution research value.
🧠 Thinking Board
A simple local board for collecting thoughts; great for drafts before turning them into publishable posts. (Notes are stored in your browser’s local storage.)
🦾 VTLA Plan B: Why Evo-RL Should Not Be the Main Framework
For VTLA, the safer path is UniVTAC/TacSL + Diffusion Policy first, then Evo-RL for real-robot continual improvement.
🧗 Friction Margin: From Climbing to DexHands
Climbing is mostly “estimate friction margin under uncertainty”. This post borrows the friction-cone / pre-slip mindset to design tactile sensing and control loops for dexterous hands.