Lingchu Unveils Embodied Large Model Deployment
创业邦 · wechat · 2026-07-16
This article focuses on Lingchu's demonstrations and partnerships at WAIC, highlighting its embodied large model capabilities rather than just mechanical movements. ### Demonstrations - The booth featured **5 Demos + 1 interactive experience**, covering scenarios like phone box assembly, World Model data conversion, and optical modules - Use cases included insertion detection, vacuum packaging, and automated boxing, emphasizing precision manipulation, long-horizon tasks, and industrial coordination ### Industrial Collaboration - Lingchu partnered with **Changfei Group** to deploy multi-robot collaboration powered by the **Psi-R2 embodied large model** for optical module inspection and packaging - The system achieves sub-millimeter positioning, yield sorting, vacuum sealing, and packing workflows ### Data and Model Strategy - The company invests heavily in human data collection, believing that "model capability determines the upper limit" - They have accumulated **100,000 hours of human data** and **open-sourced a 1,000-hour dataset** - The core logic is driving embodied AI iteration through a "data-model-scenario" closed loop The article concludes that the embodied AI race is about software and iterative capabilities, not just hardware.
More from Embodied
- Genspark launches a 2.95 mm AI voice recorder priced at $179 — Scobleizer · 2026-07-21
- Robot deployment data may look valuable, but narrow deployments create disconnected data islands — chris_j_paxton · 2026-07-21
- Mid-life crisis AI journey: dual 3090s + Qwen3.6-27b boost productivity 5x — Civil_Fee_7862 · 2026-07-21
- A talking humanoid robot demo at WAIC drew a “creepy” reaction — minchoi · 2026-07-21
- A creator says he appeared on NBC’s Today show to talk about robot fighting — chris_j_paxton · 2026-07-21
- WAIC 2026’s clearest embodied-AI signal: a server factory deployment plan targeting 10,000 units — 量子位 · 2026-07-21