# Spotter paper: Embodied model executes autonomously, VLM corrects errors on demand to reduce robot operation latency

- Source: Xbotics具身智能实验室 · 微信公众号
- Published: 2026-10-21T09:00:00.000Z
- RoboHot: https://robohot.news/en/items/f52kpq7b5ytokbg9sfr4wxxbv
- Original: https://mp.weixin.qq.com/s?__biz=MzkyMTc1NTI3Ng==&mid=2247509414&idx=1&sn=25e611d965655c2a524a2d56716264b4&chksm=c08cf873cf657b6507837a0b7282028aa2379374547e5b8b5730ff7e1b19021357545138bbba&scene=126&sessionid=1792593351#rd

## Summary

Li Long, a PhD student at Griffith University, Zhao Qichao, a PhD student at Tsinghua University's Institute for Interdisciplinary Information Sciences, and teams from Fudan University and Tongji University propose Spotter: an embodied model executes continuously and autonomously while a lightweight local model supervises in parallel, and the VLM intervenes to correct errors only when they occur, avoiding the robot waiting for VLM inference at every step.