Chemist Eye: A Visual Language Model-Powered System for Safety Monitoring and Robot Decision-Making in Self-Driving Laboratories
Published in Digital Discovery (Under Review), 2025
The integration of robotics and automation into self-driving laboratories (SDLs) can introduce additional safety complexities beyond conventional research laboratories. Chemist Eye is a distributed safety monitoring system that integrates multiple monitoring stations (RGB, depth, infrared) with a vision-language model (VLM) to detect PPE non-compliance, fires, and medical emergencies, and to assist robot decision-making (e.g., steering mobile robots away from hazards and issuing audible warnings). The system connects to third-party messaging platforms for alerts. Evaluations on an SDL with three mobile robots show hazard detection at ~97% and decision-making performance at ~95%.
Download the preprint (arXiv:2508.05148)
Recommended citation: Munguia-Galeano, F., Zhou, Z., Veeramani, S., Fakhruldeen, H., Longley, L., Clowes, R., & Cooper, A. I. (2025). Chemist Eye: A Visual Language Model-Powered System for Safety Monitoring and Robot Decision-Making in Self-Driving Laboratories. arXiv:2508.05148.
Download Paper