Google Research introduces VERID framework to ground AI agents in biological wet lab reality

Google Research presents the VERID framework, a semantic digital twin of the laboratory workcell that tracks robotic geometry, well volumes, and reagent identities in real time. The system enables AI agents to perform safe physical experimentation through live 3D state tracking, virtual dry runs, and hardware safety checks, preventing physically impossible or logistically invalid commands. Demonstrated on aptamer and de novo protein design prototypes, the framework allows closed-loop scientific discovery where AI agents can ideate, test, and learn from real-world wet lab feedback.