The development
What happened
Google Research's October 16 report describes three projects combining AI proposals with laboratory state and experimental feedback. VERID tracks equipment and materials in a digital twin, while separate aptamer and protein-design studies explore repeated design-test cycles.
Key facts
Built on PyLabRobot, VERID tracks robot geometry, reagents and well volumes, with synchronized video and action logs.
Simulation and hardware checks reject invalid instructions before robot execution.
Google reports confirmed aptamer hit rates of 33.5% and 42.1% on two targets after three rounds and correction of experimental artifacts.
A human-agent workflow reported more than eightfold faster structure prediction and single-digit nanomolar FAP binders; these are research results, not clinical outcomes.
The context
Why it matters
Physical feedback can expose errors invisible to a software-only agent. Google says testbeds are controlled and noninfectious, with biosecurity screening. Combining the components into a fuller closed-loop system remains future work under scientific direction.