Benchmarking Autonomy in Scientific Experiments: A Hierarchical Taxonomy for Autonomous Large-Scale Facilities
This work proposes the Benchmarking Autonomy in Scientific Experiments (BASE) Scale, a 6-level taxonomy (Levels 0-5) for autonomous Large-Scale User Facilities. Unlike owner-operator models, User Facilities require zero-shot deployment where agents must operate immediately. The Inference Barrier (Level 3) is identified as the critical latency threshold where decisions shift from scalar feedback to semantic digital twins, providing facility directors and funding bodies with a standardised metric to assess risk and quantify experimental workflow intelligence.
Autonomous experimentation
Artificial intelligence (AI)
Large-scale facilities
Taxonomy
Robotics
A 6-level taxonomy (BASE Scale) for benchmarking autonomy in scientific experiments, specifically adapted for the unique constraints of Large-Scale User Facilities.
James Le Houx 