UUnified Layered Architecture (ULSA) and SDL-BI: A Governance-Embedded Benchmarking Framework for Autonomous Self-Driving Laboratories (2022–2026)ntitled Item

Autonomous Self-Driving Laboratories (SDLs) have transitioned from experimental automation platforms to adaptive cyber-physical discovery systems between 2022 and 2026. Demonstrations across solid-state materials synthesis, continuous-flow chemistry, synthetic biology, nano photonics, and AI-integrated manufacturing confirm the technical viability of closed-loop experimentation at operational scale.Despite rapid technical progress, SDL infrastructure remains architecturally fragmented. Performance metrics are reported heterogeneously across domains, and governance mechanisms are inconsistently embedded within experiment-planning systems. Institutional comparability and capital allocation, therefore, lack a standardized evaluative structure.This manuscript introduces three primary contributions:Unified Layered SDL Architecture (ULSA) — A formal seven-layer cyber-physical abstraction representing SDL systems as interacting modules: Goal Specification (G), Planning Agent (P), Orchestration (O), Robotic Execution (R), Sensing (S), Data & Model Retraining (D), and Governance & Constraints (V).Self-Driving Laboratory Benchmark Index (SDL-BI) — A composite benchmarking framework integrating seven normalized performance dimensions: Throughput, Optimization Efficiency, Discovery Yield, Resource Efficiency, Robustness, Model Intelligence, and Return-on-Discovery.Governance-Embedded Deployment Protocol — A structured integration model embedding constraint-aware planning, human override protocols, compliance logging, and audit-ready retraining pipelines within the autonomy loop.The paper synthesizes deployments reported between 2022 and 2026 and formalizes SDLs as constrained adaptive control systems with embedded learning dynamics. It provides a domain-neutral reference architecture and benchmarking scaffold for institutional adoption, cross-platform comparison, and regulatory integration.Autonomous laboratories represent a structural transformation in scientific production. Architecture must be unified. Metrics must be normalized. Governance must be embedded.This work establishes that foundation.