Dual-Agent Framework for Cross-Model Verified Translation of Natural-Language Protocols into Robotic Laboratory Platform

Researchers developed a dual-agent AI framework that translates natural-language biological protocols into executable commands for robotic laboratory platforms, bridging the semantic gap between human-written experiments and automated systems. The system uses a Parser Agent to structure protocols and a Validation Agent to verify accuracy, with successful demonstration on real microplate-based experiments.

This research addresses a fundamental challenge in laboratory automation: converting human-readable experimental protocols into machine-executable instructions. The semantic gap between natural language and robotic control systems has historically required manual translation, limiting scalability of autonomous laboratories. The proposed framework leverages multiple large language models in a verification loop, where a Parser Agent generates structured representations and a separate Validator Agent cross-checks completeness and accuracy. This heterogeneous approach demonstrates that AI systems perform better when specialized for distinct subtasks rather than attempting end-to-end translation. The framework's real-world validation through Bradford protein assays on actual robotic platforms moves beyond theoretical demonstrations into practical deployment territory. The study's systematic evaluation of 7 parser models and 3 validator types reveals how model scale impacts translation reliability—a critical insight for organizations building similar systems. The research directly impacts the emerging self-driving laboratory market, where automated experiment execution accelerates drug discovery, materials science, and synthetic biology research. By reducing manual intervention and translation errors, this framework makes high-throughput automated experimentation more accessible to smaller research institutions. The accuracy-latency trade-offs documented in the study provide a template for similar protocol translation systems. Looking ahead, broader adoption depends on standardizing protocol representations across different laboratory platforms and expanding the framework to handle more complex experimental designs beyond microplate-based assays.

• →Dual-agent AI framework successfully translates natural-language biology protocols into robotic laboratory commands with cross-model verification for accuracy
• →Heterogeneous LLM validation outperforms single-model end-to-end approaches by decomposing the translation task into specialized subtasks
• →Real-world validation demonstrates autonomous execution of Bradford protein assays, proving practical viability beyond theoretical demonstrations
• →Framework's systematic evaluation of model scale and validator types establishes benchmarks for future protocol translation systems
• →Self-driving laboratory automation becomes more accessible by eliminating manual protocol-to-command translation bottlenecks