The ADePT framework for assessing autonomous laboratory robotics

Abstract

Laboratory robotics is advancing from routine automation toward autonomous systems capable of intelligent decision-making and flexible execution. This perspective outlines key milestones and introduces the ADePT framework, which defines four core dimensions of robotic capability proficiency: adaptability and learning, dexterity, perception, and task complexity. We discuss future directions for self-driving laboratories, including robot-centric, end-to-end robotic integration, and collaborative human-robot environments. These scenarios highlight the importance of technological enablers and evolving regulatory paradigms. By connecting present technologies to emerging system configurations, this work offers a foundation for designing autonomous laboratory ecosystems that support scientific discovery and operational efficiency.