Human Body as a Model for Distributed Intelligence in Self-Driving Laboratories

Self-driving laboratories are emerging as a transformative paradigm for accelerating scientific discovery through the integration of artificial intelligence, automation, and robotics. However, the intrinsic complexity and unpredictability of experimental environments challenge the feasibility of fully centralized laboratory control architectures. In this perspective, we propose the human body as a conceptual model for the design of distributed-intelligence self-driving laboratories. Inspired by the organization of the nervous system, we distinguish two complementary levels of control: a high-level cognitive layer responsible for reasoning, planning, and scientific decision-making, and a distributed adaptive layer responsible for local operational optimization, sensing, and real-time reaction to environmental changes. We discuss how such an architecture can enable laboratories to dynamically adapt to unpredictable events including precipitation, thermal runaway, viscosity changes, robotic handling failures, or evolving reaction conditions. Concrete implementation examples are presented, including adaptive instrumentation, distributed sensing, automated method selection, Bayesian optimization, multimodal robotic perception, and local orchestration frameworks currently under development at Swiss Cat+ at EPFL. We further propose a two-layer architecture composed of an interpretation and compilation layer coupled to a self-adaptive operational layer capable of autonomous execution and optimization. Finally, we discuss how this distributed-intelligence paradigm could provide the basis for future cloud-distributed laboratory infrastructures combining centralized generative AI systems with locally adaptive laboratory environments. Overall, this perspective argues that distributed adaptive intelligence, rather than purely centralized control, may constitute a critical foundation for the next generation of scalable, resilient, and autonomous scientific laboratories.