WHEN EXPERIMENTS RUN THEMSELVES: AN AUTOMATED, MODULAR, INTEGRATIVE SYSTEM FOR CLOSED-LOOP EXPERIMENTAL CONTROL

Complex neuroscience experiments increasingly rely on extended training, adaptive task logic, and tight coordination between sensors, effectors, and neural recording or manipulation systems. In behavioral paradigms, experiments that require frequent manual intervention are labor-intensive, introduce stress, and limit standardization, constrain experimental duration and throughput. As a result, continuous learning, circadian-scale monitoring, and large-cohort studies remain difficult to achieve. While existing systems reliably acquire data or control individual devices, few support automated, scalable experimental control that integrates complex logic, heterogeneous hardware, and long-term autonomy, or that can operate many experiments in parallel. To address this gap, we developed MICS (Modular Interactive Control System), an open-source hardware–software platform for high-throughput, closed-loop experimental control and long-term autonomy across multiple setups. MICS enables the implementation of complex task logic through a modular architecture that abstracts sensors, effectors, and external experimental systems as interchangeable functional units, allowing experimenters to construct rich, state-dependent environments while minimizing technical overhead. MICS orchestrates experimental flow using finite state machine logic embedded on Linux-based Raspberry Pi platforms, providing millisecond-scale control, real-time monitoring, and adaptive task progression based on subject behavior. Integrated logging, centralized supervision, and automated alerting support continuous operation while preserving experimental state and context alongside acquired data. We demonstrate MICS in multi-day experiments that integrate automated mouse training with electrophysiological recordings and closed-loop optogenetic manipulation. By combining integrative experimental control with scalability and autonomy, MICS provides a versatile and cost-effective platform for reproducible, large-scale experimental neuroscience across laboratories, species, and research domains.