Human-in-the-Loop Control of Adaptive and Autonomous Experiments with Bluesky

• Phillip Maffettone (Brookhaven National Lab)

**Abstract**: This presentation explores the active work integrating artificial intelligence and robotics in scientific research facilities, focusing on the National Synchrotron Light Source II. By leveraging the open-source Bluesky project and the Robotic Operating System 2 (ROS2), we demonstrate a framework that enhances experimental control and orchestration while addressing the operational challenges of large-scale user facilities. We showcase the successful implementation of a robotic arm for sample handling at a beamline, achieving a 90% repeatability rate. Additionally, we elaborate on the bluesky-adaptive package, which extends Bluesky’s capabilities to support adaptive and autonomous experiments. Through various case studies, we highlight the optimization of beamline alignment, improved experimental outcomes, and the orchestration of distributed multimodal experiments. This ongoing development includes collaborators from across the 6 light and neutron sources in the DOE complex. This work paves the way for [semi-]autonomous scientific research environments, which remain flexible and leverage human ingenuity. **Speaker Biography**: Phil Maffettone is a computational scientist at the National Synchrotron Light Source II at Brookhaven National Laboratory. He received a DPhil from Oxford studying Bayesian methods for modeling disordered matter from multimodal analysis and built the brain on the world’s first mobile robotic scientist at the University of Liverpool. His current research focuses on integrating robotics and algorithms at large scientific user facilities. He is the project lead for NSLS-II Experimental Tools-III Infrastructure for Experiment Orchestration, a software project building foundational tools for the construction and orchestration of new beamlines.