Driving discoveries via collaboration
Modern scientific experiments are becoming increasingly complex, generating enormous datasets and requiring rapid decision making. To overcome these challenges, researchers at Brookhaven Lab are developing new artificial intelligence (AI) techniques that automate experiments, using capabilities like machine learning to drive experiments more efficiently. Looking toward the next generation of AI, Brookhaven is also creating new research facilities and capabilities that push AI beyond pure automation and towards collaboration with scientists, enabling more accurate experiments.
Machine learning is a form of AI in which computers learn by gaining experience rather than being programmed. Computers running on machine learning algorithms process and organize large datasets, automatically improving their ability to perform tasks without human intervention over time. Scientists are leveraging and developing new machine learning techniques to make discoveries in all areas of the Lab’s research, from materials science to medicine.
Automated experiments enabled by AI are increasing the rates of data collection and analysis, but scientists envision the next generation of AI systems as a more collaborative experience, integrating humans into the decision-making process and increasing experimental accuracy. In order to realize this goal of human-machine collaborations, scientists and AI systems need better access to real-time experimental information. Brookhaven is expanding its research on AI and applied math across the Lab’s divisions to enable the delivery of this information.
Brookhaven is integrating AI computing capabilities into accelerators, detectors, and sensors across the Lab’s research facilities, including the upcoming Electron-Ion Collider and other new experiments. By creating the potential for analyzing data at their source, Brookhaven aims to optimize machine performance and more efficiently steer complex experiments.