Imagine a world where experiments can run faster, smarter, and even partly on their own. That’s exactly what a new UKRI-STFC project is aiming to do—by using artificial intelligence (AI) to help run experiments and make sense of huge amounts of scientific data.

Led by STFC’s Ada Lovelace Centre (ALC), the Autonomous Laboratories project brings together experts from STFC Scientific Computing, Diamond Light Source, and the Physical Sciences Data Infrastructure Programme (PSDI). With almost £125,000 of funding from EPSRC’s AI for Science grant, the research team is working on ways to use AI to change how experiments are performed, sorting information and making smart decisions during experiments and improve how scientists analyse results.

Today’s research facilities produce massive amounts of data, and it’s becoming harder for scientists to keep up. The new project tackles this challenge head-on. By using AI, researchers can process data more effectively and make decisions during experiments, rather than waiting until the experiment is finished.

One key focus is a technique called diffraction, which helps scientists study the structure of materials. Because it’s used in many different experiments, it’s the perfect test case. Success here could benefit a wide range of UK research centres, including national facilities like Diamond Light Source.

The project isn’t just about making things faster—it’s also about making science smarter. AI can help spot patterns in data, guide experiments in real time, and produce high-quality datasets that are easier for other scientists to use. This will make research more reliable and easier to repeat, which is essential for scientific progress.

There are also clear practical benefits. By improving efficiency, the new approach could improve throughput, cut costs and provide new ways for industry to quickly access advanced measurements. It may even allow more experiments to be carried out remotely, meaning scientists won’t always need to travel to the lab.

ALC director, Dr Paul Quinn, said, “Ultimately, the project is a testbed for the future of discovery science. By combining advanced AI tools with cutting-edge research facilities, it aims to show how ‘smart labs’ could become the norm. If successful, it will help keep the UK at the forefront of scientific innovation—and change how experiments are done for years to come.”

ALC and Scientific Computing will continue the work with Diamond to develop a new autonomous system, with a view to becoming production ready in the near future. PSDI has led a related project to produce a prototype for AI-ready data pipelines and repositories, ensuring stored data becomes usable data for AI systems.