CERN operates some of the most complex scientific machinery ever built, relying on intricate control systems and generating petabyte upon petabyte of research data. Operating this equipment and performing analysis on the data gathered are both intensive tasks. Therefore, CERN is increasingly looking to the broad domain of artificial-intelligence (AI) research to address some of the challenges encountered  in dealing with data, particle beam handling, and in the up keep of its facilities.

The modern domain of artificial intelligence came into existence around the same time that CERN did, in the mid-’50s. It has different meanings in different contexts, and CERN is mainly interested in task-oriented, so-called restricted AI, rather than general AI involving aspects such as independent problem-solving, or even artificial consciousness. Particle physicists were among the first groups to use AI techniques in their work, adopting Machine Learning (ML) as far back as 1990.  Beyond ML, physicists at CERN are also interested in the use of Deep Learning to analyse the data deluge from the LHC.

Dealing with a data deluge

Even before the Large Hadron Collider began colliding high-energy beams of protons in 2010, the particle-physics community began to collect unprecedented quantities of data. Particles collide within the LHC’s detectors up to 40 million times a second, each collision event generating about a megabyte of data: far too much to store without some filtering.

Not only do the scientists have to programme their data acquisition systems to select the right events for further analysis while discarding the uninteresting data, they also have to examine trillions of stored collision events looking for signatures of rare physics phenomena. They have therefore turned to one sub-domain of AI, called machine learning (ML), to improve the efficiency and efficacy of these tasks. In fact, the LHC’s four major collaborations ‑ ALICE, ATLAS, CMS and LHCb ‑ have formed the Inter-experimental Machine Learning (IML) Working Group to follow developing trends in ML. Researchers are also collaborating with the wider data-science community to organise workshops to train the next generation of scientists in the use of these tools, and to produce original research in Deep Learning. ROOT, the software program developed by CERN and used by physicists around the world for analysing their data, also comes with machine-learning libraries.

Operating in extreme environments

Experimental facilities at CERN may be temporarily classified as high-radiation zones, preventing human intervention to perform repairs or to replace equipment. CERN has therefore developed autonomous robots to operate in these zones, which include the tunnel containing the LHC. The Engineering department at CERN, which builds and maintains these robots, uses AI techniques to help the robots navigate on their own and make decisions on what actions to take inside the radiation environments.

In addition to using AI robotics for the maintenance of its complex machines, predicting component failure and for safety applications, CERN has recognised the importance of involving external AI expertise in projects undertaken at the laboratory.

Much of the collaboration with these experts is through CERN openlab, a public-private partnership that enables CERN to work with world-leading researchers and companies working on AI. CERN openlab has launched several machine-learning projects ranging from improving industrial control systems to simulating the conditions inside particle detectors following high-energy collisions to investigations of future Quantum Machine Learning (QML) algorithms. CERN openlab also contributes to the use of CERN’s AI resources for humanitarian operations and recently organised a conference on the ethics associated with our increasing use of AI in the world.