Behind the Scenes at CERN: An Account From ATLAS

On 4 July, 2012, the ATLAS experiment presented a preview of its updated results on the search for the Higgs Boson. Photo: CERN©

by Khoo Teng Jian

Despite the massive media coverage of the Higgs boson discovery on 4 July 2012, it is probably fair to say that the process of “discovering” a new particle remains a mystery to most. As one of the 5000 physicists nominally involved in the result from the ATLAS1 and CMS2 experiments, I hope to offer some insight into the background behind the boson buzz – a look at the lives of experimental high energy physicists.

Two activities occupy the bulk of our time: coding and communicating. Interpreting detector data requires many layers of analysis software that successively refine the “picture” taken of each collision event; writing, improving and debugging this software is the main activity of most ATLAS members. We also spend a significant amount of time reporting results to and seeking advice from one another, via e-mail, Skype, “kopitiam-style”, or in the ubiquitous meetings that are scheduled all day and every day. This description may sound more reminiscent of an IT company than of a physics lab, but a great deal of physics is embedded in all these activities.

Little needs to be said about the experience of writing analysis code, but the motivations for the large and sustained effort are worth discussing. Experimental particle physics is the industrially automated equivalent of first year university physics. Special relativistic mechanics allows us to combine the trajectories of “final state” particles observed in the detector into their “parent” precursors – something like tracing all the twinkly bits in an exploded firework to figure out how much gunpowder was in the firework to begin with – an operation that must be repeated on trillions of events. It’s also necessary to describe and correct for defects in the detector, equipment failures and other evolving conditions (the experiments will run for decades), so there’s no end to the adjustments that need to be made.

While much maligned for taking us away from “productive work” (i.e., coding), meetings can be a much more lively affair. Just as in any collaboration, the exchange of information and ideas is critical; we simply have to manage it on a larger scale. Those responsible for monitoring detector components need to report any problems that might affect the reconstruction of physics objects (electrons, muons, photons). The reconstruction experts then devise recipes to correct for these, in addition to calibrating measurements of particle energies, momenta etc. These recipes are finally passed on to the physics analysis groups, which carry out the searches and parameter measurements that are our main focus.

CMS detector that was opened during installation. Photo: CERN©

How precisely does a meeting at CERN proceed? Speakers present a great many slides and plots, while audience members all over the globe tap away on laptops (perhaps writing their own presentations). New results are announced, their consequences are discussed, and suggestions may be made for solutions to problems or cross-checks of unexpected occurrences. While the senior members of the collaboration often do the bulk of the talking, many presentations are made by fresh-faced PhD students, and anyone who wishes to may comment, establishing a very democratic environment.

Particularly important are the approval meetings for new results to be published. All members of the experiment are sworn to secrecy – collaboration information is private unless explicitly approved for release. This is less a matter of protecting trade secrets or retaining juicy information, and more to prevent misuse of information by outsiders who may not know how to interpret it correctly. Authors of a conference note or article must convince the collaboration that they have produced a solid physics result without bias or error, which can be an arduous process, taking weeks if not months. Once more, comment is open to the whole collaboration, and all concerns must be addressed before publication.

A final, though less regular, activity is taking shifts to operate the detector. During data-taking, the LHC and experiments run 24/7, squeezing every last drop of data from the circulating beams. Shift teams of a dozen physicists spend eight hour stints in the control room to ensure that the delicate detector components do not risk damage as the beams are powered up and manoeuvred into collisions. A stray beam could bore a hole through equipment worth tens of thousands of Euros. While we have no direct control over the beams, we do ensure that fragile detector components remain switched off until conditions are safe, and continue to monitor them for errors and malfunctions.

As the LHC approaches its first long shutdown, the race for results to present at the 2013 winter conferences is on its final lap. Though the machine will take a two-year break for repairs and upgrades, for the physicists there will be no rest, as data sit unanalysed, and preparations must be made for the 2015 restart. But that’s a story for another day!

About the Author:

Khoo Teng Jian is KL-born, Penang-bred, and an Old Free. A graduate of Williams College, Massachusetts, he has just completed his PhD thesis in experimental particle physics at the University of Cambridge. As a member of the ATLAS collaboration, he searches for supersymmetric particles and investigates reconstruction techniques involving invisible objects. In October, he will take up a Junior Research Fellowship at Jesus College, Cambridge. He can be contacted at [email protected]. Find out more about Teng Jian by visiting his Scientific Malaysian profile at


[1] ATLAS, standing for A Toroidal LHC Apparatus, is the biggest particle detector experiments of the Large Hadron Collider.

[2] CMS, standing for Compact Muon Solenoid, is one of the seven particle detector experiments of the Large Hadron Collider, and the heaviest one.

Leave a Reply

Your email address will not be published. Required fields are marked *

* Copy This Password *

* Type Or Paste Password Here *