Meet Anton, a Data Science PhD student at CERN!

I first encountered CERN during my university studies, like most Swedish students, I did not even know that CERN offered student programmes.

Tell us a little about yourself and what brought you to CERN?

Heyo! I am Anton, I am 27 years old, and I am currently a doctoral student at CERN in my first year, affiliated with the Technical University of Vienna (TU Wien) in Austria. I am, however, by nationality Swedish, and graduated last year (2022) from a diploma programme in Engineering Physics and a M.Sc. in Machine Learning, from KTH Royal Institute of Technology in Stockholm, Sweden.

At this point, you might wonder what Engineering Physics is. In Sweden, Engineering Physics (sv. Teknisk Fysik) is essentially 30% maths, 30% physics, 30% engineering courses like strength of materials, fluid mechanics, etc, and then finally some programming sprinkled in there. I have therefore always found that my odd combination of skills didn’t fit into the conventional categories like physicist, software engineer, mechanical engineer, etc… However at CERN, I have found that my very specific skill set is uniquely suited for the problem that I am now trying to solve.

In fact, when I first encountered CERN during my university studies, like most Swedish students, I did not even know that CERN offered student programmes. During an excursion to CERN with ESN Lausanne during an exchange semester (2018), the guide mentioned during the introductory talk that CERN had summer student programmes. The next year (2019) I applied, I did not hear anything from recruitment for 4 months and completely forgot it. But then out of the blue, I was selected for a project in DevOps. A year later after the summer studentship, during my penultimate year in my masters (2020), I applied for, and was selected for another studentship at CERN - a technical studentship in computational physics. Lastly, by the time I was finishing up my masters (2022), I was contacted by another supervisor, without having applied for anything, for a doctoral project in data science. And here I am in 2023, unintentionally continuing my speedrun of all of the CERN student programmes.


What do you do at CERN?

In my 3 studentships at CERN, I have worked / am working with many different things, that has, and is still giving me, the opportunity to test and upgrade all my skill sets.

During my summer studentship in 2019, I was designing a CI (continuous integration) pipeline for integration tests for the DIRAC computer grid interware software that the LHCb (EP-LBC) experiment uses. Looking back now, that pipeline was a disaster, and should have not been used at all - but I learned a lot! Among others:

  • GitLab CI
  • Working in the Linux terminal, and bash scripting
  • Docker and containerised integration tests
  • Maintenance of large code bases

In my technical studentship 2020-2021, I was working for the Beams and RF Studies section (BE/SY-RF-BR), prototyping a “full stack” pipeline for doing longitudinal phase space tomography in PS Booster. In basic terms, I was implementing a method to create “x-ray pictures” of proton and ion beams in the PS Booster accelerator (the first circular accelerator in the chain that injects to the LHC), 4 times every 0.3 seconds, and logging the resulting “images” to the accelerator logging system, enabling on-line diagnostics of the beam. In this project I learned:

  • Developing and distributing Python packages, both for independent computation, leveraging CERN APIs and accelerator control middleware, and PyQt GUIs to be used in the control system.
  • Leveraging C++ and OpenMP for parallel processing of numpy array data
  • Working with the CERN Accelerator control systems for acquisition, processing, and publishing / logging data, in everything from Python to C++ and Java.
  • To present my work in a way that people from various different fields (physics, engineers, software developers) could appreciate my work

Now since February 2023, I am a doctoral student in the Data Science for Beam Operation section in the Control Software and Services group, in the Beams Department (BE-CSS-DSB). I am prototyping ways to use machine learning to do feed-forward corrections of the magnetic fields in the CERN SPS (the last accelerator in the LHC injector chain, a 7 km long particle accelerator), caused by a material phenomenon called hysteresis that occur in the magnet iron cores. This phenomenon, at present, deteriorates the quality of the particle beam in a measurable way, due to the actual magnetic fields in the accelerator not being what it is programmed to be. Specifically, I will be able to get to:

  • Apply machine learning (specifically timeseries prediction) for physics problems
  • Deploy neural networks to work in real-time in the CERN accelerator control system.
  • Work together with operations and various equipment (magnet, power converter, IT, …) groups to collect measurements, develop, test and deploy the new compensation system
  • Measure “warm” magnets, namely the SPS main dipoles, quadrupoles, sextupoles and octupoles, and process it to something that a neural network can work with

What is working at CERN like for you?

My mother visited a few months ago, and her first impression of the CERN Campus was “wow, this feels like a university campus” - and I think it is true! There are so many young people from all ages 18 and up, and you see them everywhere. Everyone is at CERN to learn, everyone ranging from job shadowers, to summer students, and all the way to senior staff. Very few know exactly what they are doing at all times - and this is kind of the point. If you know everything, it is more difficult to find new things to learn.

In fact, CERN is the perfect place for learning, and I have seen my entire time at CERN as a huge learning experience. At CERN, I do not feel the pressure so much to perform (as it is more often seen in the industry), but rather to learn, and to teach others what I have learned. Because this is what Science is. There is a vast variety of people from different nationalities, educational and cultural backgrounds, with different amounts of experience. This has given, and continues to give me the perfect opportunity to meeting like-minded people and experts from various fields to learn from them - and not only about engineering and physics!

The majority of my practical engineering, research, and presentation skills at this point in my career comes from my studentships at CERN, and I am eternally grateful, especially to my summer student supervisors who initially took a chance on me.

What advice would you give potential applicants?

CERN does way more things than particle physics! Actually, the vast majority of personnel you see at CERN are not particle physicists! CERN needs everyone, all kinds of engineers (electrical, mechanical, civil, chemical, …), software developers, physicists, administration, legal, economics, outreach, and everything in between, and especially people with interdisciplinary skills. You might not think so, but there is a decent chance that there is a project CERN that suits your profile uniquely.

Lastly, if you are still reading, I want to give you - the reader - the most important (but also the most obvious) advice of all:


Inspired? If you're a student looking for an internship opportunity to complement your studies, check out our different technical, doctoral and administrative programmes and apply!

CERN. Take part!