Kevin Einsweiler is a senior scientist at Lawrence Berkeley National Lab (LBNL). He joined the ATLAS Collaboration in 1993, playing an instrumental role in bringing US institutes into the LHC programme. He served as ATLAS Pixel Project Leader (2005-2009), Physics Coordinator (2011-2013) and Upgrade Coordinator (2014-2019). In this interview, Kevin shares how he became a particle physicist, discusses the formation of the LHC collaborations, and looks ahead to the next era of the ATLAS experiment

What sparked your first interest in physics? Was there a particular event or inspiring teacher?

I suppose my physics "origin story" is a little different, because it started in a car instead of a classroom. Because of a late birthday and a skipped year, I started university at 16 and lived at my parent’s home as an undergraduate. This meant that, every day, I would get a ride to the university with a neighbour (and father of my best friend) who was a physics professor. His name was Hans Courant and he led an incredibly interesting life; he came from a very prolific family of physicists and had worked on the Manhattan Project and at CERN. We would spend the drives having these fascinating conversations about particle physics. One day, during my second year in university, he invited me to visit Argonne National Lab to see an experiment he was working on (this was summer 1975). It was a small fixed-target experiment using polarised protons. This was not an experiment out of a textbook, but I got to wire-wrap some of the data acquisition (DAQ) system and write PDP assembly language programs. I was completely hooked.

He sounds like an incredible influence. How did this initial interest transition into a PhD in experimental physics?

It wasn’t immediate. I knew I was fascinated by physics and wanted to study it, but my undergraduate studies originally pushed me towards wanting to be a theorist. I think that’s pretty common for students: you spend all of your time learning concepts and solving simple physics problems, and end up imagining that could be a career.

While I really loved the mathematics behind quantum mechanics – and still find it incredibly beautiful – I also knew I loved building stuff. Perhaps it runs in the family: my grandfather was a construction contractor, my father and brother are urban planners. There was always work going on at home that I, even at a young age, loved getting involved in. Even so, it took me a while to reconcile that the hands-on side of physics was what I enjoyed most.

Luckily, I had this realisation before picking a graduate school. I decided on Stanford for my Ph.D, specifically because they had an accelerator and were launching new experiments. That gave me a chance to get in on an experiment on the ground level – designing, prototyping, building, operating and then doing my thesis on the data.

That’s something a lot of ATLAS people haven’t experienced – seeing an experiment through from start to finish.

That's right. It’s a multi-decade process now and involves hundreds, if not thousands, of people. My thesis experiment (SPEAR Mark III) had maybe 30 people on it, and only a half a dozen institutes. So I was able to work on a huge range of things – working with multiple kinds of electronics, writing the data acquisition system, and developing tracking chambers and calorimeters. It was really exciting and a great environment to start off in.

After graduating in 1984, you came to CERN to join the UA2 experiment on CERN’s Super Proton Synchrotron (SPS). How did that come about and how did you then transition into the proto-collaborations for the LHC?

There was a funny sort of a pipeline between SLAC and UA2 in those days – and part of that pipeline was Peter Jenni [UA2 physicist who went on to be the first ATLAS spokesperson]. I got to know Peter in 1979, when he was at SLAC working on the Mark II experiment. He was a more senior physicist, and I was a young graduate student. The Mark II collaboration had many strong physicists and engineers, several of whom were people I considered as mentors. That group had a lot of connections to the UA2 team and it was common knowledge that if you wanted to become an expert in instrumentation, you absolutely had to go to CERN and work on UA2. So that’s what many of us did.

UA2 was really a great experiment and you can still see the influence it had on ATLAS today. Not just in the detector design, but also in establishing our community. It was where a lot of ATLAS people started their careers and we got used to its very congenial environment. There were never any barriers between young people and senior people – a lot of great physicists, but no major egos. We all ended up forming close personal and professional relationships.

These ties carried over to the first LHC proto-collaborations. When you have a group of people with a proven track record of working well together, of course you want to build your collaboration around them. I also think there was a desire to carry forward the structure and spirit we’d fostered at UA2. Other LHC collaborations were under consideration at the time, which would have had very rigid, top-down hierarchies. That way of doing physics is just completely anathema to me; it leads to decisions being made that have nothing to do with physics. Luckily, the LHC collaborations avoided taking that route.

That said, I missed the early days of ATLAS’ formation. I left CERN in early 1990 to work at LBNL, and totally immersed myself in work on the Superconducting Super Collider (SSC) by joining the SDC experiment, becoming the Physics Coordinator for several years, and producing the physics part of the Technical Design Report (TDR) in 1992. There wasn’t a lot of belief in the LHC programme from many US scientists at the time. It was thought, rather naively, that because the SSC had been under development for so long and used more mature magnets and much larger rings than the LHC, that it was definitely the best way to go. Its sudden cancellation was one of those traumatic experiences that completely changes everything you do afterwards. It was an unbelievable situation. Hundreds of our colleagues got their pink slips, all at the same time. So many people had to leave physics for good.

How did the field adapt to this major cancellation? Did it impact the way you approached projects?

It really did change everything. Those of us still employed had to completely reorient our dreams around the LHC. It was the only possible path forward into the TeV range and, for the US, it was still far from a sure thing. In that period, my whole professional life revolved around ensuring that the US could participate fully in the LHC, and that the LHC and ATLAS actually got built. This intense dedication was common at the time, because this was a life-or-death situation for our field as seen from the US side. We referred to ourselves as SSC refugees, looking for a safe harbour.

In the Berkeley group, we chose to look forward and organised initial discussions between our group (George Trilling, Murdock Gilchriese, and myself) to interview both ATLAS and CMS. During that process, we were encouraged by Peter Jenni to use our silicon-detector expertise in the inner detector area, but told that joining ATLAS meant joining the international team – detectors would not be “owned” by national groups. After our experience with losing SDC, we were enthusiastic about this approach. In discussions with CMS, the dialog was oriented towards “owning” parts of the detector that would fit national budgets. We chose ATLAS, and never looked back!

From my perspective, we urgently needed to define how the US would become a collaborator in the LHC programme. What would be the scale of involvement? How much money would be involved? Could we even get our colleagues and funding agencies on board with the idea of joining the Large Hadron Collider project at CERN?

I had the good fortune to serve on a subpanel of the High Energy Physics Advisory Panel (HEPAP) that was formed to address the issues above. This was the so-called “Drell panel”, chaired by the SLAC theorist Sid Drell, formed in 1994. This panel was unusually well-connected to high levels in the US government. As a young scientist, I focused on physics and technical issues. I had been serving on the CERN LHC Experiments Committee (LHCC) from its start in 1992, and was part of the technical management of SDC. A key point of contention was the LHC’s lower energy that required more intense beams to explore the TeV scale. There were many sceptics who thought it was impossible to do serious physics research with more than ~1 hadron collision at a time. Yet the LHC required more than 20 hadron collisions per beam crossing to achieve its physics goals. That’s hard to imagine nowadays, given the massive number of collisions we see per LHC bunch crossing, but it was still unclear to many in the early 90s. I felt like a big part of my job on the sub-panel was to convince my colleagues of the feasibility of operating in this very high pileup regime. Our panel formally endorsed the US role in the LHC in early 1995, which was a major step forward from the SSC catastrophe.

What was your first role in the ATLAS Collaboration?

I began work on electronics for the ATLAS Pixel detector in 1994, prototyping ideas for front-end designs and readout architectures. Little did I know this would be the start of my long odyssey into beyond-state-of-the-art silicon technology. Over the subsequent 15 years, I saw the Pixel detector go from a pioneering concept to a successful prototype and finally into a working instrument inside the ATLAS experiment. The experience really cemented my belief in building ambitious hardware.

In 2005, I was elected Pixel Project Leader and moved back to CERN with my family. We were facing several new technical and resource challenges at the time – all of which had to be resolved before the looming LHC start-up deadline. We successfully installed the Pixel detector in June 2007, just in time. The Endcap Toroids were to be installed in July 2007 and, if the Pixel installation had fallen behind schedule, we would have risked not being ready for first collisions. The Pixel detector was ready for commissioning in 2008. This final push over the finish line was only possible thanks to the Pixel team’s tireless efforts.

With the Pixel detector complete, you soon transitioned to data analysis – becoming ATLAS’ Standard Model convener in 2009 and Physics Coordinator in 2011. What were the main physics priorities in those early days?

As soon as we began recording data, we launched a strategy we dubbed “re-observing the Standard Model”. The first step was to actually understand our signals, learning how to reconstruct/identify/measure tracks, jets, photons, leptons, missing energy, etc. in the ATLAS data. In parallel, we were making the first reference measurements of Standard Model particles. Our group published the very first ATLAS physics analysis paper, on minimum bias events, measuring basic distributions of track parameters. By the end of 2011, we had an unprecedentedly detailed picture of the Standard Model in a new energy regime. Moving to the Physics Coordinator level was an enormous step, and the lead up to the Higgs observation was a unique lifetime experience – this story has been told many times already. I consider Physics Coordinator to be the best and most exciting job in ATLAS, and am grateful for the opportunity I had to be in the right place at the right time.

What amazed me was the change in the group dynamics at this time. We’d spent years preparing a lot of these analyses and, when we had no data, that work was driven by the most senior people. Once we had data, all of a sudden the plots started coming from completely different corners of the community. Fellows, graduate students and postdocs left all the senior people in the dust – myself included – producing remarkably high-quality physics results.

I think it had a massive impact on their confidence in their future careers. Because even though they were part of a huge collaboration, they were very empowered. It was an environment where they had to do a lot of thinking on their feet, constantly inventing new tricks. There was no playbook to follow. If they found that a bit of software wasn't working, for example, it was all on them to come up with ways to get around the problem. From my side, it was really exciting to watch these young people grow their skills and absolutely thrive.

You moved back into detector development in 2013, first working on the Insertable B-layer sub-detector and then focussing on ATLAS upgrades for the High-Luminosity LHC (HL-LHC). What inspired the move?

I just really believed in the HL-LHC programme and wanted it to succeed. By that time, I had collected a unique assortment of skills both in physics and detector design, and there was a lot of work to be done on the upgrades. Although there had been plenty of discussions, I felt the Collaboration hadn’t made enough progress in turning our ideas into concrete projects. What would the final designs be? How many people would we need to build these upgrades? How much money would it cost? Could we actually make the things we were talking about? And then, could we go out to funding agencies to convince them?

It was clear that we needed to take a fresh look at our HL-LHC upgrade plans and ask some hard questions. At the time, my main concern was: were we being sufficiently ambitious? This was also echoed by the LHCC reviewers of our initial Letter of Intent for the upgrades.

That’s interesting – because typically the feedback experimentalists get is that they are being too ambitious.

That’s true, but it’s important to also match the scale of the project. Data-taking in HL-LHC will generate around 95% of ATLAS’ lifetime dataset – clearly the upgraded detector needs to be the best it can possibly be.

In 2014, we developed a weekly task force that would explore some outside-the-box ideas. In particular, I thought that we could really improve our forward detectors in order to study vector boson fusion in detail. This led to the design of the new Inner Tracking detector, which will be unlike anything that has been built for a hadron collider. We also explored extensions to the detector’s magnetic field coverage, improving how we track muons, and identified precision timing as another key point and led to the development of the High-Granularity Timing Detector. Not all of the ideas made the cut into the final HL-LHC detector, but many did, helping to ensure ATLAS would continue to lead in the coming decades.

I was certainly among those pushing for ambitious designs, wanting to improve the ATLAS detector as much as we possibly could. ATLAS is a huge collaboration of extremely talented people – I believe we can do the work. Now, some elements of the upgrades may end up being beyond our reach. But I didn’t want the field to look back on these decisions, 15 years from now, and say “why didn’t they do more?” We still don’t know what will come next after the HL-LHC. We need to take full advantage of the present HL-LHC opportunities!

You ended up becoming ATLAS Upgrade Coordinator in October 2014. How did you set out to put these ambitious designs on schedule for HL-LHC?

Within two weeks of my election, and before my term started, we launched a new “scoping document” process, jointly with CMS and the CERN Director of Research. This was a formal procedure to evaluate the cost and feasibility of the upgrades. We wanted to put forward a range of options for different costs and ambitions: low, medium and high. This would then be used to define the final scope of the upgrade project. It was not yet clear how much an HL-LHC detector should cost, and there is a tendency in the field to define the scope of a project by the budget available. That’s really not the right approach for a project as large as this one. This scoping process gave us a chance to explore several options, balancing cost against the possible physics performance. By examining three scenarios, we could make a very clear and compelling case for the most expensive option. This process took about a year, and by September 2015 we had a very clear idea of what we would build.

This set us up well for the next stage: creating the upgrade TDRs. I am very proud that, despite several challenges, we completed and published all six upgrade TDRs by the end of 2017. We were then able to formally submit the Memorandums of Understanding to funding agencies in 2019. This was an extremely quick turnover, considering the number of R&D, performance and physics studies each report required. The ATLAS upgrade community is extremely dedicated, and I have every confidence in their success going forward.

What do you think are the key challenges for ATLAS going forward?

I think Run 3 will be a critical phase for the ATLAS Collaboration; there’s a lot of work we have to do in parallel. A lot of dedication will be needed to construct all of our new detectors. This work will be an investment in the future, as the HL-LHC is the most important project of the global accelerator-based particle physics programme. That said, the start of the HL-LHC is still a long way off. We need to keep up the pace of our data analysis, as the future of our field depends on the success of our current physics programme.

ATLAS Portraits is a series of interviews presenting collaborators whose contributions have helped shape the ATLAS experiment. Discover more ATLAS Portraits here.