CERN and Fermilab announce big step in Deep Underground Neutrino Experiment
The largest liquid-argon neutrino detector in the world has just recorded its first particle tracks, signaling the start of a new chapter in the story of the international Deep Underground Neutrino Experiment (DUNE).
DUNE’s scientific mission is dedicated to unlocking the mysteries of neutrinos, the most abundant (and most mysterious) matter particles in the universe. Neutrinos are all around us, but we know very little about them. Scientists on the DUNE collaboration think that neutrinos may help answer one of the most pressing questions in physics: why we live in a universe dominated by matter. In other words, why we are here at all.
The enormous ProtoDUNE detector — the size of a three-story house and the shape of a gigantic cube — was built at CERN, the European laboratory for particle physics, as the first of two prototypes for what will be a much, much larger detector for the DUNE project, hosted by the U.S. Department of Energy’s Fermi National Accelerator Laboratory in the United States. When the first DUNE detector modules record data in 2026, they will each be 20 times larger than these prototypes. There will be four modules in total.
This image shows one of the first cosmic muon particle tracks recorded by the ProtoDUNE detector at CERN. Three wire planes, each of which is made up of thousands of individual wires, recorded the signal of the muon as it traveled approximately 3.8 meters through liquid argon in the detector, and the images together give scientists a three-dimensional picture of the particle’s path. Image: DUNE collaboration
It is the first time CERN is investing in infrastructure and detector development for a particle physics project in the United States.
The first ProtoDUNE detector took two years to build and eight weeks to fill with 800 tons of liquid argon, which needs to be kept at temperatures below minus 184 degrees Celsius (minus 300 degrees Fahrenheit). The detector records traces of particles in that argon both from cosmic rays and a beam created at CERN’s accelerator complex. Now that the first tracks have been seen, scientists will operate the detector over the next several months to test the technology in depth.
“Only two years ago we completed the new building at CERN to house two large-scale prototype detectors that form the building blocks for DUNE,” said Marzio Nessi, head of the Neutrino Platform at CERN. “Now we have the first detector taking beautiful data, and the second detector, which uses a different approach to liquid-argon technology, will be online in a few months.”
The technology of the first ProtoDUNE detector will be the same to be used for the first of the DUNE detector modules in the United States, which will be built a mile underground at the Sanford Underground Research Facility in South Dakota. More than 1,000 scientists and engineers from 32 countries spanning five continents — Africa, Asia, Europe, North America and South America — are working on the development, design and construction of the DUNE detectors. The groundbreaking ceremony for the caverns that will house the experiment was held in July 2017.
Inside the first ProtoDUNE detector, before it was filled with liquid argon. Photo: CERN
“Seeing the first particle tracks is a major success for the entire DUNE collaboration,” said DUNE co-spokesperson Stefan Soldner-Rembold of the University of Manchester, UK. “DUNE is the largest collaboration of scientists working on neutrino research in the world, with the intention of creating a cutting-edge experiment that could change the way we see the universe.”
When neutrinos enter the detectors and smash into the argon nuclei, they produce charged particles. Those particles leave ionization traces in the liquid, which can be seen by sophisticated tracking systems able to create three-dimensional pictures of otherwise invisible subatomic processes. (Watch an animation of how the DUNE and ProtoDUNE detectors work, along with other videos about DUNE.)
“CERN is proud of the success of the Neutrino Platform and enthusiastic about being a partner in DUNE, together with institutions and universities from its member states and beyond,” said Fabiola Gianotti, director-general of CERN. “These first results from ProtoDUNE are a nice example of what can be achieved when laboratories across the world collaborate. Research with DUNE is complementary to research carried out by the LHC and other experiments at CERN; together they hold great potential to answer some of the outstanding questions in particle physics today.”
The steel cage for one of the two ProtoDUNE detectors is outfitted with a steel top, hoisted into position by a crane. Photo: CERN
DUNE will not only study neutrinos, but their antimatter counterparts as well. Scientists will look for differences in behavior between neutrinos and antineutrinos, which could give us clues as to why the visible universe is dominated by matter. DUNE will also watch for neutrinos produced when a star explodes, which could reveal the formation of neutron stars and black holes, and will investigate whether protons live forever or eventually decay. Observing proton decay would bring us closer to fulfilling Einstein’s dream of a grand unified theory.
“DUNE is the future of neutrino research,” said Fermilab Director Nigel Lockyer. “Fermilab is excited to host an international experiment with such vast potential for new discoveries and to continue our long partnership with CERN, on both the DUNE project and the Large Hadron Collider.”
To learn more about the Deep Underground Neutrino Experiment, the Long-Baseline Neutrino Facility that will house the experiment, and the PIP-II particle accelerator project at Fermilab that will power the neutrino beam for the experiment, visit www.fnal.gov/dune.
CERN, the European Organization for Nuclear Research, is one of the world’s leading laboratories for particle physics. The organization is located on the French-Swiss border, with its headquarters in Geneva. Its member states are: Austria, Belgium, Bulgaria, Czech Republic, Denmark, Finland, France, Germany, Greece, Hungary, Israel, Italy, Netherlands, Norway, Poland, Portugal, Romania, Slovakia, Spain, Sweden, Switzerland and United Kingdom. Cyprus, Serbia and Slovenia are associate member states in the pre-stage to membership. India, Lithuania, Pakistan, Turkey and Ukraine are associate member states. The European Union, Japan, JINR, the Russian Federation, UNESCO and the United States of America currently have observer status.
Fermilab is America’s premier national laboratory for particle physics and accelerator research. A U.S. Department of Energy Office of Science laboratory, Fermilab is located near Chicago, Illinois, and operated under contract by the Fermi Research Alliance LLC, a joint partnership between the University of Chicago and the Universities Research Association, Inc. Visit Fermilab’s website at www.fnal.gov and follow us on Twitter at @Fermilab.
DOE’s Office of Science is the single largest supporter of basic research in the physical sciences in the United States and is working to address some of the most pressing challenges of our time. For more information, please visit science.energy.gov.
While his official title might be Joel Butler, Distinguished Scientist, his tenure as the CMS spokesperson required a very different skill set than most people associate with research.
“At Fermilab, I was a famous nondelegator,” Butler said. “I liked to poke my nose into everything. But that doesn’t work when you’re overseeing a collaboration with more than 3,000 members.”
Butler started working at Fermilab in 1979 and quickly found himself managing large projects and departments. But according to Butler, nothing compares to his time at CERN as the CMS spokesperson for the last two years.
“It’s an experience very few people ever have,” said Butler. “I’ve run experiments before, but the magnitude of diversity, talent and energy on CMS is spectacular. There’s nothing like it.”
This week Butler prepares to return to Fermilab. He hands the reins of the CMS experiment over to his successor, Roberto Carlin, as well as to incoming deputy co-spokespersons Patty McBride of Fermilab and Luca Malgeri of CERN. He reflects on his time as the CMS spokesperson, lessons learned and major achievements.
***
What do the day-to-day activities look like for the CMS spokesperson?
The job of the CMS spokesperson is to keep everything going and make sure problems are solved. Every week we would go through all the detector systems to find out what the problems were and make sure we were on track to resolving them. Once a week I would also inform the collaboration with status updates and upcoming events, such as special LHC runs. That’s the job: communication with collaborators, listening to collaborators, problem solving and a lot of meetings. The CMS collaboration consists of many countries and funding agencies, and they all have different projects and priorities. The job of the spokesperson is to bring everyone together and make sure that everyone is doing the best work they can do.
What were some of the challenges that come along with being the spokesperson of a large international collaboration with 3,000-plus members?
The biggest problems were not conflicts, but rather communicating complicated ideas. The CMS experiment has a tremendous number of smart people with good arguments for what they want to do and why they want to do it. But just because someone is talking, doesn’t mean that they’re communicating. Sometimes people would leave meetings without a real understanding of what we had decided. Sometimes I’d say something extreme during a meeting just to see if people were listening. Sometimes silence is taken as consent, but other times, it’s just a lack of attention. So I started doing this thing called affirmative consensus, where silence was not consent and people had to state that they agreed.
I heard that you had “office hours” in the atrium of CERN’s Building 40 every morning. Why did you decide to do that, and what was it like?
Not exactly office hours, I just hung out there around breakfast time and hoped people would come by and talk, and often they did. The problem is that as things move up the chain, certain kinds of information get diluted, and by the time it reaches the spokesperson, you don’t hear what you need to hear. People have a tendency to filter information so that only urgent matters propagate, and persistent problems and annoyances don’t move up to management. Transparency is really important in a collaboration like this, because if we don’t know what the problems are, we cannot work toward solving them. I wanted to stress that people should not be shy about discussing their problems with me, which is why I kept my door open and tried to make myself as accessible as possible. If we put the problems out there, other people can engage and help solve them.
This was your first time managing a large international collaboration. How was it different than the projects you’ve overseen in the past?
When I was the U.S. CMS program manager, from 2007 to 2013, I worked with all the U.S. universities on CMS, which is about 30 percent of the collaboration. Because it’s the same funding environment and the university systems are similar, there was a lot of commonality and overlap in world view. Now as the CMS spokesperson, take that job and add another 48 nations, which are all really different. With U.S. universities, some are very large, and some very small. But internationally, there are even bigger differences. Many international partners have been in physics for decades, while others are just starting to get into the field. A big goal of CMS and CERN is to help develop powerful new collaborators around the world and bring the ability to do science to more and more people. The world has got smart people everywhere, but some places don’t have the pre-existing infrastructure to support them. So when it comes to working with international partners, they can have many of the same problems we see in the United States, but amplified, and they can have completely new ones. At the same time, it’s also very exciting. In an international collaboration, you get new perspectives on problems and can draw on knowledge and experience from all over the world.
What are some of the challenges CMS faced over the last two years?
While we’re always making discoveries, learning new things, and progressing the field, particle physics isn’t a discipline where we’re going to completely revolutionize our understanding of the universe every day. Progress is steady, but it often takes 10 or 20 years to go from one major breakthrough to the next. There are lots of false trails and dead ends you need to explore before you eventually hit the right path. Expectations tend to be higher than justified. Just because progress is high doesn’t mean that you’ll immediately find new physics.
What are some of the major accomplishments of CMS and the LHC over the last two years?
We’ve ruled out many theories for new physics and really enhanced our understanding of the Standard Model. We made measurements of the Higgs boson that, a few years ago, we could only have dreamed of making. It shows that patience pays off. We have a magnificent opportunity to explore the subatomic laws of nature with this machine that works so well. It’s astounding, really! I’ve been around long enough to remember when turning on an accelerator could be a matter of years. But the LHC is just fabulous. We’ve been given this miraculous device, and our responsibility is to make the most of it. Stay focused, and keep going. A breakthrough could happen at any moment or could take years.
What are you doing next?
I’m now the deputy head of a CMS upgrade that will give us precision timing of particles as they pass through the detector. This can be a tremendous asset because it will help us reconstruct events and better understand what happened during the collisions. That assignment should last around two years, and after that, I’ll see where I am. I’m near the end of my career and thinking about retirement, but I’ll never stop working. This is too much fun. I remember there was this saying from Confucius that went, “If you love what you do, you’ll never work a day in your life.” It’s true!
It is hard these days not to encounter examples of machine learning out in the world. Chances are, if your phone unlocks using facial recognition or if you’re using voice commands to control your phone, you are likely using machine learning algorithms — in particular deep neural networks.
What makes these algorithms so powerful is that they learn relationships between high-level concepts we wish to find in an image (faces) or sound wave (words) with sets of low-level patterns (lines, shapes, colors, textures, individual sounds), which represent them in the data. Furthermore, these low-level patterns and relationships do not have to be conceived of or hand-designed by humans, but instead are learned directly from examples of the data. Not having to come up with new patterns to find for each new problem is the reason deep neural networks have been able to advance the state of the art for so many different types of problems: from analyzing video for self-driving cars to assisting robots in learning how to manipulate objects.
Here at Fermilab, there has been a lot of effort in having these deep neural networks help us analyze the data from our particle detectors so that we can more quickly and effectively use it to look for new physics. These applications are a continuation of the high-energy physics community’s long history in adopting and furthering the use of machine learning algorithms.
Recently, the MicroBooNE neutrino experiment published a paper describing how they used convolutional neural networks — a particular type of deep neural network — to sort individual pixels coming from images made by a particular type of detector known as a liquid-argon time projection (LArTPC) chamber. The experiment designed a convolutional neural network called U-ResNet to distinguish between two types of pixels: those that were a part of a track-like particle trajectory from those that were a part of a shower-like particle trajectory.
This plot shows a comparison of U-ResNet performance on data and simulation, where the true pixel labels are provided by a physicist. The sample used is 100 events that contain a charged-current neutrino interaction candidate with neutral pions produced at the event vertex. The horizontal axis shows the fraction of pixels where the prediction by U-ResNet differed from the labels for each event. The error bars indicate only a statistical uncertainty.
Track-like trajectories, made by particles such as a muon or proton, consist of a line with small curvature. Shower-like trajectories, produced by particles such as an electron or photon, are more complex topological features with many branching trajectories. This distinction is important because separating these types of topologies can be difficult for traditional algorithms. Not only that, shower-like shapes are produced when electrons and photons interact in the detector, and these two particles are often an important signal or background in physics analyses.
MicroBooNE researchers demonstrated that these networks not only performed well but also worked in a similar fashion when presented with simulated data and real data. The latter is the first time this has been demonstrated for data from LArTPCs.
Showing that networks behave the same on simulated and real data is critical, because these networks are typically trained on simulated data. Recall that these networks learn by looking at many examples. In industry, gathering large “training” data sets is an arduous and expensive task. However, particle physicists have a secret weapon — they can create as much simulated data as they want, since all experiments produce a highly detailed model of their detectors and data acquisition systems in order to produce as faithful a representation of the data as possible.
However, these models are never perfect. And so a big question was, “Is the simulated data close enough to the real data to properly train these neural networks?” The way MicroBooNE answered this question is by performing a Turing test that compares the performance of the network to that of a physicist. They demonstrated that the accuracy of the human was similar to the machine when labeling simulated data, for which an absolute accuracy can be defined. They then compared the labels for real data. Here the disagreement between labels was low, and similar between machine and human. (See the top figure. See the figure below for an example of how a human and computer labeled the same data event.) In addition, a number of qualitative studies looked at the correlation between manipulations of the image and the label provided by the network. They showed that the correlations follow human-like intuitions. For example, as a line segment gets shorter, the network becomes less confident if the segment is due to a track or a shower. This suggests that the low-level correlations being used are the same physically motivated correlations a physicist would use if engineering an algorithm by hand.
This example image shows a charged-current neutrino interaction with decay gamma rays from a neutral pion (left). The label image (middle) is shown with the output of U-ResNet (right) where track and shower pixels are shown in yellow and cyan color respectively.
Demonstrating this simulated-versus-real data milestone is important because convolutional neural networks are valuable to current and future neutrino experiments that will use LArTPCs. This track-shower labeling is currently being employed in upcoming MicroBooNE analyses. Furthermore, for the upcoming Deep Underground Neutrino Experiment (DUNE), convolutional neural networks are showing much promise toward having the performance necessary to achieve DUNE’s physics goals, such as the measurement of CP violation, a possible explanation of the asymmetry in the presence of matter and antimatter in the current universe. The more demonstrations there are that these algorithms work on real LArTPC data, the more confidence the community can have that convolutional neural networks will help us learn about the properties of the neutrino and the fundamental laws of nature once DUNE begins to take data.
Victor Genty, Kazuhiro Terao and Taritree Wongjirad are three of the scientists who analyzed this result. Victor Genty is a graduate student at Columbia University. Kazuhiro Terao is a physicist at SLAC National Accelerator Laboratory. Taritree Wongjirad is an assistant professor at Tufts University.