Machine learning algorithms can beat the world’s hardest video games in minutes and solve complex equations faster than the collective efforts of generations of physicists. But the conventional algorithms still struggle to pick out stop signs on a busy street.
Object identification continues to hamper the field of machine learning — especially when the pictures are multidimensional and complicated, like the ones particle detectors take of collisions in high-energy physics experiments. However, a new class of neural networks is helping these models boost their pattern recognition abilities, and the technology may soon be implemented in particle physics experiments to optimize data analysis.
This summer, Fermilab physicists made an advance in their effort to embed graph neural networks into the experimental systems. Scientist Lindsey Gray updated software that allows these cutting-edge algorithms to be deployed on data from the Large Hadron Collider at CERN. For the first time, these networks will be integrated into particle physics experiments to process detector data directly — opening the flood gates for a major jump in efficiency that will yield more precise insight from current and future detectors.
“What was a week ago just an object of research is now a widely usable tool that could transform our ability to analyze data from particle physics experiments,” Gray said.
The upgraded high-granularity calorimeter — a component of the CMS detector at the Large Hadron Collider — produces complicated images of particles generated from collisions. Researchers are working to implement graph neural networks to optimize the analysis of this data to better identify and characterize particle interactions of interest. Image courtesy of Ziheng Chen, Northwestern University
His work focuses initially on using graph neural networks to analyze data from the CMS experiment at the LHC, one of the collider’s four major particle physics experiments.
Programmers develop neural networks to sift through mountains of data in search for a specific category or quantity — say, a stop sign in a photo of a crowded street.
Normal digital photographs are essentially a giant grid of red, green and blue square pixels. After being trained to recognize what a stop sign looks like, classic neural networks inspect the whole block of pixels to see whether or not the target is present. This method is inefficient, however, since the models have to process lots of irrelevant, obfuscating data.
Computer scientists have developed new classes of neural networks to improve this process, but the algorithms still struggle to identify objects in images that are more complex than just a two-dimensional grid of square pixels.
Take molecules, for example. In order to determine whether or not a chemical is toxic, chemists have to locate certain features like carbon rings and carboxyl groups within a molecule. The photographs of the chemicals taken with X-ray chromatography machines produce 3-D images of bonded atoms, which look slightly different every time they’re viewed.
Since the data are not stored in a square grid, it’s difficult for typical neural networks to learn to identify the toxic compounds. To get around this, chemists have started employing a new set of neural networks: graph neural networks, or GNNs.
“What was a week ago just an object of research is now a widely usable tool that could transform our ability to analyze data from particle physics experiments.” – Lindsey Gray
Unlike these typical neural networks, GNNs are able to tell which pixels are connected to one another even if they’re not in a 2-D grid. By making use of the “edges” between the “nodes” of data (in this case, the bonds between the atoms), these machine learning models can identify desired subjects much more efficiently.
Gray’s vision is to bring these models and their enhanced target identification to streamline data processing for particle collisions.
“With a graph neural net, you can write a significantly better pattern recognition algorithm to be used for something as complex as particle accelerator data because it has the ability to look at relationships between all the data coming in to find the most pertinent parts of that information,” he said.
Gray’s research focuses on implementing GNNs into the CMS detector’s high-granularity calorimeter, or HGCal. CMS takes billions of images of high-energy collisions every second to search for evidence of new particles.
One challenge of the calorimeter is that it collects so much data — enough pictures to fill up 20 million iPhones every second — that a large majority must be thrown away because of limitations in storage space. The HGCal’s trigger systems have to decide in a few millionths of a second which parts of the data are interesting and should be saved. The rest get deleted.
“If you have a neural network that you can optimize to run in a certain amount of time, then you can make those decisions more reliably. You don’t miss things, and you don’t keep things that you don’t really need,” said Kevin Pedro, another Fermilab scientist working with Gray.
The HGCal detectors collect lots of different information at the same time about particle interactions, which produces some very complicated images.
“These data are weirdly shaped, they have random gaps in them, and they’re not even remotely close to a contiguous grid of squares,” Gray said. “That’s where the graphs come in — because they allow you to just skip all of the meaningless stuff.”
The CMS detector at the Large Hadron Collider takes billions of images of high-energy collisions every second to search for evidence of new particles. Graph neural networks expeditiously decide which of these data to keep for further analysis. Photo: CERN
In theory, the GNNs would be trained to analyze the connections between pixels of interest and be able to predict which images should be saved and which can be deleted much more efficiently and accurately. However, because this class of neural net is so new to particle physics, it’s not yet possible to implement them directly into the trigger hardware.
The graph neural network is well-suited to the HGCal in another way: The HGCal’s modules are hexagonal, a geometry that, while not compatible with other types of neural networks, works well with GNNs.
“That’s what makes this particular project a breakthrough,” said Fermilab Chief Information Officer Liz Sexton-Kennedy. “It shows the ingenuity of Kevin and Lindsey: They worked closely to colleagues designing the calorimeter, and they put to use their unique expertise in software to further extend the capabilities of the experiment.”
Gray also managed to write a code that extends the capabilities of PyTorch, a widely used open-source machine learning framework, to allow graph neural network models to be run remotely on devices around the world.
“Prior to this, it was extremely clunky and circuitous to build a model and then deploy it,” Gray said. “Now that it’s functional, you just send off data into the service, it figures out how to best execute it, and then the output gets sent back to you.”
Gray and Pedro said they hope to have the graph neural networks functional by the time the LHC’s Run 3 resumes in 2021. This way, the models can be trained and tested before the collider’s high-luminosity upgrade, whose increased data collection capabilities will make GNNs even more valuable.
Once the networks are up and running in one place, it should be much easier to get them working in other experiments around the lab.
“You can still apply all of the same things we’re learning about graph neural networks in the HGCal to other detectors in other experiments,” Gray said. “The rate at which we’re adopting machine learning in high-energy physics is not even close to saturated yet. People will keep finding more and more ways to apply it.”
Fermilab scientific computing research is supported by the Department of Energy Office of Science.
Fermilab is supported by the Office of Science of the U.S. Department of Energy. The 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, visit science.energy.gov.
The international Deep Underground Neutrino Experiment, or DUNE, hosted by Fermilab, will be huge. In fact, with more than 1,000 collaborators from over 30 countries and five continents, it’s the largest international science project ever hosted in the U.S.
To prepare for this massive endeavor, the particle physics community has been taking DUNE technologies for thorough test drives. Over the last decade, the particle detectors ICARUS, MicroBooNE and LArIAT at Fermilab and the ProtoDUNE detectors at CERN have all contributed in one way or another to compiling the deep background knowledge needed to build and operate DUNE, a neutrino detector that will use liquid argon and advanced electronics to capture the passage of the famously elusive particles.
In 2019, DUNE preparations entered a new stage as Fermilab established a new testing facility for DUNE detectors: The Integrated Cryostat and Electronics Built for Experimental Research Goals, or ICEBERG.
Fermilab’s ICEBERG particle detector is effectively a miniature version of component that tracks neutrinos in the international Deep Underground Neutrino Experiment, hosted by Fermilab. ICEBERG researchers confirm that the detector components for DUNE reach the specifications necessary for that experiment to be successful. Photo: Reidar Hahn, Fermilab
“DUNE’s primary goal is to measure and understand very particular properties of the neutrino, and ICEBERG is a facility where we can confirm that the detector components we’re designing reach the specifications necessary for DUNE to be successful,” said Rory Fitzpatrick, a graduate student at the University of Michigan working on ICEBERG’s photon detectors.
The most abundant matter particles in the universe, neutrinos provide a valuable testing ground for particle physics theories. They hardly interact with anything, and they oscillate between three different states as they travel.
Physics experiments such as DUNE make use of the properties of neutrinos to illuminate differences between matter and antimatter, perhaps explaining why the universe seems to be dominated by matter. Neutrinos may also teach us about the proton’s lifetime and black hole formations along the way.
Fermilab accelerators will shoot an underground beam of neutrino particles 800 miles through Earth’s crust — from Fermilab in Illinois to the Sanford Underground Research Facility in South Dakota. At each site, a detector will measure the composition of the beam and analyze how the particles have shape-shifted along their flight. Since neutrinos are so weakly interacting, the detectors have to be massive and ultrasensitive. They’re essentially giant tubs of liquid argon that get bombarded with the tiny particles. Occasionally, one of the neutrinos will interact with the argon and produce charged particles and photons, both of which are detected by various sensors in DUNE. The detector in ICEBERG is in effect a miniature version of the DUNE component that tracks these particles.
There’s no need to send highly elusive neutrinos to particle detectors while simply testing the functionality of the system. When stationed above ground, detectors can also pick up traces from cosmic rays — created when high-energy particles from outer space hit the atmosphere — much more consistently.
In many ways, ICEBERG is a crystal ball for DUNE — lending insight on its future obstacles and requirements.
The cosmic-ray signatures allow physicists to test the DUNE electronics above ground with charge-tracking and photon-detection systems. Plus, because the cosmic rays are abundant on Earth’s surface and easier to detect than neutrinos, the prototypes can be smaller and require much less precious argon.
The liquid argon used for ICEBERG would fill the bed of a pickup truck. DUNE, by comparison, requires enough argon to fill 12 Olympic-sized swimming pools. DUNE researchers are currently testing the second of several combinations of new and proven electronics with ICEBERG.
“The scientists, engineers and technical staff work together to find ways to continually improve the ICEBERG and keep all its support infrastructure running,” said Kelly Hardin, a Fermilab technician who works on all liquid-argon detectors at Fermilab.
This event display shows three views of a cosmic muon interacting with liquid argon in the ICEBERG cryostat. Image: ICEBERG collaboration
Once this series of tests ends, the chosen electronics and photon sensors are expected to be tested in one of the ProtoDUNE detectors before being mass-produced for use in DUNE.
“So far, the whole ICEBERG detector and associate infrastructure are operating properly,” said Shekhar Mishra, Fermilab scientist and ICEBERG project lead. “The measurements are coming out very nice. We’ve seen beautiful tracks and detected photons.”
The process of operating and maintaining this and other prototype detectors gets scientists ready for the big league: DUNE. An international project of its magnitude requires rigorous assurance checks and thorough preparation.
“ICEBERG has given me a chance to get my hands dirty and turn some screws,” said Ivan Caro Terrazas, a graduate student at Colorado State University working on ICEBERG’s particle-tracking systems. “It amazes me how much coordination is required for a detector as small as ICEBERG, let alone DUNE itself.”
In many ways, ICEBERG is a crystal ball for DUNE — lending insight on its future obstacles and requirements.
“Even by running ICEBERG, a micro-DUNE, we’re learning a lot about what we’re going to need to build, operate and manage this massive detector,” Mishra said. “ICEBERG is a collaborative effort of laboratories and institutions around the globe. We rely on our diverse team to push through challenges and accomplish our goals.”
Collaborators work together to make the ICEBERG particle detector a successful tool for the international Deep Underground Neutrino Experiment. Left photo, taken in 2019: Reidar Hahn, Fermilab. Right photo, taken in August 2020: Kathrine Laureto
The ICEBERG program is supported by the Department of Energy Office of Science.
Fermilab is supported by the Office of Science of the U.S. Department of Energy. The 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, visit science.energy.gov.