The present methods of rebuilding particle tracks will soon no longer suffice.AI in Particle TrackingResearch presented in the journal Computer Science by scientists from the Institute of Nuclear Physics of the Polish Academy of Sciences (IFJ PAN) in Cracow, Poland, recommends that tools built utilizing synthetic intelligence could be an effective option to existing methods for the quick restoration of particle tracks. Their debut could happen in the next two to 3 years, probably in the MUonE experiment which supports the search for brand-new physics.The concept of reconstructing the tracks of secondary particles based on hits recorded during crashes inside the MUonE detector. Credit: IFJ PANThe Complexity of Particle DetectionIn contemporary high-energy physics experiments, particles diverging from the collision point pass through succeeding layers of the detector, transferring a little energy in each. Charged particles move in it along curved lines and this is likewise how the detector aspects triggered by them, which in our jargon we call hits, will be located with regard to each other,” discusses Prof. Marcin Kucharczyk, (IFJ PAN) and right away adds: “In truth, the so-called occupancy of the detector, i.e. the number of hits per detector component, might be really high, which causes lots of problems when attempting to reconstruct the tracks of particles properly.”Experiments designed to discover new physics will clash particles at higher energies than previously, suggesting that more secondary particles will be developed in each accident.
By The Henryk Niewodniczanski Institute of Nuclear Physics, Polish Academy of Sciences March 24, 2024AI is emerging as a key tool in nuclear physics, providing options for the intricate and data-intensive job of particle track restoration. Credit: SciTechDaily.comParticles colliding in accelerators produce many cascades of secondary particles. The electronic devices processing the signals avalanching in from the detectors then have a portion of a 2nd in which to examine whether an event is of enough interest to wait for later analysis. In the near future, this demanding task may be performed utilizing algorithms based on AI.Electronics has never ever had an easy life in nuclear physics. There is a lot data being available in from the LHC, the most effective accelerator in the world, that taping all of it has never been an option. The systems that process the wave of signals originating from the detectors therefore focus on … forgetting– they rebuild the tracks of secondary particles in a fraction of a second and assess whether the crash simply observed can be overlooked or whether it deserves saving for more analysis. The existing techniques of reconstructing particle tracks will quickly no longer suffice.AI in Particle TrackingResearch provided in the journal Computer Science by researchers from the Institute of Nuclear Physics of the Polish Academy of Sciences (IFJ PAN) in Cracow, Poland, recommends that tools built utilizing synthetic intelligence might be a reliable alternative to current approaches for the fast reconstruction of particle tracks. Their debut might take place in the next 2 to 3 years, probably in the MUonE experiment which supports the look for brand-new physics.The concept of reconstructing the tracks of secondary particles based upon hits recorded throughout crashes inside the MUonE detector. Subsequent targets are marked in gold, and silicon detector layers are marked in blue. Credit: IFJ PANThe Complexity of Particle DetectionIn modern high-energy physics experiments, particles diverging from the crash point travel through successive layers of the detector, transferring a little energy in each. In practice, this suggests that if the detector consists of 10 layers and the secondary particle passes through all of them, its course needs to be rebuilded on the basis of ten points. The task is just seemingly basic.”There is normally a magnetic field inside the detectors. Charged particles relocate it along curved lines and this is also how the detector elements activated by them, which in our lingo we call hits, will be found with respect to each other,” discusses Prof. Marcin Kucharczyk, (IFJ PAN) and instantly adds: “In reality, the so-called tenancy of the detector, i.e. the number of hits per detector element, might be very high, which causes many issues when trying to rebuild the tracks of particles properly. In particular, the reconstruction of tracks that are close to each other is quite a problem.”Experiments created to find brand-new physics will collide particles at greater energies than previously, meaning that more secondary particles will be produced in each crash. The luminosity of the beams will likewise need to be higher, which in turn will increase the number of crashes per unit time. Under such conditions, classical approaches of reconstructing particle tracks can no longer cope. Artificial intelligence, which excels where particular universal patterns require to be acknowledged rapidly, can concern the rescue.AI as a Solution”The synthetic intelligence we have created is a deep-type neural network. It consists of an input layer comprised of 20 neurons, 4 surprise layers of 1,000 nerve cells each and an output layer with 8 neurons. All the nerve cells of each layer are connected to all the nerve cells of the neighboring layer. Altogether, the network has 2 million setup criteria, the values of which are set throughout the learning procedure,” explains Dr. Milosz Zdybal (IFJ PAN). The deep neural network therefore prepared was trained using 40,000 simulated particle accidents, supplemented with synthetically produced sound. Throughout the testing stage, just hit info was fed into the network. As these were originated from computer system simulations, the initial trajectories of the responsible particles were known precisely and could be compared with the restorations supplied by the artificial intelligence. On this basis, the synthetic intelligence found out to correctly rebuild the particle tracks.”In our paper, we show that the deep neural network trained on a properly prepared database is able to reconstruct secondary particle tracks as properly as classical algorithms. This is a result of great value for the development of detection methods. Whilst training a deep neural network is a lengthy and computationally demanding process, a qualified network responds quickly. Considering that it does this also with satisfying accuracy, we can think optimistically about utilizing it when it comes to genuine crashes,” stresses Prof. Kucharczyk.The MUonE Experiment and Future PhysicsThe closest experiment in which the expert system from IFJ PAN would have an opportunity to prove itself is MUonE (MUon ON Electron elastic scattering). This examines an interesting inconsistency between the measured worths of a specific physical amount to do with muons (particles that are about 200 times more enormous equivalents of the electron) and predictions of the Standard Model (that is, the model utilized to describe the world of elementary particles). Measurements brought out at the American accelerator centre Fermilab reveal that the so-called anomalous magnetic minute of muons varies from the predictions of the Standard Model with a certainty of approximately 4.2 standard variances (referred as sigma). On the other hand, it is accepted in physics that a significance above 5 sigma, corresponding to a certainty of 99.99995%, is a value considered appropriate to announce a discovery.The significance of the disparity showing brand-new physics might be considerably increased if the accuracy of the Standard Models predictions could be improved. In order to better determine the anomalous magnetic moment of the muon with its help, it would be required to understand a more precise worth of the criterion understood as the hadronic correction. Unfortunately, a mathematical estimation of this specification is not possible. At this point, the function of the MUonE experiment becomes clear. In it, researchers plan to study the scattering of muons on electrons of atoms with low atomic number, such as carbon or beryllium. The results will permit a more accurate decision of certain physical parameters that directly depend on the hadronic correction. If everything goes according to the physicists plans, the hadronic correction determined in this method will increase the confidence in measuring the discrepancy in between the theoretical and determined value of the muons anomalous magnetic moment by as much as 7 sigma– and the existence of hitherto unknown physics might end up being a reality.The MUonE experiment is to start at Europes CERN nuclear center as early as next year, however the target phase has been prepared for 2027, which is most likely when the Cracow physicists will have the opportunity to see if the expert system they have produced will do its job in rebuilding particle tracks. Confirmation of its efficiency in the conditions of a real experiment could mark the beginning of a brand-new period in particle detection techniques.Reference: “Machine Learning based Event Reconstruction for the MUonE Experiment” by Miłosz Zdybał, Marcin Kucharczyk and Marcin Wolter, 10 March 2024, Computer Science.DOI: 10.7494/ csci.2024.25.1.5690 The work of the team of physicists from the IFJ PAN was funded by a grant from the Polish National Science Centre.