Mu2e Experiment
Overview
The Mu2e is an sub-atomic based experiment owned by Fermilab. The overall experiment is about observing the products of atomic decay of muons to electrons, which would not be supported by physic's Standard Theory. In the future, they will have a large set of equipment to create a massive beam of muons("Production" in figure 1) to send through a particle detector("Tracker", in Figure 1) which would also detect any decay products of muons. If this experiment is able to find a large amount of muon to electron conversions, then it would serve as evidence to question the validity of the Standard model.
More Detail: https://www.frontiersin.org/articles/10.3389/fphy.2019.00001/full
More Detail: https://www.frontiersin.org/articles/10.3389/fphy.2019.00001/full
Figure 1: Future Muon producer and particle detector
https://www.hep.ucl.ac.uk/muons/mu2e/pictures/Mu2e-Experiment.png
https://www.hep.ucl.ac.uk/muons/mu2e/pictures/Mu2e-Experiment.png
Mu2e Lawrence Berkeley National Lab
Lawrence Berkeley National Labs have their own research under the larger Mu2e experiment. LBNL's mu2e research is creating simulations of through a future particle detector, and creating a prototype particle detector. I was a research assistant under this particle detector section of research. This particle detector was FPGA connected to sensors that measured any disruptions of a magnetic field within the prototype.
I worked in development research making FPGA code for identifying sub-atomic particles, specifically muons, going through a prototype LBNL created. I was given sensor data from muons going through the prototype and simulated data of different sub-atomic particles going through the prototype. I developed a parallel real-time machine learning identification algorithm for particle detection based off of sensor values from Muon Detector Prototype. Almost all machine learning algorithms are made to be on fast serial processing of a computer instead of slow parallel processing that can be done on a FPGA in the prototype. The algorithm I made was made after computer science's brick sort(Figure 2), which compares every pair of elements in an array and either swap the two elements or leave them where the two elements started. The data can be sorted within n computation cycles and the machine learning identification parameters can be used to identify the particles going through the prototype particle detector.
I worked in development research making FPGA code for identifying sub-atomic particles, specifically muons, going through a prototype LBNL created. I was given sensor data from muons going through the prototype and simulated data of different sub-atomic particles going through the prototype. I developed a parallel real-time machine learning identification algorithm for particle detection based off of sensor values from Muon Detector Prototype. Almost all machine learning algorithms are made to be on fast serial processing of a computer instead of slow parallel processing that can be done on a FPGA in the prototype. The algorithm I made was made after computer science's brick sort(Figure 2), which compares every pair of elements in an array and either swap the two elements or leave them where the two elements started. The data can be sorted within n computation cycles and the machine learning identification parameters can be used to identify the particles going through the prototype particle detector.
Figure 2: Brick sort example
https://www.geeksforgeeks.org/odd-even-sort-brick-sort/
https://www.geeksforgeeks.org/odd-even-sort-brick-sort/