Gabriella Azzopardi is a final year Ph.D. student based at CERN. Her thesis is being supervised by Dr Gianluca Valentino and Prof. Adrian Muscat from Department of Communications and Computer Engineering at UM and Dr Belen Maria Salvachua Ferrando from CERN.
The Large Hadron Collider (LHC) at CERN is the largest particle accelerator in the world, with a circumference of 27km, accelerating and colliding two counter-rotating beams at 6.5 TeV. The LHC is susceptible to beam losses, therefore a collimation system of 100 collimators is used to safely dispose of such losses before they damage the magnets and other sensitive equipment.
Each collimator consists of two jaws which need to be positioned precisely around the beam, by aligning them regularly. Aligning all the collimators takes hours each time, therefore Gabriella fully-automated the procedure using machine learning and other forms of data analysis. Her implementation for fully-automating the alignment successfully decreased the alignment time by 70% and requires minimal human intervention.
In October 2019, Gabriella presented the software implemented for her Ph.D. at the 17th International Conference on Accelerator and Large Experimental Physics Control Systems (ICALEPCS). The conference was held between 5 and 11 October in New York, NY, USA. Apart from her two presentations, Gabriella presented a poster on different parts of the alignment software developed. She received an overwhelming amount of positive feedback on her work and was ultimately awarded the Best Student Poster award.
Dr Valentino was invited to deliver a tutorial on Reinforcement Learning, and presented results from the experience with beam luminosity levelling for heavy-ion beams in the ALICE experiment.