BSRH Auto-alignment

This report describes the work conducted during my internship at CERN on the automation of an alignment procedure for the Beam Synchrotron Radiation Halo monitor (BSRH). BSRH is an optical diagnostic instrument that monitors the transverse beam distribution in the LHC with a particular emphasis on the low-intensity halo. Its optical system has to be properly aligned for accurate measurements, with manual alignment being time-consuming and prone to drift.

To address this, a modular system of neural networks was developed, with each network responsible for aligning one optical path defined by a pair of targets. The models achieved high accuracy for the second and third optical paths, demonstrating the feasibility of data-driven alignment in such a complex optical system.