High-Definition Imaging of Selected Hexaboards for the CMS High Granularity Calorimeter

In the near future, the CMS Experiment at CERN's Large Hadron Collider will incorporate the High Granularity Endcap Calorimeter (HGCAL) upgrade to its detector system as part of the High-Luminosity Era upgrade. HGCAL will be constructed modularly from many hexagonal-shaped readout boards (hexaboards) attached to silicon sensors.

For quality control purposes, the University of Alabama has created a pipeline to automatically accept or reject boards based on visual criteria. However, this method requires a digital image of each board. Conventional imaging, such as webcams, phones, and microscopes, fail to provide the required image quality in a single image required for both traditional and machine-learning pipelines. As such, the University of Alabama has created two automated image capturing machines. Instead of taking single-images of hexaboards, the machine takes multiple high-definition images of subsections of each board. The machines control for ambient light, and are capable of returning to any given position on a board with sub-millimeter level precision. To image one hexaborad, each machine takes hundreds of 4k images in roughly three to five minutes, then applies corrections based on the type of optics in use.