Improving charm CPV measurements with real-time data reconstruction

The violation of CP asymmetry in charm decays is a small effect, and has only recently been experimentally established in a single decay mode. This thesis describes an effort towards expanding the experimental knowledge in this field, following two parallel directions of work. The first is the development of a novel analysis method for extracting CP asymmetries in 3-body decays, that is a middle ground between a full Dalitz analysis and a counting experiment. This new method allows the extraction of precision CP parameters from individual resonances with the same statistical resolution of a Dalitz analysis but with an improved control of systematics, and is suitable for large samples. This is used to analyze a promising decay mode for a second CPV observation, $D^0 \to K_S K^\mp \pi^\pm$ , based on the largest currently available sample of those decays, collected by LHCb in Run 2. The second line of work is the development of custom real-time tracking processor, aimed at maintaining a high trigger efficiency for charm decays also in the high-luminosity environment of future LHCb runs, allowing further improvements in the precision of charm CPV measurements. This system is based on a computing architecture called ``Artificial Retina'', implemented in FPGAs, that has the potential of reconstructing full events at the crossing rate of the LHC (averaging 30 MHz). As a result of the work performed in this thesis, all the core parts of the system are now designed and tested, and a realistic demonstrator based on those parts is being built. The demonstrator will operate in parasitic mode during physics data taking at LHCb to reconstruct tracks in a portion of the detector at the full speed of 30 MHz, The first layer of this system, performing hit finding in the VELO vertex detector, has already surpassed the demonstration stage and has been adopted by LHCb for actual physics data taking already in Run 3, providing improved DAQ throughput and data compression.