<h2><strong>DRD4 - WG4 "Software" Meeting of 03.05.2024</strong></h2>
<h3>Attendants</h3>
<p>Maurizio Martinelli (MM); Lorenzo Capriotti (LC); Massimiliano Fiorini (MF); Sajan Easo (SE); Fabio Gargano (FG); Federica Cuna (FC); Viatcheslav Sharyy (VS); Roman Dzhygadlo (RD); Kunal Gautam (KG); Giacomo Volpe (GV); Nicola Nicassio (NN)</p>
<h3>Agenda</h3>
<ul>
<li><em>Welcome</em> by MM
<ul>
<li>MF the webpages will most likely be based on Confluence </li>
</ul>
</li>
</ul>
<h3>Roundtable</h3>
<p>Meeting participants present themselves and their work related to DRD4 and WG4 in particular.</p>
<ul>
<li>LC (Ferrara, LHCb) works on fast simulation of the RICH using GPU to parallelise the transport of optical photons using a software from NVIDIA. Issue: software not working anymore on their GPU, upgrading their system. Plan: repeat studies for future generation of RICH detectors.
<ul>
<li>SE comments that they are working on a similar topic collaborating with people in the UK. They have also published a paper.</li>
<li>LC is aware of the effort but did not reach out yet since his work is supposed to be parallel and more general than LHCb. Happy to collaborate. </li>
<li>SE the project was LHCb-based but can be extended.</li>
</ul>
</li>
</ul>
<ul>
<li>RD (GSI, Panda) works on the PID system for the Interested in the detector development and simulation. Currently working on algorithms to apply machine learning approaces in the reconstruction. In particular to use photon and propagation time. </li>
<li>FG + FC (INFN Bari) work on space experiments. Interested to PID and tracking for detectors based on scintillator readout with SiPM also using ML. Building a system for the identification of nuclei in space. Requires the identification of hadronic showers in calorimeters for space applications. FC working on AI algorithm for reconstructing particle trajectories in space experiments.</li>
<li>VS (CEA, CaLIPSO) developing a detector for PET using MCP and SiPM. Software tasks: 1) prepare a simulation in GEANT4 of the photocatode; 2) AI reconstruction of the detector.
<ul>
<li>MF did your simulation of the photocatode became an official class of GEANT4? VS Photocatode implemented with standard GEANT4 cluster by measuring the performance at different incidence angles to define a reasonable model. SE points out that a precise knowledge of how the photocatode is built is difficult due to the intellectual property of the photocatode builder. VS indeed, our photocathode simulation is described in some details in C.-H. Sung et al. "Detailed simulation for the ClearMind prototype detection module and event reconstruction using artificial intelligence", NIM A Volume 1053, 2023, 168357, <a href="https://doi.org/10.1016/j.nima.2023.168357">https://doi.org/10.1016/j.nima.2023.168357</a> , e-Print:2209.11587 [physics.ins-det] Section 3.3</li>
</ul>
</li>
<li>MF (Ferrara, LHCb) Simulation of custom optics for test beam. We would like to optimise the simulation of optics in GEANT. Develop faster simulation of new detectors and from photocatodes to anode and digitisation.</li>
<li>KG (VU Brussels+Zurich U, FCC + ALICE) 1) hardware for ALICE3 upgrade, testing pixel sensors jet0 tagging vertexing and PID. Considering detector concepts for FCC.</li>
<li>SE (STFC, LHCb) Simulation of LHCb RICH detectors in GEANT. GPUs simulation of optics; Reconstruction using GPUs.
<ul>
<li>MM can you give me more details on the reconstruction using GPUs? Do you focus on the porting of the likelihood or using AI for reconstructing rings?</li>
<li>SE started with likelihood, we also are interested in alternative AI approaches.</li>
</ul>
</li>
<li>GV + NN (Bari, ALICE) use SiPM for detecting Cherenkov radiation in view of TOF measurements. Software: simulation, detector setup and PID
<ul>
<li>need computational power to analyse the simulation of high-multiplicity events.</li>
</ul>
</li>
</ul>
<h3>Next Meeting</h3>
<p>31.05.2024 with topical presentations</p>