The AID2E collaboration develops software infrastructure in parallel with advancing detector optimization for the ePIC experiment. Our work builds upon the iDDS and PanDA frameworks for distributed computing while creating specialized solutions for high-performance environments including SLURM-based parallelization. These efforts support systematic exploration of detector design parameters through multi-objective optimization, enabling data-driven improvements to ePIC detector subsystems.
Core software systems and tools that enable distributed detector design optimization workflows.
Distributed workflow management system for ePIC detector simulations leveraging PanDA and iDDS frameworks
Core utilities and optimization tools for AID2E optimization workflows
Retrieval-Augmented Generation system for the Electron Ion Collider providing AI-powered documentation search and summarization
Application of AI-assisted optimization methods to specific ePIC detector subsystems.
Multi-objective Bayesian optimization for dual Ring Imaging Cherenkov detector design
Optimization of the B0 tracking system for the ePIC detector
An application of the AID2E framework to the ePIC Barrel Imaging Calorimeter (BIC).
Optimization of the central tracking detector system
Additional applications and use cases of the AID2E framework beyond ePIC detector optimization.
Bayesian optimization for the alignment of the CLAS12 RICH detector
Optimization of iron-scintillator sampling calorimeter for the Detector 2
Optimizing design of AEROGEL tile using fibers to improve mechanical properties while maintaining optical performance for RICH detectors