Towards digital twins of high-power laser systems

Optics and photonics are transforming rapidly as machine learning (ML) techniques advance in the field. These ML advancements require extensive datasets, often challenging to obtain experimentally. We present a start-to-end (S2E) software framework for high-power laser systems. Using this framework, we develop a digital twin for a complex laser system, including pulse shaping, amplification, and upconversion. We explore amplifier parameters to mimic experimental conditions and generate an ML model that achieves a 250-fold speed-up over traditional numerical simulations, paving the way for generalized digital twins and enhanced synergy between simulation and experiment.