












The project focuses on AI modeling and degradation prediction, aiming to develop advanced AI based models capable of identifying and quantifying the dominant meteorological stress factors driving photovoltaic degradation, and contributing to the recommendation of indoor testing protocols that most accurately reflect real outdoor conditions. These models will also enable real time estimation of power output and degradation in industrial PV fields equipped with emerging technologies, supporting their integration into MPPT strategies. In addition, the AI models will be embedded into a decision support and field deployment tool to facilitate the industrial rollout of perovskite and tandem modules. The work includes the analysis of both outdoor and indoor datasets, leveraging real PV field measurements to pinpoint key stress factors such as irradiance dose, thermal gradients, day–night cycling, and humidity. Finally, the developed AI/ML frameworks will be applied to the prediction of underlying physical phenomena, using regression methods, time series forecasting, and hybrid models that combine physics based and data driven approaches.