Switzerland should multiply its photovoltaic (PV) capacity by at least 6 times over the next 25 years (2050). Accurate performance assessment and optimized design are essential to maximize PV yield. By analyzing the data of hundreds of PV plants, the project aims at triggering a learning process. Focus will be dedicated to: (1) reducing uncertainties in performance loss estimates; (2) increasing statistics by enlarging the base of analyzed PV projects; (3) develop data-driven failure identification and classification algorithms (FDD). Compared to the previously SFOE-funded ASSURed-PV project, these major novel activities are planned: expansion of FDD algorithms to include irreversible failure modes, extend the scope of the analysis from small to largescale PV plants, increasing statistics.