The researchers have created a new machine learning program that could speed up designing more efficient solar cells by predicting the power-conversion efficiency (PCE) of substances for application in next-generation organic solar cells.