This project explores the use of CycleGAN for generating synthetic winter driving data from summer scenes, with the goal of improving robustness in autonomous driving systems under adverse weather conditions.