Analogous to the way crystals are built from atoms, supercrystals are built from nanocrystals (also called quantum dots). To further optimize crystallization strategies for supercrystals, high-throughput experiments will be conducted in the near future. As this process will generate a large number of images, autonomous evaluation will be essential to avoid bottlenecks in the workflow. Therefore, this project focuses on training a machine learning model that can automatically determine whether a microscopy image contains supercrystals or not.