We create new mathematical methods to design graphene structures for specific applications. We predict shape and mechanical properties of graphene flocks with defects. We use machine learning to combine large-scale and nano-scale information to optimize the pattern of defects.