The second architectural stream at the root of the Mascarillons project comes from a field of research that studies complex, aggregative architectures and urban morphologies such as traditional cities, medieval towns, medinas, slums and squatter settlements, often called “unplanned” because they show no global pattern or symmetry. The Nxi Gestatio design lab in Montreal has undertaken a research aiming to characterize and simulate these complex urbanscapes, and to use tools derived from their analysis to generate entirely new tools for architectural design and composition. Many of these tools come from Artifical Life techniques ; after many attempts that led to the production of architectural drawings and sculptures, the lab decided to study the potential of swarm intelligence models to produce self-organized three-dimensional structures of architectural relevance. The merging of this program with the flying cubes project led to the idea of building robotic flying cubes with on-board computers, sensors and motors that would be able to interact and to generate either emerging behaviors or emerging structures. This idea appeared for the first time in 2000 at Caltech, during a discussion between A. Martinoli N. Reeves and G. Théraulaz (then in Toulouse, via teleconference )and was immediately considered as an exciting and challenging research program. Existing works on structures generated by social animals, such as research works by G. Théraulaz, were seen as being of the greatest relevance to undertake this research.