Mitchell Lab 2012 


Distinguished Invited Professor: Roland Snooks


Coordinated locally by Gabriel Esquivel operated as a satellite office working remotely with feedback from the designers’ technical challenges and intellectual charge.

Project Mangers: Ryan Wilson and Adrian Cortez.



The project was to design two houses that are situated in extreme environments that necessitate the use of high performance composite construction.



This project operated on the house as a platform for exploring the generation of fibrous assemblages through high population agent-based methodologies. The re-conceptualization of the agency of matter through an understanding of swarm logic necessarily re-situates design intent from operating at the global to the local level. The posited application of swarm systems within architectural design involves encoding simple architectural decisions within a distributed system of autonomous computational entities - agents. It is the interaction of these local decisions that self-organizes design intention, giving rise to a form of collective intelligence and emergent behavior at the global scale.


The repositioning of design intent and the complex order generated by the behavioral techniques of multi-agent systems has implications for the affects which are generated as well as the nature of hierarchy within architecture. The distributed non-linear operation of swarm systems intrinsically resists the discrete articulation of hierarchies within Modern architecture and contemporary parametric component logic. The bottom up nature of swarm systems refocuses tectonic concerns on the assemblage at the micro scale rather than the sequential subdivision of program or form.


Instead we looked for an alternative organization of matter that draws from an understanding of micro-structures such as those found in butterfly wings; where color and pattern are determined through the organization of matter as a geometrical configuration rather than through chemical attributes such as pigmentation.



The design methodology involved a feedback between three main areas of design experimentation: 1. multi-agent generative design algorithms, 2. explicit digital modeling, 3. composite fiber fabrication experiments. These 3 areas of experimentation were not intended to be sequential, but instead each informed the other.


Multi-agent generative algorithms

Multi-agent algorithms based on Craig Reynolds flocking algorithms were developed and deployed as a design tool. There is an existing library of algorithms developed by Kokkugia and written in Processing ( that was used and expanded on. These will be taught through a series of tutorials in the first few weeks of the semester.


Explicit modeling

Explicit modeling operated in both the initial setup that becomes the input into the generative algorithmic process as well as a tool for further developing the organization and geometry that is generated through the algorithmic processes. The primary explicit modeling tool will be Maya subdivision surfaces.


Composite fiber fabrication prototypes

Experiments with composite fiber fabrication were undertaken to: test material properties, develop suitable material parameters for the generative algorithms and to fabricate prototype chunks. Composites of fibers and a bonding agent will be tested – such as fiberglass. Non-toxic resins and bonding agents were tested and used.