Swarm intelligence 2009 spring columbia

The major project of the Swarm Intelligence workshop is to develop a multi-agent design methodology that operates at an urban scale. An application of swarm logic to urbanism enables a shift from notions of the master-plan to that of master-algorithm as an urban design tool. This shift changes the conception of urban design from a sequential set of decisions at reducing scales, to a simultaneous process in which a set of micro or local decisions interact to generate a complex urban system. Rather than designing an urban plan that meets a finite set of criteria, urban imperatives are programmed into a set of agents which are able to self-organise.

Urban structures always act as ecologies while their design is often sequential. Swarm Urbanism seeks to develop design strategies which operate as an ecology of mutually negotiating agents, which cause a necessary rethinking of many of the hierarchies which exist in urban planning. This project necessitates the development of a minimum of two teams of agents which operate to generate two different aspects of the urban fabric (eg. road networks, public space, program, built mass).

The deliverables for this project will be as a minimum:
1. working processing swarm code
2. plans/sections/axos of the final design
3. catalogue of possible outcomes (demonstrating the extents and refinement of the system)
4. diagram of the rules encode into the agents

syllabus

This seminar will examine the role of agency within generative design processes. The course will engage algorithmic techniques in the development of a computational methodology grounded in swarm intelligence. While discussing the political and social role of agency, the workshop will focus on an abstract design methodology, recasting simple decision making ability into agents capable of self-organizing into an emergent intelligence. Scripting will form the basis for algorithmic models which enable localized interaction of agents to generate emergent topologies in the design of proto-architectural forms, structures and articulation. Unlike the typical application of swarm systems in design, this workshop will not engage simply in the mapping of these complex systems, but instead, we will mine the self-organizing potential of the systems to negotiate between a complex set of desires and parameters in the generation of architecture.

The semester will focus around two areas of research, initially developing simulations of vector based swarm systems and then using these as the basis for developing an architectural design methodology which operates within a topological substrate. This second stage of the research will shift away from any analogous relationship to an existing swarm systems and develop a design process capable of negotiating architectural inputs.

The seminar will intensively engage scripting, using and expanding on a library of agent code which will ensure that students are not slowed by the necessity of writing all the code they require. Instead the focus will be on the application and manipulation of code within the design process. This expansive library has been developed through research at Kokkugia and previous studios and seminars. The library consists of relatively simple functions (such as steering behaviors) which can be recombined in the development of more complex algorithms.

Computational design is shifting away from the reliance on heavy platforms such as Maya’s MEL scripting language into lightweight object oriented programming environments, enabling the massive iteration required for emergent processes. In anticipating this shift, the workshops will focus on the newer lightweight languages such as Processing and Python. As the research shifts into a topological substrate these languages will be transferred into Maya (Python) and RhinoScript to enable agent intelligence to be embedded into geometry.

schedule

week 1
Seminar: Introductory lecture
Instruction: introduction to processing

weekend workshop: processing - object oriented + vector libraries + intro to steering behavior

week 2
Instruction: Steering behaviors + processing in 3D

week 3
Instruction: Agent libraries and advanced steering behaviors
Critique: Desk crits for project 1B

week 4
Seminar: The algorithmic emergence of character
Critique: presentation of project progress

week 5
Seminar: The substrate of agent code

week 6
Seminar: Materializing the Swarm
Instruction: Processing Output + RhinoScript
Critique: Desk crits for project 1C

week 7
Final project review

Retrieved from "http://kokkugia.com/wiki/index.php5?title=Swarm_intelligence_2009_spring_columbia"

Swarm intelligence 2009 spring columbia

From KokkugiaWiki

Contents

workshops

desk crit sign up 04/19

kAgent code

working code

student projects

resources

news

project: swarm urbanism

syllabus

schedule

Views

Personal tools

Navigation