Abstract@ACADIA '97
Agent Generated Architecture

Jeffrey Krause
Blacksquare, San Francisco, CA


This paper will describe a behavior based artificial intelligence experiment in computer generated architectural design and will explain the internal representations and procedures of an agent based autonomous system. This is a departure from traditional (AI and architectural) top-down approaches, allowing hundreds of agents to work simultaneously—building, manipulating, and dismantling their environment. Individual agents work in collaboration, in disjunction or autonomously.

Architectural design is perhaps most commonly described by the architect as consisting of the ability to see the whole picture, to organize, to collect, to juggle, to manage, and to maintain multiple conflicting goals and values. Architecture by the preceding definition is hierarchical and top-down in nature. The agent based experiment in this paper presents an alternative design process, involving multiple autonomous agents acting distributively. The agents (objects) move through the design landscape, simultaneously collaborating, building, degenerating, and transforming their world.

Return to Abstract Table of Contents
Go to ACADIA '97 Home Page
Next abstract (in alphabetical order)
Previous abstract (in alphabetical order)