This paper describes an effort to revise the PMFserv agent architecture in order to implement J.J. Gibson's Affordance Theory. The theoretical justification for this revision is outlined along with the engineering constraints that inspired it. We describe the resulting architectural changes and the impact of those changes on the flexibility, ease of rapid scenario creation, and ability to reuse previous investments in knowledge engineering offered by our architecture. The level of effort required to build a new scenario within PMFserv both with and without the revisions suggested by Affordance Theory is compared. We conclude that Affordance Theory is an elegant solution to the problem of providing both rapid scenario development and the simulation of individual differences in perception, culture, and emotionality within the same agent architecture.
Date of this Version
affordance theory, software agents, culture, stress, emotion, crowd models, emergence
Date Posted: 06 September 2007