Automated (Re-)Design of Software Agents

Abstract

Autonomous software agents are dynamic entitities: they are capable of discovering a need for change -for additional knowledge and/or functionality on the basis of their analysis of specific situations. Agent factories are capable of redesigning and reactivating agents on the basis of the information provided by agents and/or knowledge available within the agent factories. As a result agents may evolve in ways their designers could never have pre-conceived. The artefacts themselves are dynamic in a way that can not be compared to any other type of design in current design practice. A number of prototype agents and agent factories have been built to evaluate the feasibility of this concept and its consequences.

Structuring Agents for Adaptation

“…Agents are constructed from components and data types by an automated agent factory [2]. Adapting an agent entails adapting the configuration of its components and data types.…”

“…In re-design, an initial artefact is modified according to new, additional, requirements. An existing model of re-design [35] has been used to model the adaptation process. Models and systems for re-design make use of the structure of their artefacts, the same holds for the adaptation process.…”

“…External assistance may be needed to perform the necessary modifications, e.g. by an agent factory [2].…”

Designing creative artificial systems

“…Brazier and Wijngaards describe a knowledge-level model of a self-modifying agent and apply it to a mobile information retrieval agent on the World Wide Web. In addition, Brazier and Wijngaards (2002) explain how to redesign software agents by using an agent factory to redesign and reactivate an agent on the basis of its current design, requirements provided by the agent and/or knowledge available within the agent factory. In the current implementations, the design process within the agent factory is relatively well defined.…”