Learning by Single Function Agents during Spring Design

Designing creative artificial systems

“…Artificial learning systems are available for many tasks and many domains of application. In design, examples of such systems are knowledge-based agents (e.g., Grecu & Brown, 1996), machine learning systems (e.g., Duffy, 1997), clustering mechanisms (e.g., Reffat & Gero, 2000), neural networks (e.g., Saunders & Gero, 2002), and genetic algorithms (e.g., Gero, 1996;Cho, 2002).…”

“…Grecu and Brown (1996) describe a multi-agent system for designing springs, which consists of single-function agents (SiFAs). Each SiFA is a small (knowledge-based) expert agent, which operates autonomously, negotiates bilaterally (using pre-programmed strategies) with other SiFAs to reach an agreement about a specific aspect of the spring design, and which learns to reduce the overhead of inter-agent communication.…”

Knowledge level model of an individual designer as an agent in collaborative distributed design

“…A number of researchers, but not many, have combined the research areas of design and multi-agent systems. For example, Grecu and Brown describe a system for parametric spring design [48], built from small knowledge-based systems called Single -4---Function Agents (SiFAs). The deliberately restricted capabilities of each SiFA enforce the interaction among SiFAs; only by co-operation are they able to carry out a design task.…”

“…The aforementioned research on distributed design [48,49,50] does not include explicit representations for reflective reasoning. Being able to reason about, or even from, the viewpoint of another agent is a means with which, e.g., conflicts can be prevented.…”

Design agents that learn

“…Therefore, learning is desirable for ac-D.L. Grecu and D.C. Brown quisition or change in the following areas: design concepts or concept attributes; constraints related to elements of the design; dependencies between design parameters; support in favor or against a decision; design actions and design rules; preferences in selection tasks; sequences of actions and plans; preconditions and postconditions for rules, actions and tasks; knowledge allowing the prediction of the actions of other agents; strategies for carrying out tasks; types of conflicts; heuristics to solve conflicts and to negotiate (Grecu, 1994); knowledge for the construction of shallow models of other agents.…”