One of the major problems for designers during complex design projects is integration of the various distributed and heterogeneous resources involved. These resources must cooperate with each other in a uniform and efficient way; there must be interoperability, and therefore interaccessibility, among the programs. Some standards for interoperability in engineering design environments do exist, and these can be used to clarify problems, but establishing this interoperable environment still remains as a challenging issue.
The authors offer CLOVER as a solution to the problem of cooperative intelligent design. CLOVER is a multiagent system (MAS), or, more specifically, a multiagent cooperative design environment. CLOVER is composed of many individual agents, including PMAs, DTAs, TAs, personal digital assistants (PDAs), and human agents, which are individually capable of efficient cooperation in a design project. The paper presents an architecture composed of four different types of design agents, which are separated according to their design activities, and which work within the infrastructure of CLOVER.
The authors discuss issues of interoperability and cooperation in CLOVER, and present their approach to these problems. The prototype implementation of this approach proved to be valuable, because it gave the designers an idea of whether or not the product was being built to meet the interoperability standards and cooperation requirements. The authors also discuss issues related to agent-based cooperative intelligent design, including product model, design process model, design knowledge representation, agent architecture, multiagent system architecture, inter-agent communication, and ontology.
In conclusion, in this well-written paper, it is shown that agent technology can provide a flexible and dynamic cooperative paradigm in an intelligent design environment. It is also proved that, based on the existing standards for interoperability, CLOVER can be used to improve interoperability among applications and build high-level autonomous cooperation.