Multi-Agent Task Environment Framework to Support Multidisciplinary Design and Optimization

Abstract

The multidisciplinary design and optimization process of products can be supported by automation of analysis and optimization steps. A design and engineering engine is a useful concept to structure this automation. Various discipline analysis tools analyze the particular instantiation of this product model, and an optimizer (or search engine) provides a strategy to drive the design toward a desired design, that is, compliant with required performance, a desired product behavior, or one that is desired by a (human) designer or design team. To power and orchestrate communication of the automatic analysis in this multidisciplinary design and optimization setting, an agent-based framework called TeamMate has been developed and implemented in a design and engineering engine for design and analysis of a vertical tail of an airliner. To facilitate further development of the agent-based framework, an analysis of the problem domain is performed, and a use case is discussed and reflected with literature. From findings, a new set of requirements is proposed for future framework development. Conclusively, several ideas for the next release of the agent-based framework are described. The main contributions of this article are 1) the concept of scope levels for the multidisciplinary design and optimization problem domain, 2) the definition of actor roles in a hybrid (human–computer) multifunctional team to solve the multidisciplinary design and optimization problem domain, 3) the concept of engineering services and engineering service specialist actors being directly responsible for it, 4) the concept of a multi-agent task environment framework to motor the multidisciplinary design and optimization problem-solving environment of the design and engineering engine, and 5) the lessons learned using the multi-agent task environment framework with suggestions for further work.