Using Artificial Intelligence Based Expert System for Selection of Design Subcontractors: A Case Study in Aerospace Industry

Abstract

As one of the top expectations for type certification of an aircraft, Aviation Authorities (AA) regulate design organization to establish Design Assurance System (DAS). DAS is composed of design, independent monitoring and airworthiness functions in which these functions are specialized for aerospace industry. Besides, Design Organization Approval (DOA) is a milestone to establish a rigid Design Assurance System. By this way, design organization assures aircraft development life cycle by complying with aviation regulations. To meet requirements of Design Organization Approval, Design Organization transfers its authority and technical signatories to its subcontractors to improve effectiveness of the system. So, performance of design subcontractors shall be traceable and measurable to match capability requirements of main contractor. Thus, subcontractor evaluation is a long and complicated process; survey implementation could be misleading in some cases. The purpose of this study is to propose a novel tool to measure performance of a design subcontractor according to necessities of Design Assurance System. Up to now, there is no tool to evaluate aviation design subcontractors. With this tool, contractor firm can evaluate multiple criteria in a single run. AHP is used to prioritize criteria relative to each other one-by-one. Then, for subcontractor selection and subcontractor monitoring, Artificial Neural Network (ANN) is applied to optimize decision making process. Annual Actual Data is applied in AHP model to assess current performance score of subcontractor. To have a long term judgment of this system, the model shall be applied to a design subcontractor for more than once on fixed periods such as quarterly, yearly etc.