Proceedings of the 2014 Industrial and Systems Engineering Research Conference Y. Guan and H. Liao, eds. Ontology-Based Modeling of Aircraft to support Maintenance Management System Yuchang Wu, Vahid Ebrahimipour, Soumaya Yacout Department of Mathematics and Industrial Engineering, École Polytechnique de Montréal, Montréal, Québec H3T 1J4, Canada Abstract The purpose of this paper is to present an ontology–based approach for promoting aircraft representation during maintenance management process. It describes how the problem of maintenance record heterogeneity is overcome by employing ontology-based schema approach. The approach employs ontology molding to realize the function structure of an object within causality inference during semantic extraction from the context of a maintenance report and combines OWL-Lite/RDF and international standards namely ATA ISPEC2200 and ISO 15926 in a meaningful way in order to obtain a generic system-level representation model. An inherent advantage of our approach is to cross link the identified tokens, cue words from ontologically encoded maintenance record with the causality logic according to physical-based function structure of equipment that forms fault propagation in part-component levels. Thus, the result is a generic technical understanding that enriches semantic extractor and knowledge discovery from maintenance report. At the end, OWL and RDF are employed to convert the generic human-readable interpretation to a common computer-readable representation that provides a knowledge base. The approach is finally applied to a typical aircraft system in order to illustrate its capability in dealing with heterogeneity and data inconsistency in the maintenance report, identifying domain-specific terms and common data. Keywords Maintenance Knowledge Sharing, Maintenance Data Integration, Asset Integrity Management, OWL Representation, Aircraft Representation 1. Introduction The price for a commercial aircraft can cost as much as 200 million dollars, and an additional 2 billion dollars towards operation, maintenance and support throughout the economic life, which is around 20 to 25 years [1]. Safety and reliability are always the most critical components of public transportation, especially in civil aviation industry because of the profound impact on either public concern or economy cost. Apart from the finest design and manufacture of aircraft, equipment maintenance management is the key to keep reliable and cost-effective operations [2]. Maintenance is gaining increasing attention from both of aircraft manufacturers and operators. Meanwhile, in order to comply with probably most strict rules and regulations and to ensure passengers’ safety, airlines have to undergo time-consuming and costly maintenance, repairs and overhaul procedures. According to IATA (The International Air Transport Association) 2012 report of Maintenance Cost Task Force (MCTF) statistical results, the average maintenance cost was $1,014 per flight hour, $2,547 per flight cycle and $3.4 million per aircraft. The direct maintenance cost for this group of airlines increased by 37.7% from 2009 to 2012 [3]. Reducing the maintenance cost is one of the biggest challenges for airlines. Efficient maintenance management is vital to make airlines operate profitably through decreasing the operations cost, improving the reliability of equipment and avoiding the potential hazard. Thereby, the maintenance procedures have to be optimized continuously. During the lifetime of aircraft, operator and maintenance expert deals with a wide range of situations of malfunctions that result in a production loss or safety hazards. This interaction between human and aircraft results in an in-field experience that may be recorded in linguistic words on a log sheet. The maintenance log sheet usually encompasses texted- based description of failure events and findings along with numerical indicators showing resource consumptions. This log sheet is thus a reliable source of information that provides fault diagnosis for better understanding and prediction of equipment state. Unfortunately, absence of a retrievably comprehensive knowledge base extracted from the log sheets hinders the ultimate use of the embedded knowledge. In this paper, we present a task-based