Automating Contextualized Maintenance Documentation

Currently, task-specific support information in aircraft maintenance is mostly provided using paper-based solutions, which are burdensome, slow and prone to error. Next to these issues, Aircraft Maintenance Technicians (AMTs) have very limited on-site access to support information. This leads to 15-30% of handson-aircraft time being wasted on acquiring the right information or not using maintenance documentation for task support at all, risking maintenance error. The decision by the AMT to use or not use maintenance documentation is also affected by human factors, such as time pressure and complacency, and poses an additional risk for the number one priority in aviation: safety. To overcome the weaknesses of maintenance documentation and mitigate the related human factor effects, a mobile tool will be developed to provide direct access to task-specific information. With such a system the AMT will have on-site access to the right information, spend less time on retrieving the right information and therefore will be able to make wellinformed decisions under less pressure, reducing maintenance errors. This paper describes the development of a system for a first level of contextualization of maintenance documentation. Combining a tailor-made ontology for the aerospace domain with a relational database system for Ontology-Based Data Access (OBDA), maintenance documents relevant to a specific aircraft registration mark can be identified. The system contributes to the research field of knowledge management by using OBDA for selecting relevant maintenance documents stored in a regular file folder and provides a promising start in the development of a mobile contextualized documentation tool to improve operational efficiency in the (line) maintenance environment. Future work includes advanced methods for contextualization, integration of semantic search capabilities, exploiting document referencing, implementation of end-user feedback and preferences, machine learning and verification and validation of a prototype in an operational environment.