The Analysis Method And Data Management System For Aviation Maintenance Error

With the rapid development of the aviation industry, flight safety has been paid more and more attentions. As one of the most important causes to the aircraft accidents, aviation maintenance error must be explored deeply in order to keep flight safety. The exploration includes the analysis of error modes, and the derivation of errors countermeasures, etc.The definition, characteristics and significance of aviation maintenance error are presented firstly. Then the HFACS-ME developed by U.S. Navy and MEDA by the Boeing Corporation are discussed briefly. One tool for maintenance error is designed by modofication of the Maintenance Error Decision Aid (MEDA) Results Form. Two methods for the analysis of the relationship between error mode and error cause are discussed detailedly, which are grey relational analysis and association rule mining method, respectively. The practicality of the two methods is illustrated by one example and some historical data about aviation maintenance errors.Lack of data is one of the most difficult topics of human reliability assessment. Therefore, aviation maintenance errors data collection and management is essential. One software package named Aviation Maintenance Error Information Management System is designed based on the aforementioned discussions. Firstly, the users'requirements, technical requirements and functional requirements are analyzed, and the systematical and functional structures are outlined. Secondly, the data model is constructed by using the E-R tool, and is converted into database design model. Finally, the database is accomplished by SQL Server 2005; and the interface is realized on the framework of Visual Studio 2008 and using the language C# 3.5. With the help of the package, the aviation maintenance errors data can be stored and managed with one more effictive manner, and the workload can be reduce to a great extent. The statistical analysis outcomes can provide decision support for aviation safety management and data support for further quantitative analysis.