Automated Aircraft Maintenance Data Processing and Analytics

NSERC ARD赠款的行业合作伙伴是三菱航空集团(https://mhirj.com，三菱重工-支线喷气式飞机，前身为庞巴迪公司支线喷气式飞机)。MHIRJ为全球支线飞机行业提供全面的关键运营、工程和客户支持解决方案。MHIRJ需要来自Humber学院的应用研究援助，以满足与维护相关的两个互补需求。第一个需求是自动化维护数据处理系统，该系统将与飞机运营商(航空公司)以多种格式和不同间隔提交的维护数据一起工作。这些数据将被摄取、处理、分析并作为数据库输出，用于进一步分析单个机队中飞机的个别性能，以及单个机队的整体性能与竞争运营商机队的汇总和匿名视图的比较情况。建议的解决方案可以显著减少MHIRJ人员管理和维护这一过程所需的时间和精力。只要在数据提交后24小时内就能得到输出，自动化数据处理就不是时间紧迫的。第二个需求是需要一种综合分析工具，以帮助地面维护人员根据每个航班的实时运营数据，在非常狭窄的两次航班之间的时间窗口内做出关键的、对时间敏感的“可分派”决策：“去”、“有限制地去”和“不去”。综合分析工具的设计将基于为自动维护数据处理开发的相同数据处理和分析算法。这将带来更一致的数据解释，提高飞机维修人员的效率，加快飞机的周转时间，提高安全性。亨伯学院的应用研究团队将提供文本数据挖掘、语言处理、机器学习、数据分析和数据库设计方面的相关专业知识，以帮助开发预期解决方案所需的软件算法和工具。该项目将提高MIHRJ的竞争力，并在MHIRJ和Humber学院之间建立更强有力的合作。

The industry partner for this NSERC ARD Grant is MHIRJ Aviation Group (https://mhirj.com, Mitsubishi Heavy Industries - Regional Jet, formerly Bombardier Inc. Regional Jet). MHIRJ provides comprehensive critical operational, engineering and customer support solutions for the global regional aircraft industry. MHIRJ needs applied research assistance from Humber College to address two complementary maintenance-related needs. The first need is for Automated Maintenance Data Processing system that will work with maintenance data submitted - in multiple formats and over varying intervals - by aircraft operators (airline companies). This data will be ingested, processed, analyzed, and outputted as databases for further analytics of the individual performance of aircraft in the individual fleets; as well as how the overall performance of individual fleets compare to an aggregated and anonymized view of the fleets of competing operators. The proposed solution can reduce significantly the time and effort required by MHIRJ personnel to administer and maintain the process. The automated data processing is not time critical as long as the output is available within 24 hours of data submission. The second need is for an Integrated Analytical Tool to assist ground maintenance crews in their critical and time-sensitive 'dispatchability' decision: 'go,' 'go with limitation,' and 'no go,' in a very narrow window of time between flights, based on real-time operational data from each flight. The design of the Integrated Analytic Tool will be based upon the same data processing and analytics algorithms developed for Automated Maintenance Data Processing. This will result in more consistent data interpretation, improved efficiency by aircraft maintenance crews, faster turn-around times of aircraft with improved safety. Humber College's applied research team will provide relevant expertise in textual data mining, language processing, machine learning, data analytics, and database design, to assist in developing software algorithms and tools necessary for the intended solutions. The project will improve the competitiveness for MIHRJ and establish a stronger collaboration between MHIRJ and Humber College.