Optimising Aircraft Maintenance Operations through Cloud-Based Predictive Maintenance and Analytics
Weerasekara Mudiyanselage, Dian Jr (2024)
2024
All rights reserved. This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2024060420513
https://urn.fi/URN:NBN:fi:amk-2024060420513
Tiivistelmä
The thesis was completed to overcome from a problem that was faced by Sri Lankan Air Force aircrafts used in United Nations missions in Central Africa. The main issue that the air force facing was the lack of proper methods to predict the maintenance needs of aircrafts and due to this reason air force has faced issues with providing their aircrafts continuously to the missions.
To address the given issue, a web-based application was developed to give the predictions about the maintenance type that the aircraft has to go through before the next flight. Previously, the aircraft maintenance in central Africa was based on scheduled maintenance approach. There were three main approaches of aircraft maintenance types was followed by air force as Line 1, Line 2, and Line 3 and each of these Line Types had subcategories such as lubrication, corrosion, and component maintenance. The main objective of the system was to predict the specific maintenance type needed for each aircraft to be completed before the next flight. The application is a web-based platform that the backend is managed by Node JS and Express JS and the front end of the system is handled by React JS. The Machine Learning component of predicting the maintenance type is handle by python random forest classifier. The whole system is following the micro services architecture.
The implementation has been successful, by enabling the air force to accurately predict the next type of maintenance that the aircraft has to undergo before the next flight. This achievement has improved the efficiency of the aircrafts to be in part of the missions that is conducting by United Nations in Central Africa. The result of this project provides a reliable method for proactive aircraft maintenance management, enhancing the operational readiness and reducing the downtime.
To address the given issue, a web-based application was developed to give the predictions about the maintenance type that the aircraft has to go through before the next flight. Previously, the aircraft maintenance in central Africa was based on scheduled maintenance approach. There were three main approaches of aircraft maintenance types was followed by air force as Line 1, Line 2, and Line 3 and each of these Line Types had subcategories such as lubrication, corrosion, and component maintenance. The main objective of the system was to predict the specific maintenance type needed for each aircraft to be completed before the next flight. The application is a web-based platform that the backend is managed by Node JS and Express JS and the front end of the system is handled by React JS. The Machine Learning component of predicting the maintenance type is handle by python random forest classifier. The whole system is following the micro services architecture.
The implementation has been successful, by enabling the air force to accurately predict the next type of maintenance that the aircraft has to undergo before the next flight. This achievement has improved the efficiency of the aircrafts to be in part of the missions that is conducting by United Nations in Central Africa. The result of this project provides a reliable method for proactive aircraft maintenance management, enhancing the operational readiness and reducing the downtime.