Internet of Things – A Complete Solution for Aviation’s Predictive Maintenance

Abstract

The University of South Carolina has been involved in research for the US military for helicopters and rotary aircraft for over 18 years. Majority of this work has been focused on optimizing aircraft uptime and flight readiness by leveraging condition-based maintenance (CBM), more commonly known as predictive maintenance (PM). This type of maintenance differs from other classical styles (reactive and preventive) in that it has a high reliability and a low cost. The foundation of PM in any application is data collection and storage. It begins with applying tools such as natural language processing (NLP) to historical maintenance records to determine the most critical components on the aircraft. Data mining of previously collected sensor data is then used to establish the most reliable types of condition indicators (CIs) that monitor the critical components. These thresholds from the CIs can be modified over time as more data is collected. Once a data collection scheme is in place, prognostics can be used to determine the remaining useful life of a component. Using this process, along with an optimized maintenance schedule through the maintenance steering group (MSG-3) program, helps to eliminate unnecessary maintenance actions on the aircraft, as well as, reduce the inventory of components needed for the aircraft to operate. After this maintenance scheme has been set up, the Internet of Things (IoT) can be leveraged to allow the entire process to operate within a single environment. This further develops the solution, and allows actions to be executed more quickly than if they were performed individually. The expected benefits and future development of these practices will never come to fruition unless personnel are properly educated and trained. Developing a culture of predictive maintenance practices in an aviation environment is necessary to ensure success of this solution.