Predictive Maintenance of Aircraft Engines Using Fuzzy Bolt

The Remaining Useful Life (RUL) of engines is a very important prognostic parameter that can be used to make a decision on when an aircraft engine needs to be sent for maintenance or repair. Today, there is no way to accurately estimate the RUL of an engine. Access to various sensor readings could provide more insights into RUL degradation. However, the relationship between these sensor readings obtained from flight data and the RUL of an engine is not well understood. In this paper, we attempt to provide an estimation of the engine RUL based on the time history data obtained from different sensors. A Genetic Fuzzy System, trained using Fuzzy Bolt((C)), is used to make useful estimations of RUL, which could in turn help with providing a marker for when an engine needs to be sent for maintenance. The models are trained on the NASA C-MAPSS dataset available for turbofan engines. We also compare our methodology with a similarity based model that has been proven to be one of the best models in predicting RUL on this dataset.