Predictive-Maintenance Digital Twin of Turbofan Engines Using RUL Prediction and Maintenance Optimisation

When should a jet engine be taken off the wing - as late as possible, but before it fails? This project builds a predictive-maintenance digital twin for every engine of a fleet on the NASA C-MAPSS turbofan data: it estimates the remaining useful life from the last 30 flight cycles of 14 sensors, compares classical regression, degradation and similarity models with gradient boosting, an LSTM and a 1-D CNN ensemble, keeps a Kalman-filtered RUL state per engine as new cycles stream in, and raises work orders with a cost-optimal threshold that is compared with run-to- failure and fixed-interval maintenance, for single engines and for a simulated fleet with crews and spare engines. A web fleet dashboard and a complete MATLAB version are included.

Category: Industrial Engineering

Industrial Engineering Project IE04 from ProTechSC Academy, with graphical abstract, video demonstration and project guidance.