Unplanned maintenance is a costly factor in aircraft operations. Predictive Maintenance aims at reducing the surprise effect of unplanned maintenance and thereby its associated cost. A variety of statistical models are used to estimate the remaining life, as well as sensors to gauge component condition. The application to statistical models of sensory information coming from the pilot, in the form of pilot complaints, appears to be an overlooked option worth investigating. In other words: What is the effect of pilot complaints on the predictability of component removals? This question is answered by determining relevant words in the pilot complaints using a TF-IDF analysis and use the presence of these words as covariate in the well known Proportional Hazards Model. Left truncation and right censoring is applied to limit the time-invariant nature of these covariates. The results in the form of hazard ratios indicate a hazard increase of several orders of magnitude with respect to baseline hazard. These results are put into perspective when compared when compared to the known outcome of the pilot complaints, making their added predictability seem marginal. Another adverse indication is the violation of the proportionality assumption. The magnitude of the hazard ratios do suggest that additional
measures in the from of a more in depth natural language processing and the application of time-varying covariates could bring the concept closer to practical application.