Remaining Useful Life Prognosis of Aircraft Brakes

Abstract

We investigate the performance of three different data-driven prognostic methodologies towards the Remaining Useful Life estimation of commercial aircraft brakes being continuously monitored for wear. The first approach utilizes a probabilistic multi-state deterioration mathematical model i.e. a Hidden Semi Markov model whilst the second utilizes a nonlinear regression approach through classical Artificial Neural Networks in a Bootstrap fashion in order to obtain prediction intervals to accompany the mean remaining life estimates. The third approach attempts to leverage the highly linear degradation data over time and uses a simple linear regression in a Bayesian framework. All methodologies, when properly trained with historical degradation data, achieve excellent performance in terms of early and accurate prediction of the remaining useful flights that the monitored set of brakes can safely serve. The paper presents a real-world application where it is demonstrated that even in non-complex linear degradation data the inherent data stochasticity prohibits the use of a simple mathematical approaches and asks for methodologies with uncertainty quantification.

Bayesian Network Training Method to Produce a Condition-Based Maintenance Strategy in Aviation Maintenance Programs

“…Modern aircraft collect performance data through the Airplane Condition Monitoring System (ACMS) (Sun et al, 2019). Oikonomou et al (2022) state that new-generation aircraft such as Boeing 787 and Airbus 350 are equipped with thousands of sensors for recording and monitoring health conditions. One example is the aircraft's brake system: "…the aircraft itself measures the position of the actuators when clamped to the carbon discs and infers the carbon thickness from this measurement" (Oikonomou et al, 2022, p. 2).…”

“…For validation of CBM-BN output, it is a process to validate the CBM-BN model's performance by comparing observed and predicted output. Many researchers use various validation methods such as absolute deviation (MAD) (Dinis et al, 2019), mean absolute percentage error (MAPE), root mean squared error (RMSE) (Oikonomou., 2022), mean squared error (MSE) (Ferreiro & Arnaiz, 2011), and relative error (Sun et al, 2019).…”

“…The CBM-BN is a new educational material that enables students to understand stochastic future failures by considering heterogeneous information (Dinis et al, 2019), not just relying on predetermined failure rates given by manufacturers. Once collegiate students understand the practical understanding of CBM-BN, they can develop further advanced CBM models by applying fundamental Bayesian theorems such as Bayesian linear regression (BLR) (Oikonomou et al, 2022) and Naïve Bayes (Saeidi et al, 2019).…”

“…As the aircraft ages, structures and instruments that are fatigued by friction and vibrations eventually reach their failure point. To prevent a part failure during the aircraft's operation time and to continuously operate in safe conditions, the aviation industry has adopted and implemented a time-based maintenance strategy (Oikonomou et al, 2022). However, these traditional time-based maintenance programs have efficiency issues that increase operational costs and downtime (Prajapati et al, 2012;Oikonomou et al, 2022).…”

Explainable Data-Driven Method Combined with Bayesian Filtering for Remaining Useful Lifetime Prediction of Aircraft Engines Using NASA CMAPSS Datasets

“…However, it lacks the capability of individual asset assessment since each asset may behave differently under different loads or operating conditions. Prognostics aims to solve this problem by predicting the system state and reliability of each asset and its estimated failure time, and providing confidence and valuable information to the operators [5]. This information proves beneficial to the operator because maintenance intervention is only performed when needed [6].…”