Uncertainty-Aware Prognosis via Deep Gaussian Process

Abstract

The problem of Uncertainty Quantification (UQ) is of paramount importance when Machine Learning (ML) and Deep Learning (DL) models are deployed in the real world. In the context of safetycritical applications, such as the prediction of the Remaining Useful-Life (RUL) of infrastructure and industrial assets, the relevance of effective UQ approaches is even higher, given the potentially catastrophic consequences or substantial costs associated with maintenance decisions that are performed either too late or too early. Particularly in such safety critical application contexts, transparency and reliability are essential requirements and any ML-based solution not providing meaningful uncertainty estimates would not fully satisfy such desiderata. However, most ML and DL techniques used for RUL estimation are often not designed to perform UQ, thus limiting their effective applicability in real-world scenarios. To address this limitation, in this paper we investigate the performance of a recently proposed class of algorithms, Deep Gaussian Process (DGPs), for predicting the remaining useful lifetime (RUL). DGPs provide uncertainty estimates associated with their predictions, yet retaining the expressive power and learning capabilities of modern DL methods. DGPs are able to scale to very large datasets, which was a limitation of some of the previous algorithms but has increasingly become an essential requirement for many industrial applications and, in particular, for RUL prediction tasks for which UQ is of key importance. The main contribution of this paper is a thorough evaluation and comparison of several variants of DGPs applied to the problem of RUL prediction. The performance of the DGP algorithms is evaluated on the N-CMAPSS (New Commercial Modular Aero-Propulsion System Simulation) dataset from NASA for aircraft engines. The results demonstrate that DGPs are able to provide very accurate RUL predictions along with sensible uncertainty estimates, providing more reliable solutions for (safety-critical) real-life industrial applications.

Deep reinforcement learning for predictive aircraft maintenance using probabilistic Remaining-Useful-Life prognostics

“…However, both deep ensembles and the Bayesian learning are computationally intensive, especially for deep neural networks with https://doi.org/10.1016/j.ress.2022.108908 Received 10 June 2022; Received in revised form 5 September 2022; Accepted 18 October 2022 significantly many parameters [19]. Another uncertainty quantification approach is deep Gaussian process learning, which estimates the confidence interval of RUL predictions [22]. Finally, Monte Carlo dropout is an efficient approach for uncertainty quantification [23].…”

A Benchmark on Uncertainty Quantification for Deep Learning Prognostics

“…However, very few comparative studies exist to our knowledge on the performance of UQ in prognostics with Bayesian deep learning. An exception is the recent benchmark by Biggio et al [58], but the focus is on the evaluation of deep Gaussian processes whose performance is compared only with MCD and MLP. Thus, the study does not cover DE or the recent advances in VI for BNN, and scalability is only tested on a small subset of N-CMAPSS.…”

Novel Metrics to Evaluate Probabilistic Remaining Useful Life Prognostics with Applications to Turbofan Engines

“…We use a window size of 30 flights for FD001 and FD003, of 20 flights for FD002 and of 15 flights for FD004. To obtain a probability distribution of the RUL using CNN, we additionally apply Monte Carlo dropout (Biggio et al, 2021;Gal & Ghahramani, 2016). During the training phase, we apply a dropout rate of ρ = 0.5 in each layer, with the exception of the last convolutional layer before the flatten layer, and the first convolutional layer (Gal, Hron, & Kendall, 2017).…”

“…In this line, several studies estimate the distribution of RUL, i.e., probabilistic RUL prognostics (see Figure 1-B). In (Nguyen & Medjaher, 2019) and (Biggio, Wieland, Chao, Kastanis, & Fink, 2021) the RUL distribution of turbofan engines is obtained using a Long Short-Term Memory neural network and Deep Gaussian processes, respectively. In ( the RUL distribution of aircraft Cooling Units is estimated using particle filtering.…”