SOPRENE: Assessment of the Spanish Armada’s Predictive Maintenance Tool for Naval Assets

Abstract

Predictive maintenance has lately proved to be a useful tool for optimizing costs, performance and systems availability. Furthermore, the greater and more complex the system, the higher the benefit but also the less applied: Architectural, computational and complexity limitations have historically ballasted the adoption of predictive maintenance on the biggest systems. This has been especially true in military systems where the security and criticality of the operations do not accept uncertainty. This paper describes the work conducted in addressing these challenges, aiming to evaluate its applicability in a real scenario: It presents a specific design and development for an actual big and diverse ecosystem of equipment, proposing an semi-unsupervised predictive maintenance system. In addition, it depicts the solution deployment, test and technological adoption of real-world military operative environments and validates the applicability.

A Fault-Detection System Approach for the Optimization of Warship Equipment Replacement Parts Based on Operation Parameters

“…In the arena of machine learning (ML) and especially fault detection systems, quite a few works can be found in the literature. Studies alike employ several prediction techniques, including regularized linear regression methods such as L1 (lasso) and L2 (ridge), or long short-term memory (LSTM)-based networks [ 34 ]. In this context, previous research has used artificial neuronal networks (ANN) [ 35 ] and decision trees [ 36 ], as well as support vector machines (SVM) [ 37 , 38 ] or the k-nearest neighbor technique (k-NN) as one of the most common for fault classification [ 39 ].…”

“…These logistics applications are difficult systems that require in-depth knowledge and expert analysis. Likewise, some authors have reviewed these tools proposing an alternative semi-unsupervised predictive maintenance system [ 34 ]. From our side, we focus on a novel fault detection system approach based on operational data provided by the ATAVIA application, as discussed below.…”

Data mining in predictive maintenance systems: A taxonomy and systematic review

“…Examples of this case are found on the one hand in industrial processes (Bampoula et al, 2021; Bekar et al, 2020; Chang et al, 2021; Kim et al, 2021; Kolokas et al, 2020; Lepenioti et al, 2020; Li et al, 2017; Morariu et al, 2020; Rafique et al, 2018; Ruiz‐Sarmiento et al, 2020; Susto et al, 2015; Uhlmann et al, 2018; Zschech et al, 2019), in production lines (Ayvaz & Alpay, 2021; Azab et al, 2021; Cerquitelli et al, 2021; Fathi et al, 2021; Giordano et al, 2021; Liu et al, 2021), in power plants (de Carvalho Chrysostomo et al, 2020; Khodabakhsh et al, 2018; Sun et al, 2021; Zhang, Liu, et al, 2018), in wind turbines (Chen, Hsu, et al, 2021; Leahy et al, 2018; Santolamazza et al, 2021), in ventilation systems (Fernandes et al, 2020), cryogenic pumps (Crespo Márquez et al, 2020), heat meters (Pałasz & Przysowa, 2019), press machines (Serradilla et al, 2021), or water treatment plants (Srivastava et al, 2018). Another common scenario for PdM is related to transportation: different types of land vehicles (Chen et al, 2020; Patil et al, 2021; Prytz et al, 2015; Shafi et al, 2018), aircrafts (Baptista et al, 2021; Basora et al, 2021; Ning et al, 2021; Savitha et al, 2020; Yang et al, 2017), and naval ships (Berghout et al, 2021; Fernández‐Barrero et al, 2021; Gribbestad et al, 2021) have been monitored through their electronic control units. Also, Ribeiro et al (2016) found a case that monitors train doors status.…”

“…After the phase of novelty detection, other ML methods based on unsupervised learning are employed to explain the characteristics of the anomalies detected. Other approach of AE designed to deal with time series by means of LSTM in its internal layers is explored in several works (Bouabdallaoui et al, 2021; Fernández‐Barrero et al, 2021; Ning et al, 2021). The anomalies are detected from the differences between time series at the input and at the output.…”

“…From the studies of Morariu et al (2020) and Sun et al (2021), it can be found that LSTM models learn patterns and variations in numerical measurements concerning energy consumption in order to produce forecasts. A similar approach is followed in Fern andez- Barrero et al (2021) that uses LSTMs to predict future behaviors of the monitored signals at different time horizons. In Chui et al (2021), it is proposed a hybrid method for RUL prediction composed of a generic RNN and a LSTM whose outputs are combined using weighting factors optimized by a genetic algorithm.…”