The maintenance of equipment or machinery within a company or organization is a problem that must be addressed in depth. Ineffective maintenance management can lead to long equipment downtime, low productivity, and high-cost corrective decisions. Nowadays, thanks to the combined effort of machine sensors and smart technology, maintenance can be analyzed from a new perspective. The amount of data generated from diagnosis processes has increased in recent years due to the proliferation of sensing technologies [Car19]. Predictive maintenance (PdM) employs condition-based technologies using diagnostic analysis to predict future events such as remaining useful life of a component or faults diagnosis [She05].
Machine Learning (ML) methods have emerged as a promising tool for PdM applications to prevents failures in equipment [Car19]. Recent works exploits the computing capacity of ML methods such as Artificial Neural Networks [Cre19] to predict when abnormalities can appear, Bayesian Filters [Jos20] to estimate the gradual degradation of machinery or Support Vector Machines [Cha18] to predict maintenance of vehicle fleets.
Our team has experience in the development of predictive models and, since the end of 2019, has worked on the development of a predictive maintenance platform for the Spanish land army. Thanks to this project, a data set has been collected, corresponding to the monitoring of 200 variables taken from 65 land vehicles (trucks). Based on this information, we intend to work on the following tasks:
- Development of prediction models for the values of different variables over time. Analysis and optimization of the factors that influence the prediction.
- Validation of the use of GANN for the generation of synthetic patterns. Evaluation of these artificial patterns for developing models of breakdown prediction.
[Car19] T.P Carvalho et al. (2019). A systematic literature review of machine learning methods applied to predictive maintenance. Computers & Industrial Engineering, 137.
[She05] I. Sherrington, T. Sperring, M. Williamson. (2005). Condition Monitoring as a Tool to Aid Compliance with ISO 14000. Tribology and Interface Engineering Series, Volume 48, 295-304.
[Cre19] A. Crespo Márquez, A. de la Fuente Carmona, S. Antomarioni. A process to implement an artificial neural network and association rules techniques to improve as set performance and energy efficiency. Energy 12(18), 2019.
[Jos20] J.-R. Ruiz-Sarmiento et al. A predictive model for the maintenance of industrial machinery in the context of industry 4.0. Engineering Applications of Artificial Intelligence, Volume 87, 2020.
[Cha18] Chaudhuri, A. (2018). Predictive Maintenance for Industrial IoT of Vehicle Fleets using Hierarchical Modified Fuzzy Support Vector Machine.