Share these talks and lectures with your colleagues
Invite colleaguesResearch paper
Predicting maintenance costs of an IT system using artificial intelligence models
Applied Marketing Analytics: The Peer-Reviewed Journal,
10
(1),
68-76
(2024)
Abstract
Predictive maintenance is a maintenance policy where the goal is to detect potential future maintenance risks in a system so that the maintenance process can be optimised before system faults occur. This paper describes a deep learning model that does not require domain expertise. Deep learning approaches have several benefits over explicit statistical modelling: (1) they require far less domain-specific knowledge; (2) if the underlying data-generating mechanism of assets changes, a deep learning model would only need to be retrained to learn these new changes; (3) they can capture non-linear and complex multidimensional relationships; and (4) they may outperform rule-based or statistical methods. The paper describes how the model predicts maintenance-relevant events, along with the cost of the upcoming event and the time when it will happen. The paper describes the use of a long short-term memory architecture for our deep learning model. By doing so, the cost values represent a real, quantitative value of the potential maintenance costs in the future of an asset. Event, cost and time prediction are all achieved with high accuracy. This allows for the development of maintenance solutions without the initial high degree of domain process knowledge required. The artificial intelligence model can be used to raise an alarm when the cost values exceed some threshold, when the frequency of high-cost events increases significantly over the lifetime of an asset, or when the expected cost exceeds the cost of maintenance.
Keywords: predictive maintenance; deep learning; long short-term memory; LSTM; cost prediction; time prediction
The full article is available to subscribers to the journal.
Author's Biography
Nathan Bosch is a machine learning (ML) engineer at Lyft, working on map data. Prior to joining Lyft, Nathan worked on software engineering research, specifically focusing on analysing and building predictive models around system log files, marketing research — focusing primarily on Shapley Value regression and oversampling techniques for typing tools — and applied artificial intelligence research, notably on fine-tuning large language models for telecommunications-specific retrieval tasks. Alongside his studies, Nathan worked at Ericsson (2019–22), a global telecommunications company, and Kwantum Analytics (2021–22), focusing on developing novel ML approaches for driver analysis, class imbalance correction, methods for market segmentation and language models for advanced topic modelling.
Emmanuel Okafor earned a PhD in artificial intelligence from the University of Groningen, The Netherlands, in 2019. He received an MSc in control engineering and a BEng in electrical engineering from Ahmadu Bello University (ABU), Nigeria, in 2014 and 2010, respectively. He participated in the MIT-ETT Fellowship at the Massachusetts Institute of Technology (MIT), United States, in 2022. Additionally, he worked at ABU for more than 10 years as an academic staff member. He is currently a postdoctoral researcher at the SDAIA-KFUPM Joint Research Center for Artificial Intelligence, King Fahd University of Petroleum and Minerals (KFUPM), Saudi Arabia. He has co-authored a book and written over 30 international research articles (journal articles and conference papers).
Marco Vriens is CEO of Kwantum, an analytics firm. He is the author/editor of such books as ‘The Business of Marketing Research’ (2020), ‘From Data to Decision: Handbook for the Modern Business Analyst’ (2018), ‘The Insights Advantage: Knowing How to Win’ (2012) and ‘The Handbook of Marketing Research’ (2006). His work has been published in both academic journals and industry publications.
Lambert Schomaker has been a professor in artificial intelligence at the University of Groningen since 2001. He is known for research in simulation and recognition of handwriting, writer identification, style-based document dating and other studies in pattern recognition, machine learning and robotics. He has (co)authored over 200 publications. He and his team have worked in robotics and industrial maintenance, text and speech to image transforms using generative adversarial networks, and deep learning for microscopy.
Citation
Bosch, Nathan, Okafor, Emmanuel, Vriens, Marco and Schomaker, Lambert (2024, June 1). Predicting maintenance costs of an IT system using artificial intelligence models. In the Applied Marketing Analytics: The Peer-Reviewed Journal, Volume 10, Issue 1. https://doi.org/10.69554/CSVO2679.
© Henry Stewart
Publications LLP
Publications LLP