Identifying NOx Sensor Failure for Predictive Maintenance of Diesel Engines using Explainable AI
##plugins.themes.bootstrap3.article.main##
##plugins.themes.bootstrap3.article.sidebar##
Abstract
The automotive industry is being transformed by the application of artificial intelligence and big data analysis. In particular, predictive analytics is becoming a powerful tool for anticipating component failure. This key area of research provides automotive industry manufacturers with lower warranty expenses and incremental service parts revenue while rewarding customers with higher uptime. These benefits are particularly important for commercial vehicles operations such as bus and truck fleets, since analytics led predictive maintenance can prevent inconvenient and costly interruptions of vehicle mission. Accurate prediction models for component failures are especially challenging. This paper describes one such effort for failure prediction of transit bus NOx (Nitrogen Oxides) sensors.
Stringent emissions regulations have made diesel exhaust aftertreatment systems mandatory in nearly all global markets, and NOx sensors play a critical role in control and diagnostic algorithms used by these systems. NOx is measured before and after the SCR (Selective Catalytic Reduction) system with two different components namely Engine out (EO) NOx and System out (SO) NOx. Due to differences in operating conditions and failure rates these two components were studied separately. The results of different Machine Learning algorithms were obtained and compared to get the optimal predictions. Moreover, early life and late life failures were also studied separately to differentiate between random and wear-out failure modes. Highlights of the paper are: - the data collection process, feature engineering and feature selection process, as well as explainable AI (Artificial Intelligence) built on top of the machine learning model. Efforts were also taken to keep the approach generic and not become too component specific so that it can easily be replicated for predicting other failures on other product lines or of different components.
How to Cite
##plugins.themes.bootstrap3.article.details##
Predictive Analytics, Nox Sensor Failure, Explainable AI, Decile Analysis, Machine Learning, XGBOOST, Preventive maintenance
The Prognostic and Health Management Society advocates open-access to scientific data and uses a Creative Commons license for publishing and distributing any papers. A Creative Commons license does not relinquish the author’s copyright; rather it allows them to share some of their rights with any member of the public under certain conditions whilst enjoying full legal protection. By submitting an article to the International Conference of the Prognostics and Health Management Society, the authors agree to be bound by the associated terms and conditions including the following:
As the author, you retain the copyright to your Work. By submitting your Work, you are granting anybody the right to copy, distribute and transmit your Work and to adapt your Work with proper attribution under the terms of the Creative Commons Attribution 3.0 United States license. You assign rights to the Prognostics and Health Management Society to publish and disseminate your Work through electronic and print media if it is accepted for publication. A license note citing the Creative Commons Attribution 3.0 United States License as shown below needs to be placed in the footnote on the first page of the article.
First Author et al. This is an open-access article distributed under the terms of the Creative Commons Attribution 3.0 United States License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.