Automotive fault nowcasting with machine learning and natural language processing
(2024) In Machine Learning 113(2). p.843-861- Abstract
Automated fault diagnosis can facilitate diagnostics assistance, speedier troubleshooting, and better-organised logistics. Currently, most AI-based prognostics and health management in the automotive industry ignore textual descriptions of the experienced problems or symptoms. With this study, however, we propose an ML-assisted workflow for automotive fault nowcasting that improves on current industry standards. We show that a multilingual pre-trained Transformer model can effectively classify the textual symptom claims from a large company with vehicle fleets, despite the task’s challenging nature due to the 38 languages and 1357 classes involved. Overall, we report an accuracy of more than 80% for high-frequency classes and above 60%... (More)
Automated fault diagnosis can facilitate diagnostics assistance, speedier troubleshooting, and better-organised logistics. Currently, most AI-based prognostics and health management in the automotive industry ignore textual descriptions of the experienced problems or symptoms. With this study, however, we propose an ML-assisted workflow for automotive fault nowcasting that improves on current industry standards. We show that a multilingual pre-trained Transformer model can effectively classify the textual symptom claims from a large company with vehicle fleets, despite the task’s challenging nature due to the 38 languages and 1357 classes involved. Overall, we report an accuracy of more than 80% for high-frequency classes and above 60% for classes with reasonable minimum support, bringing novel evidence that automotive troubleshooting management can benefit from multilingual symptom text classification.
(Less)
- author
- Pavlopoulos, John ; Romell, Alv ; Curman, Jacob ; Steinert, Olof ; Lindgren, Tony ; Borg, Markus LU and Randl, Korbinian
- organization
- publishing date
- 2024
- type
- Contribution to journal
- publication status
- published
- subject
- keywords
- Automotive fault nowcasting, Multilingual text classification, Natural language processing
- in
- Machine Learning
- volume
- 113
- issue
- 2
- pages
- 19 pages
- publisher
- Springer
- external identifiers
-
- scopus:85173121401
- ISSN
- 0885-6125
- DOI
- 10.1007/s10994-023-06398-7
- language
- English
- LU publication?
- yes
- id
- e338b887-8afa-40a4-8809-9006de94b4eb
- date added to LUP
- 2023-12-19 14:55:38
- date last changed
- 2025-10-14 09:44:38
@article{e338b887-8afa-40a4-8809-9006de94b4eb,
abstract = {{<p>Automated fault diagnosis can facilitate diagnostics assistance, speedier troubleshooting, and better-organised logistics. Currently, most AI-based prognostics and health management in the automotive industry ignore textual descriptions of the experienced problems or symptoms. With this study, however, we propose an ML-assisted workflow for automotive fault nowcasting that improves on current industry standards. We show that a multilingual pre-trained Transformer model can effectively classify the textual symptom claims from a large company with vehicle fleets, despite the task’s challenging nature due to the 38 languages and 1357 classes involved. Overall, we report an accuracy of more than 80% for high-frequency classes and above 60% for classes with reasonable minimum support, bringing novel evidence that automotive troubleshooting management can benefit from multilingual symptom text classification.</p>}},
author = {{Pavlopoulos, John and Romell, Alv and Curman, Jacob and Steinert, Olof and Lindgren, Tony and Borg, Markus and Randl, Korbinian}},
issn = {{0885-6125}},
keywords = {{Automotive fault nowcasting; Multilingual text classification; Natural language processing}},
language = {{eng}},
number = {{2}},
pages = {{843--861}},
publisher = {{Springer}},
series = {{Machine Learning}},
title = {{Automotive fault nowcasting with machine learning and natural language processing}},
url = {{http://dx.doi.org/10.1007/s10994-023-06398-7}},
doi = {{10.1007/s10994-023-06398-7}},
volume = {{113}},
year = {{2024}},
}