Abstract
Mitigating unplanned breakdowns in commercial vehicle fleets is critical for operational efficiency, yet challenging due to the complexity of diagnostic data. This study presents a predictive maintenance framework designed for control systems, utilizing fault codes transmitted via the J1939 CAN Bus protocol. We analyzed Diagnostic Message 1 (DM1) structures, specifically Suspect Parameter Numbers (SPNs) and Failure Mode Identifiers (FMIs), to extract interpretable features for real-time applications. The research utilizes a dataset of operational telemetry and diagnostic logs from 120 internal combustion buses over two years. We developed a structured feature engineering pipeline that incorporates temporal alignment, correlation analysis, and XGBoost-based feature weighting to track mechanical fault progression, with a specific focus on engine oil pressure events. To improve model interpretability, we integrated SHAP analysis. Our results identify cumulative runtime and braking system indicators as strong predictors for early fault detection. By converting discrete diagnostic codes into continuous indicators, the proposed framework enables direct integration of predictive intelligence within fleet management and supervisory systems.
Keywords
Machine learning Predictive Maintenance Fault Detection Feature Selection XGBoost SHAP Analysis
References
- Aeddula, O., Frank, M., Ruvald, R., Johansson Askling, C., Wall, J., & Larsson, T. (2024). AI-driven predictive maintenance for autonomous vehicles for product-service system development. Procedia CIRP, 128, 84–89. https://doi.org/10.1016/j.procir.2024.05.016
- Bekar, E. T., Nyqvist, P., & Skoogh, A. (2020). An intelligent approach for data pre-processing and analysis in predictive maintenance with an industrial case study. Advances in Mechanical Engineering, 12(5), 1687814020919207. https://doi.org/10.1177/1687814020919207
- Büyükdemir, B. N., Köroğlu, S., Ataş, B., Gökirmak, N., & Karaarslan, E. (2025). Smart fleet management in urban public transportation: Machine learning based predictive maintenance with real-time CAN-Bus data. In Proceedings of the 2025 Innovations in Intelligent Systems and Applications Conference (ASYU) (pp. 1–6). 10.1109/ASYU67174.2025.11208302
- Cavus, M., Dissanayake, D., & Bell, M. (2025). Next Generation of Electric Vehicles: AI-Driven Approaches for Predictive Maintenance and Battery Management. Energies, 18(5), 1041. https://doi.org/10.3390/en18051041
- Ercevik, A. I., & Ozbayoglu, A. M. (2025). Artificial intelligence based predictive maintenance for electric buses (arXiv:2510.23879) [Preprint]. ArXiv. https://arxiv.org/abs/2510.23879
- Alves Garcia, B. R. (2024). Combining Random Forest and SHAP for interpretability and insights in predictive maintenance (Master’s thesis, National College of Ireland, Dublin). https://norma.ncirl.ie/id/eprint/8670
- Hakami, A. (2024). Strategies for overcoming data scarcity, imbalance, and feature selection challenges in machine learning models for predictive maintenance. Scientific Reports, 14, 9645. https://doi.org/10.1038/s41598-024-59958-9
- Hancock, J. T., Khoshgoftaar, T. M., & Liang, Q. (2025). A problem-agnostic approach to feature selection and analysis using SHAP. Journal of Big Data, 12, 12. https://doi.org/10.1186/s40537-024-01041-1
- Ibrahim, H., Ali, A. M., & Attia, T. (2025). Optimizing electric bus performance via predictive maintenance: A combined experimental and modeling approach. Frontiers in Future Transportation, 5, 1506866. https://doi.org/10.3389/ffutr.2024.1506866
- Kaur, A., & Sarmadi, M. (2024). Comparative analysis of machine learning techniques for imbalanced genetic data. Annals of Data Science, 12(5), 1553–1575. https://doi.org/10.1007/s40745-024-00575-8
- Lins, R. G., de Freitas, T. N., & Gaspar, R. (2025). Methodology for commercial vehicle mechanical systems maintenance: Data-driven and deep learning based prediction. IEEE Access, 13, 33799–33812. https://doi.org/10.1109/ACCESS.2025.3543600
- Lundberg, S. M., & Lee, S.-I. (2017). A unified approach to interpreting model predictions. In I. Guyon, U. Von Luxburg, S. Bengio, H. Wallach, R. Fergus, S. Vishwanathan, & R. Garnett (Eds.), Advances in Neural Information Processing Systems (Vol. 30). Curran Associates, Inc. https://proceedings.neurips.cc/paper_files/paper/2017/file/8a20a8621978632d76c43dfd28b67767-Paper.pdf
- Mahale, Y., Kolhar, S., & More, A. S. (2025). A comprehensive review on artificial intelligence driven predictive maintenance in vehicles: Technologies, challenges, and future research directions. Discover Applied Science, 7, 243. https://doi.org/10.1007/s42452-025-06681-3
- Nagy, J., & Lakatos, I. (2024). Predictive maintenance and predictive repair of road vehicles—Opportunities, limitations and practical applications. Engineering Proceedings, 79(1), 27. https://doi.org/10.3390/engproc2024079027
- Rana, S. (2025). AI-driven fault detection and predictive maintenance in electrical power systems: A systematic review of data-driven approaches, digital twins, and self-healing grids. American Journal of Advanced Technology and Engineering Solutions, 1(1), 258–289. https://doi.org/10.63125/4p25x993
- Rasheed, R., Qazi, F., Agha, D. S., Ahmed, A., Asif, A., & Shams, H. (2024). Machine learning approaches for in-vehicle failure prognosis in automobiles: A review. VFAST Transactions on Software Engineering, 12(1), 169–182. https://doi.org/10.21015/vtse.v12i1.1713
- Syed, S. (2024). Enhancing school bus engine performance: Predictive maintenance and analytics for sustainable fleet operations. Library of Progress-Library Science, Information Technology & Computer, 44(3), 17765–17775.
- Tormos, B., Pla, B., Sánchez-Márquez, R., & Carballo, J. L. (2025). Explainable AI using on-board diagnostics data for urban buses maintenance management: A study case. Information, 16(2), 74. https://doi.org/10.3390/info16020074
- Zhang, C. (2024). Intelligent fault prediction and diagnosis of intelligent networked vehicle based on filter pruning with snapshot stacked ensemble deep neural network. In Proceedings of the 2024 4th International Conference on Mobile Networks and Wireless Communications (ICMNWC) (pp. 1–5). IEEE. 10.1109/ICMNWC63764.2024.10872049
Details
| Primary Language | English |
|---|---|
| Subjects | Planning and Decision Making |
| Journal Section | Research Article |
| Authors | |
| Submission Date | January 13, 2026 |
| Acceptance Date | February 24, 2026 |
| Publication Date | February 28, 2026 |
| DOI | https://doi.org/10.61969/jai.1862835 |
| IZ | https://izlik.org/JA67FK62DH |
| Published in Issue | Year 2026 Issue: 10 |
Aim & Scope
Journal of AI is a scientific, double-blind peer-reviewed, open-access international journal that publishes original, theoretical and/or practical articles. It aims to publish a wide range of articles in different disciplines about artificial intelligence(AI), machine learning(ML), Generative AI, Digital Twin and related topics.
Journal of AI is open to papers on various topics within the field of Artificial Intelligence (AI). It invites submissions that contribute to the advancement of AI, encompassing a wide range of areas such as cognition and AI, automated reasoning and inference, case-based reasoning, commonsense reasoning, computer vision, constraint processing, ethical AI, heuristic search, human interfaces, intelligent robotics, knowledge representation, machine learning, multi-agent systems, natural language processing, planning and action, and reasoning under uncertainty. The journal seeks to publish both research findings and innovative approaches to AI problems, with an emphasis on demonstrating their practical value and effectiveness.
Journal of AI intends to facilitate effective communication between scientists. It will reflect the significant advances that are currently ongoing in computer sciences. The multidisciplinary character of this field will be typified by providing readers with a broad range of articles. The articles will be original review articles, and research papers written by individual researchers and research groups which appeal to the wider international community of academics and other professionals.
The journal aims to keep related researchers updated on the developments in a wide range of topics reporting experiments, techniques and ideas that advance the understanding of various areas of Artificial Intelligence (AI).
Presents the latest developments in Artificial Intelligence and related topics,
Emphasizes emerging and multidisciplinary fields and international trends in research and development about Artificial Intelligence,
Covers special topics in line with the progression of worldwide scientists,
Journal of AI accepts articles in English.
Journal of AI started its publication life in 2023 with this title. It will be published once a year, continuously. Open access has adopted a double-blind pre-review system and a free publication policy. No registration, processing or publication fees are required from the authors.
Journal of AI intends to publish original research papers in all main branches of Artificial Intelligence. Articles that are not directly about Artificial Intelligence but are related to the topics listed below are also included in our journal scope. However, these publications should involve the development of new technology rather than the use of existing methods. We invite submissions on a wide range of research topics, including, but not limited to:
- AI Application
- AI Technique
- AI Theory
- AI Ethics
- AI planning
- Artificial neural networks
- Natural language processing
- Computer/Machine vision
- Intelligent machine/agent
- Machine learning
- Digital Twin
- Digital Asset
- Deep learning
- Robotics
- Genetic algorithms
- Knowledge representation and reasoning
- Evolutionary computing
- Expert systems
- Internet of Things (IoT)
- Blockchain
- Web 3
- Digital Currency
- Non-Fungible Token (NFT)
- Marketplace /Digital Commerce
- Metaverse Infrastructure
- Gaming and Entertainment
- Digital Humans
- Virtual Reality
- Augmented Reality
- XR
- Artificial Intelligence (ChatBots, Generative AI and LLM-related papers)
- Emerging Technologies and Applications for Metaverse
- Decentralized Architectures for Metaverse
- Al-driven Ecosystem in Metaverse
- Cryptocurrency, Token Solutions for Metaverse
- Data security and privacy mechanisms for Metaverse
- Fraud-Detection in Metaverse
- Cyber Security for Metaverse
- Operations and Maintenance in Metaverse
- Smart Contracts for Metaverse
- Data Analytics to Identify Malicious Behaviors in Metaverse
- Decentralized and Collaborative Al for Metaverse
- Interoperability for Metaverse
- Generative AI, ChatBots and LLMs
- Immersive Hologram Technology
Author Guidelines
Paper Types:
1) Research Articles
2) Review Articles*
* Journal of AI prioritizes publishing research articles.
Limits and Constraints for Authors
- An article that has been previously rejected for any reason is not taken into consideration and is rejected directly.
- An author can publish a maximum of one article in the Journal of
AI within two years, regardless of the author's order in the article. - The maximum number of authors in an article should be 3. In cases where more authors are required, a cover letter stating the reasons should be prepared for the editor and uploaded to the system at the time of submission.
- Journal of
AI is an international journal. It strictly follows the geographical distribution of the authors of the articles published in the journal. In case of intense demand for article publication from a certain country and geography, the publisher or editor has the right to reject these papers or consider them for future issues to ensure geographical diversity. - In such cases, if the authors do not comply with the rules or upload the article to the system without stating the reasons, the editor/editorial board may reject the article without taking it into the evaluation process and giving a reason.
Articles will be DIRECTLY REJECTED WHEN the paper does not satisfy any of the following:
- The article's relationship with the scope of the journal is NOT sufficient.
- The purpose of the study and the research question is NOT given in the context or is NOT clear. The research question should be given in sufficient detail to understand it clearly
- Methods - There should be a methods section that describes the approach clearly.
- Contribution to the field - The paper should have a contribution to the field and clearly describe it.
- Ethics - If there is any sign of plagiarism or a repeated/well-known study in the literature.
- The findings/results are discussed sufficiently in a scientific cause-effect relationship by the purpose.
Manuscript Preparation
The manuscript should be written in clear, concise and grammatically correct English. It is recommended that you ask colleagues to read over your paper prior to submission to ensure it is of a high standard and conforms to a high level of scientific writing.
The maximum length is typically 15 pages for research papers and 20 pages for surveys(review papers).
Journal of AI is a template-free journal at submission.
Accepted papers should be written using the standard JAI format (Download the Microsoft Word Template ).
Download the Microsoft Word Template by clicking or using the link https://dergipark.org.tr/tr/download/journal-file/28312
Other Required Files:
- Author Publishing Agreement :
https://dergipark.org.tr/tr/download/journal-file/28291 - Permission Verification Form :
https://dergipark.org.tr/tr/download/journal-file/28313 - Author Response Form (The file is to be uploaded to the system to respond to the requests of the reviewers or editor during the article evaluation process): https://dergipark.org.tr/en/download/journal-file/32847
Attention to Authors
It is the responsibility of the author to obtain permissions, including paying any fees, for your figures, diagrams, schemes, tables and text. The author must also provide the appropriate acknowledgement text in the caption in the format prescribed by the copyright owner. For example: Reprinted by permission from Macmillan Publishers Ltd: [JOURNAL NAME] (reference citation), copyright (year of publication).
Submission Steps on Dergipark
- Register the system: https://dergipark.org.tr/en/register/
- Log in to the system: https://dergipark.org.tr/en/login
- Follow the submission steps: https://dergipark.org.tr/en/journal/4169/submission/step/manuscript/new
Order of Manuscript
The manuscript should be presented in the following order.
Title Page
This should contain the title of the contribution and the names and addresses of the authors. The full postal address, e-mail address, telephone and facsimile number of the author who will receive correspondence and check the proofs should be included.
Running Head
The running head or short title is the shortened version of your manuscript title. Not only does this help to identify your manuscript during the evaluation process, but it also acts as the manuscript title on the journal homepage providing information at a glance for people who are reading the journal.
Abstract
All manuscripts must include a brief but informative Abstract. It should not exceed 300 words and should describe the scope, hypothesis or rationale for the work and the main findings. The abstract should allow the reader to quickly have a clear idea about the rationale for the work, the experiments conducted and the results of those experiments before reading the rest of the manuscript. Both common and scientific names should be included; the authorities are not given if they appear in the title. References to the literature and mathematical symbols/equations should not be included.
Keywords
Keywords (3-5) should be provided below the Abstract to assist with indexing the article.
Introduction
The Introduction should briefly indicate the objectives of the study and provide enough background information to clarify why the study was undertaken and what hypotheses were tested.
Materials and Methods
This section should be concise but provide sufficient detail of the material used and equipment and the procedure followed to allow the work to be repeated by others.
Results
Results should be presented in a logical sequence in the text, tables and figures. Repetitive presentation of the same data in tables and figures should be avoided. The results should not contain material appropriate to the Discussion. All tables, graphs, statistical analyses and sample calculations should be presented in this section.
Discussion
The results should be discussed in relation to any hypotheses advanced in the Introduction. Comment on results and indicate possible sources of error. Place the study in the context of other work reported in the literature. Only in exceptional cases should the “Results and Discussion” sections be combined. Refer to graphs, tables and figures by number. This helps tie the data into the text in a very effective manner. Authors should also take future research and limitations into account in the Discussion section.
Conclusion
The main conclusions of the experimental work should be presented. The contribution of the work to the scientific community and its economic implications should be emphasized.
Acknowledgement
This title is mandatory in the article. The source of financial support must be acknowledged. Authors must declare any financial support or relationships that may pose a conflict of interest in the cover letter submitted with the manuscript. Technical assistance may also be acknowledged.
Funding Information
This title is mandatory in the article. The authors should acknowledge the funders of this manuscript and provide all necessary funding information.
Author Contributions
This title is mandatory in the article. Authors are required to include a statement of responsibility in the manuscript that specifies each author's contribution. The level of detail varies; some disciplines produce manuscripts that comprise discrete efforts readily articulated in detail, whereas other fields operate as group efforts at all stages.
Conflict of Interest
This title is mandatory in the article. A conflict of interest exists when judgment regarding the research is influenced by factors such as financial gain or personal relationships. All authors must disclose any financial, personal or other associations that may affect or be perceived to influence their work.
Data Availability
This title is mandatory in the article. A sample text can be as follows. "The datasets generated and analyzed during the current study are available from the corresponding author upon reasonable request.". If there is any specific information about the data, it should be shared here.
References
It is the Authors responsibility to ensure that the information in each reference is complete and accurate. Only published and “in press” references should appear in the reference list.
Tables
Tables should be self-contained and the data should not be duplicated in figures. Tables should be numbered consecutively. Each table should be presented on a separate page with a comprehensive but concise legend above the table. Tables should be double-spaced and vertical lines should not be used to separate columns. Column headings should be brief, with units of measurement in parentheses. All abbreviations should be defined in footnotes. Use superscript letters (not numbers) for footnotes and keep footnotes to a minimum. *, **, *** should be reserved for P values.
Figures
Only necessary illustrations should be included. All illustrations (line drawings and photographs) are classified as figures. Figures should be cited in consecutive order in the text. Figures should be sized to fit within the column or the full text width. Line figures should be supplied as sharp, black and white or colour diagrams, drawn with a computer graphics package. Photographs should be sharp and magnifications should be indicated on photographs using a scale bar. Graphics should be supplied as high-resolution (at least 300 d.p.i.) electronic files. Digital images supplied as low-resolution cannot be used and will not be accepted. The legend should incorporate definitions of any symbols used and all abbreviations and units of measurement should be explained so that the figure can be understood without reference to the text.
Abbreviations and Units
SI units as outlined in the latest edition of Units, symbols and Abbreviations: A Guide for Medical and Scientific Editors and Authors (Royal Society of Medicine Press, London), should be used wherever possible. Statistics and measurements should always be given in figures; except where the number begins the sentence. When the number does not refer to a unit measurement, it is spelt out, except where the number is greater than nine. Use only standard abbreviations. The word Figure should be shortened to Fig. unless starting a sentence.
Online Submission
You can submit your prepared manuscript using our online submission system which you can access using the link given below.
Submit your Article by clicking or using the link https://dergipark.org.tr/en/journal/4169/submission/step/manuscript/new
Licensing Agreement
CC BY license allows reusers to distribute, remix, adapt, and build upon the material in any medium or format, so long as attribution is given to the creator. The license allows for commercial use.
You are free to:
Share — copy and redistribute the material in any medium or format
Adapt — remix, transform, and build upon the material for any purpose, even commercially.
Review Process
https://dergipark.org.tr/en/pub/jai/page/15057
Post-Production Manuscript Revisions
https://dergipark.org.tr/en/pub/jai/page/15064
Retractions
https://dergipark.org.tr/en/pub/jai/page/15059
Policy - Generative AI usage key principles
https://dergipark.org.tr/en/pub/jai/page/17844
Ethical Principles and Publication Policy
About - Publisher
About - About Journal
About - Aim & Scope
About - Statistics
About - Indexes
About - Archive
About - Journal Metrics
About - Journal Boards
About - Contact
Ethics - Publication Ethics
Ethics - Research Ethics
Ethics - Ethics Committee Approval
Policy - Publication Frequency of Periodicals
Policy - Duties and Responsibilities
Policy - Author Guidelines
Policy - Peer Review Process
Policy - Open Access Policy
Policy - Special and Supplemental Issue Policy
Policy - Retractions and Corrections
Policy - Conflict of Interest
Policy - Plagiarism Policy
Policy - AI & Generative AI usage key principles
Policy - Indexing and Abstracting
Policy - Copyright Policy
Policy - Price Policy
Policy - Archiving Policy
Policy - Repository Policy
Policy - Can editors publish in their own journal?
Price Policy
İzmir Academy Association the owner of the Journal of AI covers most of the charges. We would like to thank the Izmir Academy Association for its contributions to science and publishing.
Fee
NO submission, processing, page, coloured figures, publication, or open-access charges are received from authors.
All articles are freely available as early online and on the day of publication on the journal’s website.
Journal of AI articles are Open Access (OA) and fully compliant with funder OA mandates; they are free to read without a subscription.
Journal of AI does not receive payments from authors under any name. The reviewers, editors and other board members also work on a voluntary basis without any fee. If the publisher has an opportunity, it can give awards or make payments to the editors, reviewers and other board members.<br>
Language Editing Services Fee
The responsibility for the grammar and language of the articles in the article belongs to the authors.
Our journal, editors or reviewers do not provide any paid services. If there is a need for language correction and proofreading, the author can have them done by independent organizations and individuals. Our journal has no responsibility for that process.
Fast Track Review Process
Journal of AI does NOT support Fast Track Review Process.
Such a process is NOT available in the journal.
This service, paid or free, is NOT included in the Journal of AI.
Indexes
Journal Boards
Publisher
info@izmirakademi.org
Phone : +90 555 597 19 84
Address : Buyukkale Mahallesi Kerimoglu Sokak No:40-2
35904 Tire / IZMIR
TURKIYE
Editor in Chief
International Editorial Board
International Scientific Board
Journal of AI
is indexed and abstracted by
WoS Research Commons, DOAJ, OpenAIRE, ERIHPLUS, Google Scholar, Harvard Hollis, Scilit, ROAD
Publisher
Izmir Academy Publishing
www.izmirakademi.org
Journal Home Page | Aim & Scope | Author Guidelines | Policies | Indexing and Abstracting | Journal Boards | Contact