The need for intelligent predictive maintenance (PdM) systems that provide data privacy and operational efficiency has been made clear by Industry 4.0 and smart manufacturing. The scalability of traditional, centralized learning models is constrained by bandwidth, high communication costs, limited scalability, and data security, especially on Industrial Internet of Things (IIoT) platforms. This study addresses these challenges by proposing a lightweight and distributed predictive maintenance framework based on split learning (SL) that employs the sample convolution and interaction network (SCINet) to identify faults in distributed AC motor systems. In our system, each motor uses four sensors (vibration, current, temperature, and ambient) that capture real-time data, which is processed on-site, and only intermediate representations are sent to a cloud server where the model is trained. With the deep neural network (DNN) divided into edge devices and cloud, we can minimize the exposure of raw data, as well as the communication overhead. Experimental results exhibit that the proposed framework achieves 99.35 per cent accuracy as well as 0.0022 loss in the classification of the five types of failures (normal, overcurrent, misalignment, stop rotation, and heavy load), which is better than centralized training (99.04 per cent accuracy and 0.0216 loss), while reducing communication overhead and computational burden on edge devices. The major contributions are: a privacy-conserving SL architecture that does not require the exchange of raw data, feature extraction using SCINet to better model temporal dependencies, and bandwidth saving through data transmission through compression. The experimental findings prove that SL not only improves fault detection but also meets the key IIoT requirements, namely, data security, computational efficiency, and real-time performance. The paper offers a scalable solution to intelligent maintenance in Industry 4.0 that can be extended further in the future by the introduction of hybrid federated-split learning to cross-factory collaboration.
Keywords
Predictive MaintenanceIndustrial Internet of Things (IIoT)Convolutional Neural Network (CNN)SensorsSCINet.
References
P. Karuppusamy, “Machine learning approach to predictive maintenance in the manufacturing industry comparative study,” Journal of Soft Computing Paradigm (JSCP), vol. 2, no. 4, pp. 246-255, 2020, [Online]. Available: https://doi.org/10.36548/jscp.2020.4.006.
A. R. Zarzoor, N. A. S. Al-Jamali, and D. A. A. Qader, “Intrusion detection method for the Internet of Things based on the spiking neural network and decision tree method,” International Journal of Electrical and Computer Engineering, vol. 13, no. 2, pp. 2278-2288, 2023, [Online]. Available: https://doi.org/10.11591/ijece.v13i2.pp2278-2288.
O. Alfarraj, “Internet of things with bio-inspired co-evolutionary deep-convolution neural-network approach for detecting road cracks in smart transportation,” Neural Computing and Applications, vol. 37, no. 35, pp. 29061-29071, 2025, [Online]. Available: https://link.springer.com/article/10.1007/s00521-020-05401-9.
M. Cakir, M. A. Guvenc, and S. Mistikoglu, “The experimental application of popular machine learning algorithms on predictive maintenance and the design of an IIoT-based condition monitoring system,” Computers & Industrial Engineering, vol. 151, p. 106948, 2021, [Online]. Available: https://doi.org/10.1016/j.cie.2020.106948.
M. Sohaib, S. Mushtaq, and J. Uddin, “Deep learning for data-driven predictive maintenance,” in Vision, Sensing, and Analytics: Integrative Approaches, pp. 71-95, Cham: Springer International Publishing, 2021, [Online]. Available: https://doi.org/10.1007/978-3-030-75490-7_3.
Z. Li, Q. He, and J. Li, “A survey of deep learning-driven architecture for predictive maintenance,” Engineering Applications of Artificial Intelligence, vol. 133, p. 108285, 2024, [Online]. Available: https://doi.org/10.1016/j.engappai.2024.108285.
Q. Duan, S. Hu, R. Deng, and Z. Lu, “Combined federated and split learning in edge computing for ubiquitous intelligence in the internet of things: State-of-the-art and future directions,” Sensors, vol. 22, no. 16, p. 5983, 2022, [Online]. Available: https://doi.org/10.3390/s22165983.
Z. Zhao, Y. Mao, Y. Liu, L. Song, Y. Ouyang, X. Chen, and W. Ding, “Towards efficient communications in federated learning: A contemporary survey,” Journal of the Franklin Institute, vol. 360, no. 12, pp. 8669-8703, 2023, [Online]. Available: https://doi.org/10.1016/j.jfranklin.2022.12.053.
S. M. S. Mohammad Abadi, S. Zawad, F. Yan, and L. Yang, “Speed Up Federated Learning in Heterogeneous Environments: A Dynamic Tiering Approach,” IEEE Internet of Things Journal, 2024, [Online]. Available: https://doi.org/10.1109/JIOT.2024.3487473.
R. W. Abdalah, O. F. Abdulateef, and A. H. Hamad, “A Predictive Maintenance System Based on Industrial Internet of Things for Multimachine Multiclass Using Deep Neural Network,” Journal Europeen des Systemes Automatises, vol. 58, no. 2, 2025.
A. Jalali, M. Soltany, M. Greenspan, and A. Etemad, “Learning Time-Series Representations by Hierarchical Uniformity-Tolerance Latent Balancing,” arXiv preprint arXiv:2510.01658, 2025, [Online]. Available: https://doi.org/10.48550/arXiv.2510.01658.
Z. Hu, T. Zhou, B. Wu, C. Chen, and Y. Wang, “A review and experimental evaluation of split learning,” Future Internet, vol. 17, no. 2, p. 87, 2025, [Online]. Available: https://doi.org/10.3390/fi17020087.
S. Guo, X. Zhang, F. Yang, T. Zhang, Y. Gan, T. Xiang, and Y. Liu, “Robust and Privacy-Preserving Collaborative Learning: A Comprehensive Survey,” arXiv, arXiv:2112.10183, 2021, [Online]. Available: https://doi.org/10.48550/arXiv.2112.10183.
Z. Hu, T. Zhou, B. Wu, C. Chen, and Y. Wang, “A review and experimental evaluation of split learning,” Future Internet, vol. 17, no. 2, p. 87, 2025.
G. H. Lokesh, G. S. Hukkeri, N. Z. Jhanjhi, and H. Lin, Split Federated Learning for Secure IoT Applications: Concepts, Frameworks, Applications, and Case Studies, 2024, [Online]. Available: https://doi.org/10.1049/PBSE025E.
Z. Lin, G. Qu, X. Chen, and K. Huang, “Split learning in 6G edge networks,” IEEE Wireless Communications, vol. 31, no. 4, pp. 170-176, 2024, [Online]. Available: https://doi.org/10.1109/MWC.014.2300319.
J. Ryu, D. Won, and Y. Lee, “A Study of Split Learning Model,” in IMCOM, pp. 1-4, Jan. 2022, [Online]. Available: https://janicejihyeon.github.io/files/IMCOM2022_Jihyeon.pdf.
M. Liu, A. Zeng, M. Chen, Z. Xu, Q. Lai, L. Ma, and Q. Xu, “Scinet: Time series modeling and forecasting with sample convolution and interaction,” Advances in Neural Information Processing Systems, vol. 35, pp. 5816-5828, 2022.
J. Zhang, C. Zhang, S. Xu, G. Liu, H. Fei, and L. Wu, “Remaining life prediction of bearings based on improved IF-SCINet,” IEEE Access, 2024, [Online]. Available: https://doi.org/10.1109/ACCESS.2024.3355978.
K. M. Hosny, A. Magdi, A. Salah, O. El-Komy, and N. A. Lashin, “Internet of things applications using Raspberry Pi: A survey,” International Journal of Electrical & Computer Engineering, vol. 13, no. 1, pp. 902-910, 2023, [Online]. Available: https://doi.org/10.11591/ijece.v13i1.pp902-910.
A. Halbouni, T. S. Gunawan, M. H. Habaebi, M. Halbouni, M. Kartiwi, and R. Ahmad, “Machine learning and deep learning approaches for cybersecurity: A review,” IEEE Access, vol. 10, pp. 19572-19585, 2022, [Online]. Available: https://doi.org/10.1109/ACCESS.2022.3151248.
R. H. Abood and A. H. Hamad, “Multi-Label Diabetic Retinopathy Detection Using Transfer Learning Based Convolutional Neural Network,” [Online]. Available: https://doi.org/10.54216/fpa.170221.
