{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,13]],"date-time":"2026-08-13T12:28:54Z","timestamp":1786624134373,"version":"build-2736575974"},"reference-count":34,"publisher":"MDPI AG","issue":"16","license":[{"start":{"date-parts":[[2026,8,13]],"date-time":"2026-08-13T00:00:00Z","timestamp":1786579200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["62232009"],"award-info":[{"award-number":["62232009"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]},{"award":["62232009"],"award-info":[{"award-number":["62232009"]}],"id":[{"id":"https:\/\/ror.org\/01h0zpd94","id-type":"ROR","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Sensors"],"abstract":"<jats:p>Industrial sensor-based fault diagnosis often requires continuous data acquisition, local data processing, and timely model inference on edge devices. Although deep learning-based diagnostic methods have achieved promising performance, many existing approaches rely on cloud-centered processing pipelines that introduce communication overhead and potential data privacy concerns. This paper presents a CPU\u2013NPU heterogeneous edge fault diagnosis framework for industrial sensor data. The framework runs on an RK3588 local edge device and includes SQLite- and RingBuffer-based data management, sliding window generation, micro-batch construction, and model inference. The CPU is responsible for data access and buffering, preprocessing, and micro-batch preparation, while the NPU executes fault diagnosis models using the RKNN runtime environment. By performing inference locally, the framework reduces the continuous transmission of raw sensor data and supports real-time fault diagnosis under resource-constrained edge devices. Experimental results demonstrate high consistency between ONNX-based CPU inference and RKNN-based NPU inference after model conversion. Furthermore, the effects of different data input paths and micro-batch configurations are systematically evaluated. A cross-platform comparison between server-class CPU\/GPU execution and embedded NPU deployment is also conducted in terms of latency, throughput, and energy efficiency. The results show that RingBuffer-based streaming input significantly reduces data access overhead, while the effectiveness of NPU acceleration depends on both model structure and micro-batch size. The cross-platform results further demonstrate the energy efficiency advantages of the RK3588 platform, making it more suitable for practical deployment in resource-constrained edge scenarios. These findings provide practical insights for deploying fault diagnosis models on heterogeneous edge devices.<\/jats:p>","DOI":"10.3390\/s26165125","type":"journal-article","created":{"date-parts":[[2026,8,13]],"date-time":"2026-08-13T11:33:46Z","timestamp":1786620826000},"page":"5125","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["A CPU\u2013NPU Heterogeneous Edge Fault Diagnosis Framework for Industrial Sensor Data"],"prefix":"10.3390","volume":"26","author":[{"ORCID":"https:\/\/orcid.org\/0009-0000-1770-7908","authenticated-orcid":false,"given":"Kangli","family":"Xu","sequence":"first","affiliation":[{"name":"Department of Computer Science and Technology, Tsinghua University, Beijing 100084, China"},{"name":"Beijing National Research Center for Information Science and Technology, Tsinghua University, Beijing 100084, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0004-1738-7907","authenticated-orcid":false,"given":"Haozhou","family":"Wang","sequence":"additional","affiliation":[{"name":"Beijing National Research Center for Information Science and Technology, Tsinghua University, Beijing 100084, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6844-6127","authenticated-orcid":false,"given":"Chao","family":"Li","sequence":"additional","affiliation":[{"name":"Department of Computer Science and Technology, Tsinghua University, Beijing 100084, China"},{"name":"Beijing National Research Center for Information Science and Technology, Tsinghua University, Beijing 100084, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0006-5838-3747","authenticated-orcid":false,"given":"Hongxuan","family":"Liu","sequence":"additional","affiliation":[{"name":"Beijing National Research Center for Information Science and Technology, Tsinghua University, Beijing 100084, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-9390-3097","authenticated-orcid":false,"given":"Chunxiao","family":"Xing","sequence":"additional","affiliation":[{"name":"Department of Computer Science and Technology, Tsinghua University, Beijing 100084, China"},{"name":"Beijing National Research Center for Information Science and Technology, Tsinghua University, Beijing 100084, China"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1968","published-online":{"date-parts":[[2026,8,13]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"799","DOI":"10.1016\/j.ymssp.2017.11.016","article-title":"Machinery health prognostics: A systematic review from data acquisition to RUL prediction","volume":"104","author":"Lei","year":"2018","journal-title":"Mech. Syst. Signal Process."},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"6551","DOI":"10.1007\/s11831-025-10478-3","article-title":"Artificial Intelligence Based Contemporary Computational Paradigms for Intelligent Machining Processes: Review on Assessment, Progress and Recent Development","volume":"33","author":"Laghari","year":"2026","journal-title":"Arch. Comput. Methods Eng."},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"93155","DOI":"10.1109\/ACCESS.2020.2990528","article-title":"Bearing fault detection and diagnosis using case western reserve university dataset with deep learning approaches: A review","volume":"8","author":"Neupane","year":"2020","journal-title":"IEEE Access"},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"273","DOI":"10.1023\/A:1022627411411","article-title":"Support-vector networks","volume":"20","author":"Cortes","year":"1995","journal-title":"Mach. Learn."},{"key":"ref_5","first-page":"281","article-title":"Some methods for classification and analysis of multivariate observations","volume":"Volume 5","author":"MacQueen","year":"1967","journal-title":"Proceedings of the Fifth Berkeley Symposium on Mathematical Statistics and Probability, Volume 1: Statistics"},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"2278","DOI":"10.1109\/5.726791","article-title":"Gradient-based learning applied to document recognition","volume":"86","author":"LeCun","year":"1998","journal-title":"Proc. IEEE"},{"key":"ref_7","doi-asserted-by":"crossref","first-page":"533","DOI":"10.1038\/323533a0","article-title":"Learning representations by back-propagating errors","volume":"323","author":"Rumelhart","year":"1986","journal-title":"Nature"},{"key":"ref_8","unstructured":"Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A.N., Kaiser, \u0141., and Polosukhin, I. (2017). Attention is all you need. Advances in Neural Information Processing Systems 30, NeurIPS."},{"key":"ref_9","doi-asserted-by":"crossref","unstructured":"Xu, K., Li, C., Dong, W., Xing, C., and Sun, Y. (2025, January 10\u201312). Comparison of Traditional and Large Model Methods for Bearing Fault Classification Tasks. Proceedings of the 2025 Global Reliability and Prognostics and Health Management Conference (PHM-Xian), Xi\u2019an, China.","DOI":"10.1109\/PHM-Xian66756.2025.11427647"},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"11093","DOI":"10.1109\/JIOT.2023.3239944","article-title":"Edge computing on IoT for machine signal processing and fault diagnosis: A review","volume":"10","author":"Lu","year":"2023","journal-title":"IEEE Internet Things J."},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"738","DOI":"10.1016\/j.jmsy.2021.03.005","article-title":"Big data analytics for intelligent manufacturing systems: A review","volume":"62","author":"Wang","year":"2022","journal-title":"J. Manuf. Syst."},{"key":"ref_12","doi-asserted-by":"crossref","first-page":"7208","DOI":"10.1109\/TII.2021.3138510","article-title":"Deep learning-based cloud\u2013edge collaboration framework for remaining useful life prediction of machinery","volume":"18","author":"Jing","year":"2021","journal-title":"IEEE Trans. Ind. Inform."},{"key":"ref_13","doi-asserted-by":"crossref","first-page":"3820","DOI":"10.1109\/COMST.2025.3527641","article-title":"Mobile edge intelligence for large language models: A contemporary survey","volume":"27","author":"Qu","year":"2025","journal-title":"IEEE Commun. Surv. Tutor."},{"key":"ref_14","doi-asserted-by":"crossref","first-page":"119","DOI":"10.1007\/s10462-024-10748-9","article-title":"Artificial intelligence and edge computing for machine maintenance-review","volume":"57","author":"Bala","year":"2024","journal-title":"Artif. Intell. Rev."},{"key":"ref_15","doi-asserted-by":"crossref","first-page":"5618","DOI":"10.1109\/JSEN.2017.2727638","article-title":"Analysis of Statistical Time-Domain Features Effectiveness in Identification of Bearing Faults From Vibration Signal","volume":"17","author":"Nayana","year":"2017","journal-title":"IEEE Sens. J."},{"key":"ref_16","doi-asserted-by":"crossref","unstructured":"Juodelyte, D., Cheplygina, V., Graversen, T., and Bonnet, P. (2022, January 14\u201318). Predicting bearings degradation stages for predictive maintenance in the pharmaceutical industry. Proceedings of the 28th ACM Sigkdd Conference on Knowledge Discovery and Data Mining, Washington, DC, USA.","DOI":"10.1145\/3534678.3539057"},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"1793","DOI":"10.1109\/TIE.2015.2509913","article-title":"Motor bearing fault detection using spectral kurtosis-based feature extraction coupled with K-nearest neighbor distance analysis","volume":"63","author":"Tian","year":"2015","journal-title":"IEEE Trans. Ind. Electron."},{"key":"ref_18","doi-asserted-by":"crossref","first-page":"3772","DOI":"10.1109\/TII.2020.3014422","article-title":"Toward self-supervised feature learning for online diagnosis of multiple faults in electric powertrains","volume":"17","author":"Senanayaka","year":"2020","journal-title":"IEEE Trans. Ind. Inform."},{"key":"ref_19","first-page":"443","article-title":"An improved bearing fault diagnosis method using one-dimensional CNN and LSTM","volume":"64","author":"Pan","year":"2018","journal-title":"Stroj. Vestn.-J. Mech. Eng."},{"key":"ref_20","doi-asserted-by":"crossref","first-page":"379","DOI":"10.1016\/j.neucom.2022.04.111","article-title":"A time series transformer based method for the rotating machinery fault diagnosis","volume":"494","author":"Jin","year":"2022","journal-title":"Neurocomputing"},{"key":"ref_21","doi-asserted-by":"crossref","unstructured":"Liu, Y., Wang, Y., Yang, K., Sun, P., Liu, J., Du, Y., Wu, J., Boukerche, A., Leung, V.C., and Hu, X. (2025). Edge-Cloud Collaborative Computing on Distributed Intelligence and Model Optimization: A Survey. arXiv.","DOI":"10.1109\/COMST.2026.3669216"},{"key":"ref_22","doi-asserted-by":"crossref","unstructured":"Pham, M.T., Kim, J.M., and Kim, C.H. (2020). Deep Learning-Based Bearing Fault Diagnosis Method for Embedded Systems. Sensors, 20.","DOI":"10.3390\/s20236886"},{"key":"ref_23","doi-asserted-by":"crossref","first-page":"8287","DOI":"10.1109\/JSEN.2019.2911299","article-title":"In Situ Motor Fault Diagnosis Using Enhanced Convolutional Neural Network in an Embedded System","volume":"20","author":"Lu","year":"2020","journal-title":"IEEE Sens. J."},{"key":"ref_24","doi-asserted-by":"crossref","unstructured":"Huang, Y., Liang, S., Cui, T., Mu, X., Luo, T., Wang, S., and Wu, G. (2024). Edge Computing and Fault Diagnosis of Rotating Machinery Based on MobileNet in Wireless Sensor Networks for Mechanical Vibration. Sensors, 24.","DOI":"10.3390\/s24165156"},{"key":"ref_25","doi-asserted-by":"crossref","first-page":"5649","DOI":"10.1007\/s12206-023-1007-3","article-title":"Lightweight fault diagnosis method in embedded system based on knowledge distillation","volume":"37","author":"Gong","year":"2023","journal-title":"J. Mech. Sci. Technol."},{"key":"ref_26","doi-asserted-by":"crossref","first-page":"147494","DOI":"10.1109\/ACCESS.2025.3600678","article-title":"Efficient Bearing Fault Diagnosis for Edge Computing Using Grayscale Spectrograms and Hybrid Neural Model Compression","volume":"13","author":"Nguyen","year":"2025","journal-title":"IEEE Access"},{"key":"ref_27","doi-asserted-by":"crossref","unstructured":"Zhou, Z., Qiao, Y., Lin, X., Li, P., Wu, N., and Yu, D. (2025). A Deployment Method for Motor Fault Diagnosis Application Based on Edge Intelligence. Sensors, 25.","DOI":"10.20944\/preprints202411.2240.v1"},{"key":"ref_28","doi-asserted-by":"crossref","unstructured":"Jayanth, R., Gupta, N., and Prasanna, V. (2024). Benchmarking Edge AI Platforms for High-Performance ML Inference. arXiv.","DOI":"10.1109\/HPEC62836.2024.10938499"},{"key":"ref_29","doi-asserted-by":"crossref","first-page":"30625","DOI":"10.1038\/s41598-025-16043-z","article-title":"Real time wire rope detection method based on Rockchip RK3588","volume":"15","author":"Qian","year":"2025","journal-title":"Sci. Rep."},{"key":"ref_30","unstructured":"ONNX Runtime Developers (2026, July 28). ONNX Runtime. Version: 1.23.2. Available online: https:\/\/onnxruntime.ai\/."},{"key":"ref_31","unstructured":"(2026, July 28). RKNN Toolkit2. Version: 2.3.2. Available online: https:\/\/github.com\/airockchip\/rknn-toolkit2."},{"key":"ref_32","unstructured":"Case Western Reserve University Bearing Data Center (2026, June 21). CWRU Dataset. Available online: https:\/\/engineering.case.edu\/bearingdatacenter\/download-data-file."},{"key":"ref_33","unstructured":"(2026, July 21). Machinery Failure Prevention Technology Society. Available online: https:\/\/www.mfpt.org\/fault-data-sets\/."},{"key":"ref_34","unstructured":"Bai, S., Kolter, J.Z., and Koltun, V. (2018). An Empirical Evaluation of Generic Convolutional and Recurrent Networks for Sequence Modeling. arXiv."}],"container-title":["Sensors"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/1424-8220\/26\/16\/5125\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,13]],"date-time":"2026-08-13T11:36:53Z","timestamp":1786621013000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/1424-8220\/26\/16\/5125"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,8,13]]},"references-count":34,"journal-issue":{"issue":"16","published-online":{"date-parts":[[2026,8]]}},"alternative-id":["s26165125"],"URL":"https:\/\/doi.org\/10.3390\/s26165125","relation":{},"ISSN":["1424-8220"],"issn-type":[{"value":"1424-8220","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,8,13]]}}}