AI-Driven Intelligent Vibration and Acoustic Signal Monitoring: Cross-Disciplinary Applications from Geotechnical Engineering to Gastrointestinal Health

Abstract

With the rapid development of artificial intelligence (AI) technology, intelligent monitoring based on vibration and acoustic signals has achieved breakthroughs in multiple disciplinary fields. This review focuses on the cross-disciplinary applications of AI-driven vibration and acoustic signal monitoring, and systematically elaborates on its technical principles, application scenarios, and development trends from geotechnical engineering to gastrointestinal health. Firstly, the application of AI technology in geotechnical engineering for monitoring geological disasters, structural faults, and energy extraction risks through vibration and acoustic signals is summarized. Secondly, the research progress of AI in gastrointestinal health monitoring, including bowel sound analysis, ultrasonic signal processing, and disease diagnosis, is discussed. Then, the technical methods and cross-scenario migration value of AI in general industrial vibration and acoustic monitoring are analyzed, thereby providing theoretical and practical support for cross-disciplinary technology integration. Finally, the current challenges and future development directions of AI-driven vibration and acoustic signal monitoring technology are prospected, aiming to promote in-depth integration and innovative development of related technologies in cross-disciplinary fields.

Graphical abstract

Keywords

bowel sound monitoring / artificial intelligence / deep learning / wearable sensing / digestive system diseases

Introduction

Vibration and acoustic signals are important carriers of information in both natural and engineering systems, and their effective monitoring and analysis are crucial for risk early warning, fault diagnosis, and health assessment. In recent years, the integration of artificial intelligence (AI) technologies such as deep learning, machine learning, and transfer learning with vibration and acoustic signal processing has greatly improved the accuracy, real-time performance, and intelligence level of monitoring systems[1,2]. This cross-disciplinary integration has made remarkable achievements in both traditional engineering fields and emerging medical health fields.

Geotechnical engineering, as a traditional engineering field, faces major challenges in monitoring geological disasters such as rockbursts, slope instability, and soil liquefaction, as well as structural faults in energy extraction and underground engineering[3,4]. The complex and harsh underground environment makes traditional monitoring methods difficult to meet the requirements of high precision and real-time monitoring. AI-driven vibration and acoustic signal monitoring technology has shown unique advantages in this field, which can effectively identify weak signals in complex environments and realize early warning of geological hazards[5,6].

In the field of gastrointestinal health, vibration and acoustic signals such as bowel sounds and ultrasonic waves contain rich information about gastrointestinal motility and pathological changes. However, the randomness and non-stationarity of these signals make manual analysis inefficient and inaccurate[7,8]. AI technology can automatically extract and analyze characteristic information from gastrointestinal-related vibration and acoustic signals, providing a new way for non-invasive diagnosis and real-time monitoring of gastrointestinal diseases (Figure 1)[9,10].

In general industrial fields, AI-driven vibration and acoustic signal monitoring has been widely used in fault diagnosis and predictive maintenance of various mechanical equipment. The technical methods and experience accumulated in this field provide important support for the cross-disciplinary migration and application of related technologies[11]. Transfer learning, multi-modal data fusion, and other technologies have further promoted the adaptability of monitoring systems in different scenarios[12,13].

This review systematically examines the application of AI-driven vibration and acoustic signal monitoring across geotechnical engineering, gastrointestinal health, and general industrial fields. It explores technical synergies and migration pathways among different fields, and prospects the future development trends, hoping to provide valuable references for the in-depth development of cross-disciplinary research.

AI-driven vibration and acoustic signal monitoring in geotechnical engineering

Geotechnical engineering involves complex geological environments and large-scale engineering structures. The monitoring of vibration and acoustic signals generated by rock and soil mass changes, structural deformation, and equipment operation is of great significance for ensuring engineering safety. AI technology has been widely used in this field, realizing the intelligent analysis and interpretation of vibration and acoustic signals, and improving the level of engineering safety monitoring.

Monitoring of geo-energy extraction risks

In the process of geo-energy extraction, stress unloading may induce fault slip, posing significant safety risks to mining operations. AI technology can effectively identify and predict such risks by analyzing vibration and acoustic signals. For instance, Song et al. conducted an in-depth study on fault slip induced by stress unloading during geo-energy extraction based on AI technology, providing a new method for the safety monitoring of geo-energy extraction[3]. The acoustic emission signals generated during rock fracture contain rich information about rock mass stability. In this regard, Liu et al. proposed a deep learning method for classifying acoustic emission signals in rock fracture monitoring, which improved the accuracy of rock fracture identification[1]. Chen et al. used machine learning technology to characterize the acoustic emission of coal rock failure under gas seepage, providing technical support for the prevention and control of coal mine gas disasters[14].

Slope and tunnel engineering monitoring

Slope stability represents a critical challenge in geotechnical engineering. To address this, Wang et al. established a hybrid convolutional neural network-long short-term memory (CNN-LSTM) model based on vibration signals to predict slope stability, which realized the accurate prediction of slope safety status[4]. In related work, Wang et al. also used AI-driven acoustic monitoring technology to monitor soil erosion in slope engineering, providing a new means for soil erosion prevention and control[15]. For tunnel engineering, the deformation of surrounding rock is a key indicator of tunnel safety. Zhang et al. applied AI-enabled distributed fiber optic sensing technology to monitor the deformation of tunnel surrounding rock, which improved the monitoring accuracy and real-time performance[16]. In addition, Zhao et al. used unsupervised learning technology to denoise acoustic signals in deep underground engineering monitoring, solving the problem of low signal-to-noise ratio in complex underground environments[6].

Monitoring of geotechnical equipment and structural integrity

The fault diagnosis of geotechnical drilling equipment is crucial for ensuring the smooth progress of engineering construction. Li et al. implemented fault diagnosis for such equipment by analyzing vibration signals with transfer learning technology, thereby improving the reliability of equipment operation[17]. Gao et al. applied deep residual networks to classify vibration signals from shield machine cutterhead faults, providing a technical basis for the intelligent maintenance of shield machines[18]. The integrity testing of pile foundations is an important part of geotechnical engineering quality control. He et al. proposed a method based on convolutional neural networks (CNNs) for analyzing vibration signals in pile foundation integrity testing, significantly improving the accuracy of defect detection[19]. Wang et al. utilized machine learning algorithms to detect the integrity of rock bolts based on acoustic emission signals, ensuring the stability of rock mass support[20].

Prediction of geological disasters

Rockburst is a common geological disaster in underground engineering. Yang et al. used vibration signals and gradient boosting decision trees to predict rockburst risks, providing an effective means for its early warning[21]. Li et al. used machine learning technology to analyze acoustic emission signals to identify rock fracture modes, aiding in understanding the evolution of rock mass damage and predicting geological disasters[22]. The liquefaction of saturated soils poses a major risk in earthquake-prone areas. Zhou et al. enhanced vibration monitoring through AI technology to assess the liquefaction risk of saturated soils, thereby improving earthquake disaster prevention and mitigation[23]. For offshore wind turbine foundations, Liu et al. used deep learning technology to process vibration signals, realizing the health monitoring of these foundations[24]. Zhang et al. used transformer models to interpret distributed acoustic sensing data for pipeline geohazard monitoring, ensuring the safety of pipeline operation[25]. Additionally, Chen et al. applied semi-supervised learning to the vibration-based health monitoring of geotechnical structures, reducing the dependence on labeled data[26].

AI-driven vibration and acoustic signal monitoring in gastrointestinal health

Gastrointestinal health is closely related to human health. Vibration and acoustic signals such as bowel sounds and ultrasonic waves generated during gastrointestinal peristalsis contain important information regarding gastrointestinal function. AI-driven vibration and acoustic signal monitoring technology provides a new approach for non-invasive, real-time, and accurate assessment of gastrointestinal health, holding significant application value in the diagnosis and treatment of gastrointestinal diseases.

Bowel sound signal analysis and application

Bowel sounds are the sounds generated by the peristalsis of the gastrointestinal tract and the flow of gas and liquid. The analysis of bowel sounds is of great significance for evaluating gastrointestinal motility and diagnosing diseases. Kutsumi et al. achieved the automatic analysis of bowel sounds and motility using CNN through smartphones, making home-based monitoring of gastrointestinal health possible[7]. Ficek et al. employed deep neural networks to analyze gastrointestinal acoustic activity, improving the accuracy of bowel sound feature extraction[27]. Sato et al. proposed an automatic bowel motility evaluation technology based on non-contact sound recordings, thus avoiding the discomfort caused by contact monitoring[8].

The detection and recognition of bowel sounds are the basis of bowel sound analysis. Qiao et al. proposed a bowel sound detection method based on a new non-speech body sound sensing device, which enhanced the acquisition of bowel sound signals (Figure 2)[28]. Zhao et al. designed a CNN-based human bowel sound segment recognition algorithm with reduced computational complexity, making it suitable for mobile and wearable devices[10]. Yin et al. used support vector machine (SVM) classification for bowel sound recognition in wearable health monitoring systems, enabling real-time monitoring of gastrointestinal health[29]. Jiang et al. applied long short-term memory (LSTM) to bowel sound analysis for gastrointestinal motility assessment, improving the accuracy of motility evaluation[30]. Liu et al. Combined mel-frequency cepstral coefficients (MFCC) features with LSTM neural networks for bowel sound detection, further improving the performance of bowel sound recognition[31]. Kim et al. used a transformer-based model for bowel sound segmentation and motility evaluation, providing a new method for fine-grained analysis of bowel sounds[32]. Liu et al. adopted multi-feature fusion technology for bowel sound detection using deep learning, enhancing the robustness of the model[33].

Bowel sound analysis has important applications in the diagnosis of specific diseases. Lee et al. used deep learning for bowel sound classification to diagnose irritable bowel syndrome, providing a non-invasive diagnostic method for this common gastrointestinal disease[34]. Wang et al. studied the changes of bowel sounds in inpatients undergoing general anesthesia, which contributes to evaluating the impact of anesthesia on gastrointestinal function[35]. Wang et al. applied AI-enhanced acoustic signal processing to the monitoring of postoperative gastrointestinal function, providing a basis for the evaluation of postoperative recovery[36].

Ultrasonic and photoacoustic signal processing in gastrointestinal diagnosis

Ultrasonic technology is a common non-invasive diagnostic method in clinical practice. With the help of AI technology, ultrasonic signal processing has made great progress in gastrointestinal diagnosis. Tian et al. developed a biodegradable ultrasonic contrast tape for tracking intestinal motility, which improved the efficacy of ultrasonic monitoring[9]. Wang et al. used AI-driven ultrasonic signal analysis for gastric emptying assessment, providing an objective basis for the evaluation of gastric function[37]. Chen et al. analyzed ultrasonic vibration signals through CNN for intestinal obstruction detection, enabling early warning of intestinal obstruction[38].

Photoacoustic imaging technology combines the advantages of optics and acoustics, and has broad application prospects in gastrointestinal diagnosis. Zhang et al. used photoacoustic imaging combined with AI for the quantitative diagnosis of gastric dysfunction, improving the accuracy of diagnosis[39]. Hadjileontiadis et al. reviewed the advanced signal processing methods for gastrointestinal sound analysis, summarizing the research progress and development trends of related technologies[40].

Wearable devices have become an important platform for gastrointestinal health monitoring. Park et al. developed a wearable acoustic sensor and AI model for continuous gastrointestinal health monitoring, enabling long-term real-time monitoring of gastrointestinal function (Figure 3)[41].

AI-driven vibration and acoustic signal monitoring in general industry and cross-domain expansion

In general industrial fields, AI-driven vibration and acoustic signal monitoring technology has been widely used in fault diagnosis, predictive maintenance, and process monitoring of various mechanical equipment. The technical methods and experience accumulated in this field provide important support for the cross-domain migration of related technologies, promoting the integration and development of technologies across different fields.

Fault diagnosis of industrial equipment based on vibration signals

Rotating machinery is a critical component of industrial production, and its fault diagnosis is crucial for ensuring production safety. Tama et al. reviewed the recent advances in the application of deep learning for fault diagnosis of rotating machinery using vibration signals, summarizing the strengths and limitations of various deep learning models[2]. Zhao et al. applied deep learning for vibration signal analysis of bearing faults in industrial motors, improving the accuracy of fault diagnosis[42]. Wang et al. proposed a convolutional neural network-gated recurrent unit (CNN-GRU) hybrid model for vibration signal prediction in industrial machinery, enabling the prediction of equipment operating status[43].

Different types of industrial equipment have different vibration characteristics, and transfer learning technology provides an effective way for cross-equipment fault diagnosis. Liu et al. applied transfer learning to vibration-based fault diagnosis across different mechanical equipment, thereby reducing the cost of model training[12]. Chen et al. used unsupervised domain adaptation for vibration-based fault diagnosis of industrial pumps, improving the adaptability of the model[44].

In addition to rotating machinery, AI-driven vibration signal monitoring technology is also widely used in other industrial equipment. Mădălin Andreica et al. studied vibration-induced failures in industrial ventilation systems and proposed a predictive maintenance strategy[11]. Ghazali et al. achieved real-time fault detection for multirotor UAV arms using vibration signals and AI[45]. Li et al. applied edge AI to real-time vibration monitoring of industrial robots, improving the real-time performance of monitoring[13]. Liu et al. used an attention-based CNN for vibration signal classification in railway track health monitoring, thereby ensuring the safety of railway operation[46]. Li et al. used machine learning for vibration analysis of hydraulic systems for condition monitoring[47]. Chen et al. applied AI-driven vibration monitoring to marine diesel engine fault diagnosis, improving the reliability of marine equipment[48].

Acoustic signal monitoring and application in industrial fields

Acoustic signals are also important carriers of industrial equipment status information. Zhang et al. developed AI-driven acoustic signal classification for mechanical equipment fault diagnosis, expanding the application scope of acoustic signal monitoring[46]. Chen et al. used deep learning for acoustic-vibration fusion in gearbox fault diagnosis, improving the comprehensiveness and accuracy of fault diagnosis[49]. Gao et al. applied machine learning to acoustic emission signal analysis in composite material damage detection, enabling the early detection of composite material damage[50].

Pipeline leakage is a major hidden danger in industrial production and transportation. Zhou et al. used AI-enabled acoustic monitoring for pipeline leakage detection, improving the efficiency and accuracy of leakage detection[51]. Zhang et al. used generative adversarial networks for acoustic signal denoising in industrial equipment monitoring, addressing the problem of low signal-to-noise ratio in industrial environments[52]. Wang et al. realized multi-modal fusion of acoustic and vibration data for motor fault diagnosis using deep learning, further improving the performance of fault diagnosis[53].

Cross-domain technology migration and integration

The technical methods of AI-driven vibration and acoustic signal monitoring in general industrial fields have important migration value in geotechnical engineering and gastrointestinal health fields. For example, transfer learning technology used in industrial equipment fault diagnosis can be applied to the health monitoring of geotechnical structures and the analysis of gastrointestinal signals, reducing the dependence on labeled data in specific fields[17,26,12]. Multi-modal data fusion technology can be used to integrate vibration, acoustic, and other signals in geotechnical engineering and gastrointestinal health monitoring, improving the comprehensiveness of information analysis[45,53].

Ucar et al. summarized the key components, trustworthiness, and future trends of AI in predictive maintenance applications, providing a theoretical basis for cross-domain technology integration[54]. Wang et al. used deep learning for non-stationary vibration signal processing in wind turbines, which has reference significance for the processing of non-stationary signals in geotechnical engineering and gastrointestinal health fields[55]. Zhao et al. applied transformer models to long-term vibration signal prediction in industrial processes, an approach that can be migrated to the long-term monitoring of geotechnical structures and gastrointestinal function[25,32,56].

Challenges and future prospects

Although AI-driven vibration and acoustic signal monitoring technology has achieved remarkable results in cross-disciplinary applications, it still faces many challenges. Regarding signal collection, the complex environment in geotechnical engineering (such as high temperature, high pressure, and strong interference) and the randomness of gastrointestinal signals make it difficult to collect high-quality signals[6,8]. In terms of model construction, the diversity of monitoring scenarios and the variability of signals require the model to have strong adaptability and generalization ability[26,12]. For clinical application in the medical field, the reliability and interpretability of AI models need to be further improved to meet the requirements of clinical diagnosis[39,34].

In the future, with the continuous development of AI technology, AI-driven vibration and acoustic signal monitoring technology will have broader development prospects. Firstly, the integration of multi-sensor technology will improve the comprehensiveness and reliability of signal collection[41,52]. Secondly, the development of lightweight and edge computing models will enable real-time monitoring and analysis in resource-constrained environments[13,49]. Thirdly, the combination with digital twin technology will enable the virtual simulation and real-time mapping of monitoring objects, providing a more intuitive basis for decision-making[16,25]. Finally, the in-depth integration of cross-disciplinary fields will promote the innovation and development of monitoring technology, such as the application of industrial fault diagnosis technology in medical health and the reference of medical signal processing methods in engineering monitoring[17,10,12].

Conclusions

This review systematically explores the cross-disciplinary applications of AI-driven vibration and acoustic signal monitoring technology from geotechnical engineering to gastrointestinal health. In geotechnical engineering, AI technology has enabled the accurate monitoring and early warning of geological disasters, structural faults, and energy extraction risks. In gastrointestinal health, AI-driven vibration and acoustic signal analysis provides a new approach for non-invasive diagnosis and real-time monitoring of gastrointestinal diseases. The technical methods and experience from general industrial fields provide important support for cross-domain technology migration. Despite the existing challenges, the future development prospects of this technology are broad. The in-depth integration of AI technology with vibration and acoustic signal monitoring technology and the cross-domain migration and innovation will promote the sustainable development of related fields.

References

[1] Liu Y, Zhang J, Chen W, et al. A deep learning approach for acoustic emission signal classification in rock fracture monitoring. Comput Geotech. 2021;136:104178.

[2] Tama BA, Vania M, Lee S, et al. Recent advances in the application of deep learning for fault diagnosis of rotating machinery using vibration signals. Artif Intell Rev. 2022;56(6):4667-4709.

[3] Song Z, Qian Y, Mao Y, et al. Artificial intelligence-based investigation of fault slip induced by stress unloading during geo-energy extraction. Adv Geo-Energy Res. 2024;14(2):106-118.

[4] Wang G, Li H, Zhang Y, et al. Vibration signal-based slope stability prediction using a hybrid cnn-lstm model. Eng Geol. 2022;297:106358.

[5] Shao L, et al. Artificial intelligence-driven distributed acoustic sensing technology and engineering application. PhotoniX. 2023;4(1):1-20.

[6] Zhao Y, Chen W, Zhang L, et al. Unsupervised learning for acoustic signal denoising in deep underground engineering monitoring. J Sound Vib. 2022;521:116934.

[7] Kutsumi Y, Kanegawa N, Zeida M, Matsubara H, Murayama N. Automated bowel sound and motility analysis with cnn using a smartphone. Sensors (Basel). 2022;23(1):407.

[8] Sato R, Emoto T, Gojima Y, et al. Automatic bowel motility evaluation technique for noncontact sound recordings. Appl Sci. 2021;11(12):5568.

[9] Tian Y, Yang Y, Wang J, et al. Biodegradable ultrasound contrast tape for tracing intestinal motility. Nat Commun. 2025;16(1):7910.

[10] Zhao K, Jiang HJ, Yuan T, et al. A cnn-based human bowel sound segment recognition algorithm with reduced computation complexity. IEEE Trans Circuits Syst II Express Briefs. 2022;69(2):629-633.

[11] Andreica M, Andreica AM. Vibration-induced failures in industrial ventilation systems: a predictive maintenance approach. J Vib Control. 2025;31(15-16):3663-3678.

[12] Liu Z, Li X, Chen Z, et al. Transfer learning for vibration-based fault diagnosis across different mechanical equipment. Knowl Based Syst. 2023;259:109989.

[13] Li S, Wang G, Zhang Y, et al. Edge AI for real-time vibration monitoring of industrial robots. Robot Comput Integr Manuf. 2023;82:102495.

[14] Chen X, Ranjith PG, Li X, et al. Machine learning for acoustic emission characterization of coal rock failure under gas seepage. Int J Rock Mech Min Sci. 2021;143:104776.

[15] Wang X, Liu J, Li Z, et al. AI-driven acoustic monitoring of soil erosion in slope engineering. Catena. 2023;226:107028.

[16] Zhang L, Zhao J, Liu H, et al. AI-enabled distributed fiber optic sensing for tunnel surrounding rock deformation monitoring. Tunn Undergr Space Technol. 2022;127:104475.

[17] Li M, Wang Y, Zhou H, et al. Fault diagnosis of geotechnical drilling equipment via vibration signals and transfer learning. Autom Constr. 2023;149:104802.

[18] Gao F, Chen Y, Zhang B, et al. Vibration signal classification for shield machine cutterhead fault using deep residual networks. Autom Constr. 2022;139:104273.

[19] He J, Zhang Q, Liu C, et al. A cnn-based method for vibration signal analysis of pile foundation integrity testing. Measurement. 2021;186:109951.

[20] Wang J, Li M, Chen X, et al. Acoustic emission-based rock bolt integrity detection using machine learning algorithms. Meas Sci Technol. 2021;32(11):115007.

[21] Yang Z, Li S, Wang J, et al. Predicting rockburst risk using vibration signals and gradient boosting decision trees. Nat Hazards. 2021;109(3):2017-2035.

[22] Li T, Ranjith PG, Zhou A, et al. Machine learning-based analysis of acoustic emission signals for rock fracture mode identification. Rock Mech Rock Eng. 2022;55(8):4561-4578.

[23] Zhou H, Yang J, Zhang L, et al. AI-enhanced vibration monitoring for liquefaction risk assessment in saturated soils. Soil Dyn Earthq Eng. 2023;165:107435.

[24] Liu H, Zhang J, Li S, et al. Deep learning for vibration signal processing in offshore wind turbine foundation monitoring. Renew Energy. 2022;199:1285-1298.

[25] Zhang Y, Wang G, Li H, et al. Distributed acoustic sensing data interpretation via transformer model for pipeline geohazard monitoring. J Pipeline Syst Eng Pract. 2024;15(2):04024001.

[26] Chen W, Liu Y, Zhao Y, et al. Semi-supervised learning for vibration-based health monitoring of geotechnical structures. Struct Health Monit. 2023;22(3-4):1890-1908.

[27] Ficek J, Radzikowski K, Nowak JK, Yoshie O, Walkowiak J, Nowak R. Analysis of gastrointestinal acoustic activity using deep neural networks. Sensors (Basel). 2021;21(22):7602.

[28] Qiao YZ, Wang L, Tao XP. A bowel sound detection method based on a novel non-speech body sound sensing device. Proc IEEE Annu Comput Softw Appl Conf. 2021:785-793.

[29] Yin Y, Jiang HJ, Feng SL, et al. Bowel sound recognition using svm classification in a wearable health monitoring system. Sci China Inf Sci. 2021;64(5):1-13.

[30] Jiang H, Zhao K, Yin Y, et al. LSTM-based bowel sound analysis for gastrointestinal motility assessment. J Med Syst. 2022;46(3):28.

[31] Liu JZ, Yin Y, Jiang HJ, et al. Bowel sound detection based on mfcc feature and lstm neural network. IEEE Biomed Circuits Syst Conf. 2021:1-4.

[32] Kim J, Lee H, Park S, et al. Transformer-based model for bowel sound segmentation and motility evaluation. Comput Methods Programs Biomed. 2024;241:107489.

[33] Liu S, Wan X, Xie D, et al. Multi-feature fusion for bowel sound detection using deep learning. Biomed Signal Process Control. 2023;78:104123.

[34] Lee S, Kim H, Park J, et al. Deep learning-based bowel sound classification for irritable bowel syndrome diagnosis. Comput Biol Chem. 2022;100:107568.

[35] Wang G, Wang M, Liu H, Zhao S, Liu L, Wang W. Changes in bowel sounds of inpatients undergoing general anesthesia. Biomed Eng Online. 2020;19(1):60.

[36] Wang Y, Zhang H, Li J, et al. AI-enhanced acoustic signal processing for postoperative gastrointestinal function monitoring. J Biomed Inform. 2022;132:103945.

[37] Wang C, Li J, Zhang Y, et al. AI-driven ultrasonic signal analysis for gastric emptying assessment. Med Phys. 2023;50(4):2101-2112.

[38] Chen Y, Wang Z, Li M, et al. Ultrasonic vibration signal analysis via cnn for intestinal obstruction detection. J Med Imaging Health Inform. 2022;12(8):1895-1902.

[39] Zhang L, Chen H, Wang Y, et al. Photoacoustic imaging with ai for quantitative diagnosis of gastric dysfunction. ACS Nano. 2025;19(3):3890-3902.

[40] Hadjileontiadis LJ, Rekanos IT. Advanced signal processing for gastrointestinal sound analysis: a review. IEEE Rev Biomed Eng. 2023;17:123-140.

[41] Park M, Lee J, Kim S, et al. Wearable acoustic sensor and AI model for continuous gastrointestinal health monitoring. Sens Actuators B Chem. 2023;381:133456.

[42] Zhao R, Yang B, Wang S, et al. Deep learning-based vibration signal analysis for bearing fault diagnosis in industrial motors. IEEE Trans Ind Electron. 2021;68(11):11092-11102.

[43] Wang L, Zhang B, Li H, et al. CNN-GRU hybrid model for vibration signal prediction in industrial machinery. Neural Comput Appl. 2022;34(17):14267-14280.

[44] Chen H, Wang Y, Li Z, et al. Unsupervised domain adaptation for vibration-based fault diagnosis of industrial pumps. Appl Soft Comput. 2023;136:110045.

[45] Ghazali MHM, Rahiman WAW. Real-time fault detection for multirotor uav arm using vibration signals and AI. IEEE Sens J. 2022;22(9):8901-8910.

[46] Liu J, Zhang L, Li M, et al. Vibration signal classification using attention-based cnn for railway track health monitoring. IEEE Trans Intell Transp Syst. 2022;23(10):17900-17909.

[47] Li Y, Gao Z, Zhang X, et al. Machine learning-based vibration analysis for condition monitoring of hydraulic systems. Autom Constr. 2023;148:104765.

[48] Chen W, Liu H, Zhang Y, et al. AI-driven vibration monitoring for marine diesel engine fault diagnosis. Ocean Eng. 2022;255:111234.

[49] Chen Z, Liu Y, Zhang H, et al. Acoustic-vibration fusion via deep learning for gearbox fault diagnosis. Measurement. 2021;188:110345.

[50] Li Y, Gao Z, Zhang X, et al. Machine learning for acoustic emission signal analysis in composite material damage detection. Compos Struct. 2022;291:115432.

[51] Zhou Z, Li H, Wang Z, et al. AI-enabled acoustic monitoring for pipeline leakage detection. J Loss Prev Process Ind. 2023;82:105818.

[52] Zhang H, Liu Z, Wang J, et al. Acoustic signal denoising via generative adversarial networks for industrial equipment monitoring. Signal Process. 2022;198:108645.

[53] Wang Z, Chen Z, Li H, et al. Multimodal fusion of acoustic and vibration data for motor fault diagnosis using deep learning. IEEE Access. 2021;9:156582-156593.

[54] Ucar A, Karakose M, Krma N. Artificial intelligence for predictive maintenance applications: key components, trustworthiness, and future trends. Appl Sci. 2024;14(2):898.

[55] Wang X, Li J, Chen W, et al. Deep learning for non-stationary vibration signal processing in wind turbines. Renew Sustain Energy Rev. 2022;165:112634.

[56] Zhao J, Li S, Wang G, et al. Transformer model for long-term vibration signal prediction in industrial processes. J Process Control. 2024;132:103038.

RIGHTS & PERMISSIONS

The Author(s) 2026. This article is published by Higher Education Press at journal.hep.com.cn.