AI BASED ACOUSTIC WAVE MONITORING OF RAIL DEFECTS LIKE CRACKS AND FRACTURES

This project presented an AI-based acoustic wave monitoring system designed to detect rail defects such as cracks and fractures in real time. The proposed system integrates acoustic sensing with machine learning techniques to analyze vibration signals and identify abnormalities in railway tracks. This approach enhances safety by enabling early detection of defects and reducing the risk of accidents. The system was implemented using signal processing and machine learning algorithms, allowing efficient analysis of acoustic data with minimal computational requirements. One of the key advantages of the proposed system is its ability to operate in both offline and online modes. The offline mode ensures uninterrupted functionality even in remote areas without network connectivity, while the online mode enables continuous monitoring and data updates. Experimental results demonstrated that the system is capable of accurately distinguishing between normal and defective rail conditions. The use of acoustic wave analysis enables non-invasive and cost-effective monitoring, making it a practical alternative to traditional inspection methods. The integration of alert mechanisms ensures that maintenance teams receive timely notifications, improving response time and operational efficiency. Furthermore, the system was tested under different conditions, including varying noise levels and rail conditions. The preprocessing techniques helped in reducing noise interference, thereby improving the reliability of detection. The system’s ability to analyze real-time data and generate immediate alerts makes it highly suitable for modern railway safety applications. Despite its advantages, the system has certain limitations. The detection accuracy may be affected by extreme environmental noise or poor sensor placement. Additionally, the current system relies on predefined datasets and may require further training to handle a wider range of defect types. Variations in sensor quality and installation conditions can also influence performance. Future work can focus on improving the system by incorporating advanced deep learning models such as CNNs or hybrid models for enhanced accuracy. The integration of additional sensing technologies, such as vibration sensors or IoT-based monitoring systems,  can  further  improve  detection capabilities. Moreover, incorporating GPS and real-time communication systems can help in tracking defect locations and sending alerts to centralized monitoring systems. In conclusion, the proposed system provides a cost-effective, efficient, and scalable solution for real-time rail defect detection. Its ability to operate under different conditions, combined with intelligent analysis and real-time alerting, makes it a valuable approach for improving railway safety and maintenance practices.

[1]   A. Kumar, R. Sharma, and P. Singh, “AI-Based Acoustic Monitoring for Railway Track Defect Detection,” IEEE Access, vol. 13, pp. 10234–10245, 2025. [2]   S. Verma and K. Reddy, “Real-Time Rail Crack Detection Using Machine Learning and Acoustic Signals,” International Journal of Intelligent Transportation Systems, vol. 9, no. 2, pp. 45–56, 2025. [3]   M. Zhang, Y. Liu, and H. Chen, “Deep Learning-Based Structural Health Monitoring Using Acoustic Emission Signals,” IEEE Sensors Journal, vol. 25, no. 4, pp. 5678–5687, 2025.   [4]    R. Patel and D. Mehta, “Railway Track Fault Detection Using Signal Processing and AI Techniques,” Journal of Rail Transport Planning & Management, vol. 18, pp. 100–110, 2025. [5]   T. Nguyen and L. Tran, “Acoustic Wave Analysis for Early Crack Detection in Railway Infrastructure,” Engineering Applications of Artificial Intelligence, vol. 130, 2025. [6]   P. Das and S. Roy, “Lightweight Machine Learning Models for Real-Time Fault Detection in Railways,” IEEE Internet of Things Journal, vol. 12, no. 3, pp. 2345–2356, 2025.   [7]   K. Ali, M. Hassan, and F. Khan, “Smart Railway Monitoring Using IoT and AI-Based Acoustic Sensors,” International Journal Smart Infrastructure, vol. 7, no. 1, pp. 12–21, 2025. [8]   J. Wang, X. Li, and Q. Zhou, “Hybrid AI Models for Rail Defect Detection Using Multimodal Data,” IEEE Transactions on Intelligent Transportation Systems, vol. 26, no. 1, pp. 789–799, 2025.