In this paper, we propose a cyber-physical sensing framework for the real-time monitoring of water pipe integrity using multimodal detection technologies. The system integrates an unmanned vehicle equipped with an embedded camera for internal visual inspection and external microphones for acoustic sensing. The mobile platform enables flexible navigation inside pipes and captures high-quality images for detecting cracks and structural defects. Concurrently, acoustic signals collected from the pipe surface are transformed into the frequency domain via fast Fourier transform, and machine learning classifiers are employed to distinguish normal flow conditions from leakage-induced anomalies. The vision module utilizes a You Only Look Once v9-based detection model trained on an augmented dataset to enhance the accuracy of crack and rupture identification, particularly in large-diameter pipes. By combining acoustic and visual sensing within a unified cyber-physical architecture, the proposed system achieves efficient, low-cost, and scalable pipe condition monitoring. This approach provides the timely detection and localization of damage, offering practical value for reducing water loss and supporting more sustainable urban water infrastructure management.