Abstract: To address the low efficiency and high safety risks inherent in traditional manual inspection of belt conveyors in bulk cargo terminals, as well as the limitations of existing inspection robots(e.g., monitoring blind spots, high operational costs, and discontinuous coverage), this paper presents a comprehensive intelligent inspection system based on distributed optical fiber sensing technology. A five-layer architecture, “Perception–Network–Platform–Application–User, ” was designed and implemented. The system integrates Distributed Acoustic Sensing(DAS) and Distributed Temperature Sensing(DTS) to continuously acquire vibration and temperature signals along the conveyor line. An intelligent fault diagnosis model for idler damage, belt tearing, and chute blockage was established by combining wavelet threshold denoising, multi-dimensional feature extraction, and a Support Vector Machine(SVM) classifier. Field validation conducted on a 1,255-meter-long belt conveyor system in a large open-pit mine demonstrated that the system achieves real-time, full-line monitoring. It attained a 100% detection rate for idler faults with an overall diagnostic accuracy of approximately 91.8%, reduced the alarm response time for belt tearing to within 5 s, and reliably identified material accumulation in chutes. Compared to conventional inspection robot solutions, the proposed system exhibits distinct advantages, including intrinsic safety, strong interference immunity, lower long-term operational costs, and truly continuous, blind-spot-free monitoring. Beyond enabling deep fault diagnosis of critical conveyor components, the technology demonstrates excellent scalability, offering a viable and effective pathway toward the unmanned and intelligent operation and maintenance of bulk cargo terminal conveying systems.