Conveyors are considered as a critical asset in industrial and are most used in many fields like mining, manufacturing, and material industries transport, however unexpected failures can lead a big loss for companies by a long downtime in production lines, in safety hazards, and production losses mainly when it about unsecure places like underground mining projects. This paper presents an intelligent conveyor inspection system that combines infrared thermography and deep learning-based computer vision for the aim of monitoring. The proposed pipeline consists of four main components: the first, data acquisition using thermal and RGB cameras, the second, thermal analysis for the detection of overheated idlers, the third, visual inspection using a convolutional neural network (CNN) benchmarked against a Support Vector Machine (SVM) baseline to identify belt defects such as cracks, tears, and defected material, and the fourth, a monitoring platform that stores inspection data in a centralized database and visualizes operational insights through Power BI dashboards. By integrating thermal and visual information, the system provides a comprehensive assessment of conveyor health and enables real-time detection of fault alerts. Experimental results on a dataset of 10,307 annotated images demonstrate that CNN achieves an accuracy of 98.21% and an AUC of 0.99 during image processing for defect detection against an accuracy of 97.35% and AUC of 0.84 for SVM. Unlike existing conveyor inspection solution that rely on single thermal monitoring or visual inspection independently, the proposed framework combines thermal diagnostics, CNN-based defect detection, centralized monitoring, real-time alerting and business intelligence dashboard into a unified inspection architecture. The solution provides an effective support that used to earlier maintenance and contributes to development of more predictive maintenance systems for industrial conveyor operations