Traditional fault diagnosis and repair procedures in automotive air-conditioning systems rely heavily on manual inspection, often resulting in time-consuming processes and inconsistent outcomes. This paper proposes a novel, automated fault diagnosis and repair procedure to replace the conventional approach. The proposed system's core involves using an endoscope-based diagnostic tool to capture images from hard-to-reach areas without requiring dashboard disassembly. These images are pre-processed using fundamental image processing techniques and analyzed using deep learning algorithms for automatic fault identification. The detected fault types are then mapped to a newly developed diagnostic framework, enabling accurate and efficient repair actions. Experimental results demonstrate that the proposed method significantly enhances diagnostic accuracy, reduces repair time, and improves overall maintenance quality. This approach lays the groundwork for intelligent and automated maintenance strategies in modern automotive systems.