Commonly used cable trench localization and navigation methods have suffered from the issue of incorrect object recognition and have been primarily associated with the limitations of maps and markers for navigation and localization. With the introduction of Augmented Reality (AR) for cable trench localization and navigation, it helps to somewhat enhance such challenges, yet the past review highlighted the lack of an in-depth understanding of the AR system’s actual position estimation and the RD estimation. This study adds to the existing perception by assessing AR-based cable trench infrastructure localization and navigation more comprehensibly in real-world situations. We present framework design of cable trench localization and navigation system based on AR from the perspective of realistic. We discuss the localization accuracy and stability evaluation of the system and the key technologies of the system including visual localization, spatial perception, real-time updating and sensor fusion. The study confirms that the AR-based cable trench localization and navigation system has higher localization accuracy compared to traditional methods. The location of cable trench can be recognized accurately and it can determine the accurate navigation path planning accordingly.