부산대학교 도서관

In smart cities, utility infrastructures are strategically placed underground for reasons of convenience, safety, and aesthetics. The aging of these underground utilities emphasizes the importance of accurate mapping for inspections and maintenance. This survey paper extensively explores a range of sensors specifically designed for the detection of underground utilities. The manual operation of these sensors, requiring skilled personnel, is characterized by its slow pace, susceptibility to human errors, and safety concerns in potentially hazardous environments. To overcome these challenges, researchers are actively engaged in developing robotic solutions that seamlessly incorporate machine learning algorithms, aiming to achieve both rapid and accurate underground mapping. The paper offers a comprehensive overview of these robotic solutions, emphasizing their integration with diverse sensors to enable autonomous and efficient mapping of underground infrastructure. The shift toward automated solutions is deemed crucial to meet the growing survey demands in rapidly developing urban environments, ensuring safety, accuracy, and speed in modern underground utility mapping practices.