LARGE-SCALE FIELD VERIFICATION TEST OF SIMPLE PAVEMENT INSPECTION TECHNOLOGY

Abstract

In recent years, comprehensive and periodic inspections and asset management based on the results of these inspections have been actively implemented to steadily realize preventive maintenance of road infrastructure. Pavement maintenance and management, which covers a huge road network, urgently requires the development and application of new technologies to reduce the cost and labor of pavement inspection. In this study, we developed a simple pavement inspection technology using a commercially available camera and deep learning, evaluated its measurement accuracy on a national highway in Hokkaido, and verified the introduction benefits through a large-scale field verification test. As a result, it was found that the accuracy of our technology is equivalent to that of the existing technology. Furthermore, we showed that the cost of periodic pavement inspections and the total number of workers can be reduced by 50 to 80% by using our technology.