A team of researchers from the Korea Institute of Civil Engineering and Building Technology (KICT) has created a pothole identification program to expedite pothole repairs.

The "Road Pothole Filtering Program" was established to create an emergency response restoration system for the occurrence of potholes on roadways, which can damage automobiles and even cause traffic accidents.

Source: KICT

As such, the KICT team developed the new server-based pothole filtering program, which the team intends to add to the Public Data Management System (PDMS). According to developers, the program is capable of reviewing initial pothole information detected by the pothole detection program.

As an enhancement to a pothole detection program operated by the Land Management Office, under the Ministry of Land, Infrastructure and Transport, which is a mobile-based application for detecting potholes, the researchers have implemented a technique for filtering out objects that tend to resemble potholes like tires, lane markings and shadows — which are difficult to discern from actual potholes using mobile devices alone.

To make those distinctions, the KICT team has created and trained an artificial intelligence (AI) algorithm to disregard objects other than potholes from the pothole information initially transmitted to the server. Once distinguished, that verified pothole data is then automatically transmitted to personnel in the Land Management Office every three hours. The KICT team suggests that the new grogram will expedite the processing time of emergency pothole restoration on roadways.