“A road like the surface of the moon” is an expression of a British who is filling a pothole every 19 seconds. The UK Transportation Authority has allocated a budget of 500 million pounds in 2021-22 to fill 10 million potholes. 

The United States is estimated to have 55 million potholes as of 2020, with an annual traffic accident handling cost of KRW 1 trillion. India had more than 3,500 traffic accidents caused by potholes in 2020. In Korea, 87% of all roads are expected to age by 2039. There are many road hazards and traffic accidents around the world, but roads are managed in manual ways such as civil complaints and patrols. With the development of fourth industrial technologies like AI, big data, and IoT, the need and interest in smart road management are also increasing.

Dareesoft is a real-time road hazard information service that offers road hazard information as a Service (RiaaS). Dareesoft’s technology detects road hazards that threaten driver safety with AI road analyzer, analyzes data, and provides hazard information to the general public.

The AI Road Analyzer, equipped with ultra-small and ultra-light AI models, detects 12 road hazards such as potholes, erased road markings, falling objects, and garbage in real-time while driving at 110 km/h, and includes driver assistance functions such as ADAS, anti-drowning sensors, and emergency calls. 

The detected hazard information is transmitted to the control system after secondary precise analysis from the cloud-based server.
The hazard information, shared instantly from the control system, is used for hazard notification service and road maintenance. As of the end of November, more than 1.4 million km of roads nationwide were detected, and the cumulative number of hazard information detection exceeded 5 million.

This year, the Korea Expressway Corporation’s highway high-density AI model development project, the Korea Land and Geospatial Informatix Corporation’s digital twin pilot content project, Anyang City and Changwon City conducted a pilot service operation project to build road hazard data, as well as local OEMs, map and navigation companies and insurance companies.

Dareesoft’s AI data pipeline is automatically learning, analyzing, and upgrading itself, said Elizabeth Row, CEO of Dareesoft. “Through this accumulated data and technology competitiveness, we will expand our detection items and detection performance to develop services into essential services in the era of autonomous driving.”

Dr. Ryu Seung-ki, head of the Department of Highway & Transportation Research at the Korea Institute of Civil Engineering and Building Technology, said, “There has been researching at home and abroad on automatic detection of potholes and road cracks by applying AI vision technology, but we are very happy to expand our scope to various projects.”