Hankook Tire released a record of the Road Hazard Prediction & Detection Solution’s development process through its official YouTube Channel.

Through a partnership with SK Planet, Hankook developed Road Hazard Prediction & Detection Solution that preemptively detects potential hazardous risks on the road. The solution essentially analyzes road noise acquired from a moving vehicle through artificial intelligence (AI) and deep learning technology to identify risk factors such as rainfall, snow and black ice.

The danger of black ice, for instance, is that it is difficult to detect due to its highly transparent characteristics. It can cause life-threatening accidents, particularly on highways. By identifying hazardous conditions in advance, the Road Hazard Prediction & Detection Solution can help reduce the risk of accidents caused by adverse conditions formed through weather and other causes.

During the data mining process, Hankook conducted approximately 200 test drives from 20 individual proving grounds including Geumsan G’Trac in Korea to acquire 600 audio samples. These audio samples were converted into 45,000 units of data which were used in developing the technology and improving accuracy.