KT to Showcase 'Traffic Flow Optimization Solution' at ITS World Congress

[Asia Economy Reporter Lim Hye-sun] KT announced on the 19th that it will introduce digital twin and artificial intelligence (AI)-based intelligent transportation technologies at ITS World Congress 2022, held in Los Angeles, USA, until the 23rd (Korean time), in collaboration with Bucheon City.

ITS World Congress is an exhibition and academic conference with authority in the field of intelligent transportation systems (ITS), known as the world transportation Olympics. ITS World Congress has been held annually since it began in Paris, France, in 1994, and was held in Korea in Seoul (1998) and Busan (2010). KT will exhibit the ‘KT Road Twin,’ developed independently by its Convergence Technology Institute, applied to the actual traffic environment of Bucheon City, along with the signal system.

KT Road Twin consists of a ‘wide-area traffic simulator’ that virtualizes all real roads and traffic situations, and an ‘AI optimal signal engine’ that learns traffic signal systems in advance to derive optimized signals. To derive optimal signals, it is necessary to learn the changing traffic volume by adjusting signals at more than 16 intersections every second, making the number of cases nearly infinite and the learning itself difficult. KT solved this problem through over three years of AI research and development.

KT also completed a demonstration using KT Road Twin in collaboration with Bucheon City and Bucheon Wonmi Police Station. The AI-derived optimal traffic signals were applied for two weeks from 6 a.m. to 10 p.m. on pilot roads within Bucheon City, resulting in an average traffic throughput increase of 4.75 to 8% compared to before.

KT and Bucheon City expect to reduce approximately 14.7 billion KRW in annual traffic congestion costs and about 1,000 tons of carbon emissions based on the pilot roads. KT explained that if expanded to 288 intersections throughout Bucheon City, it would reduce about 350.5 billion KRW in annual traffic congestion costs and approximately 23,000 tons of carbon emissions.

This content was produced with the assistance of AI translation services.