 |
| An SK Telecom autonomous car drives on a road in Songdo to collect road information necessary to create HD maps. / Courtesy of SK Telecom |
By Jun Ji-hye
SK Telecom will establish a self-driving infrastructure powered by fifth-generation (5G) networks in Songdo, an international business district in the western port city of Incheon, the firm said Monday.
The nation's top mobile carrier signed a business agreement with the Incheon Free Economic Zone (IFEZ) earlier in the day for a project to turn the IFEZ into a smart city.
Under the agreement, the two parties will join hands to set up the 5G-based self-driving infrastructure, nurture related industries, and establish a data hub.
SK Telecom said it plans to create high-definition (HD) maps in an area spanning 132.9 square kilometers, based on the firm's high-end 5G networks.
HD maps, exclusively designed for autonomous vehicles, provide accurate information on the road ahead as well as the surrounding environment.
The maps, which offer information related to traffic lanes, the speed limit and road conditions in real time, are essential for level 4 or higher level autonomous driving systems, according to the company.
The Society of Automotive Engineers (SAE) classifies the capabilities of autonomous vehicles ranging from Level 0 to Level 5.
The Level 4 system does not require human intervention in most situations, though a human driving option remains for difficult circumstances or the preference of the driver, while Level 5 refers to a stage in which human driving is completely eliminated.
SK Telecom said it will build a 5G infrastructure optimized for a self-driving environment by the end of October.
"SK Telecom will provide a blueprint for the future of 5G-based mobility and smart cities beginning with the HD map project in Songdo," said Park Jin-hyo, chief technology officer of SK Telecom.
![[INTERVIEW] SpaceX's Starlink in Kyiv is wake-up call for Seoul](https://img.koreatimes.co.kr/upload/thumbnailV2/dc6a1a7c92114ec298cef2e3b8e04989.jpg/dims/resize/178/optimize)
