Two-Month Demonstration at Pangyo Office
Cost Reduction and Lower Hacking Risks
Advancing Infrastructure Development for the 6G Era
SK Telecom announced on the 7th that it has successfully completed a demonstration of autonomous driving robot technology based on its Telco Edge AI infrastructure.
SK Telecom announced on the 7th that it has successfully completed a demonstration of autonomous driving robot technology based on its Telco Edge AI infrastructure. The photo shows SK Telecom researchers working at the Pangyo office. [Photo by SKT]
This demonstration consisted of various research and development efforts aimed at securing core technologies for indoor goods transportation and delivery robots, and was conducted over two months starting in June at the SKT Pangyo office building.
Based on the Telco Edge AI infrastructure established in the office building, SKT enabled autonomous driving robots to perform various tasks. The AI technology was advanced to process diverse information received from sensors such as cameras and IMU (Inertial Measurement Unit) as the robots navigated the complex interior of the building, and related robot technologies and ultra-precise positioning technologies were verified.
For the robots’ autonomous driving, SKT applied its own VLAM (Visual Localization And Mapping) technology. VLAM is an AI-based positioning technology that analyzes camera images mounted on the robot in real time to determine precise locations, supporting stable movement of the robot even in complex environments.
The Telco Edge AI infrastructure linked to the robots played a crucial role in maximizing the accuracy and efficiency of autonomous driving robots, thereby enhancing service performance.
Additionally, the Telco Edge AI infrastructure demonstrated effectiveness in providing high security and scalability while replacing the existing high-cost On-Device Vision AI equipment. The robot solution utilizing edge AI technology processes data collected by the robot immediately at the edge without transmitting it to a central server, making it suitable for privacy protection and data security enhancement.
By using edge AI, since device data is not transmitted to a central server, the risk of hacking or data leakage is reduced, allowing sensitive information to be securely protected. Furthermore, edge AI technology improves efficiency in terms of maintenance cost reduction and scalability by updating only the edge server rather than individual robots.
SKT expects to lower the manufacturing cost of robots and improve battery efficiency while performing complex tasks requiring high-performance AI computation by utilizing the Telco Edge AI infrastructure. As the autonomous driving robots’ object recognition capabilities improve and edge AI processes route calculation tasks in real time, the burden on robot hardware decreases, minimizing battery consumption. This significantly helps extend robot operation time and reduce maintenance frequency.
SKT plans to actively develop various business models based on Telco Edge AI. It is expected to serve as a foundation to lead innovation in diverse fields such as manufacturing, logistics, and services, meeting increasingly complex and diversified market demands in the future.
SKT will build a testbed to provide an environment for testing edge AI-based performance and effects, striving to secure the technological capabilities necessary for designing Telco Edge AI infrastructure suitable for mobile carrier environments. In addition, through research and development activities and technology verification including global standardization, multilateral cooperation, intellectual property acquisition, and academic paper publication, SKT plans to expand the Telco Edge AI ecosystem.
Ryu Tak-gi, head of infrastructure technology at SK Telecom, said, "We will lead the evolution to 6G AI wired and wireless infrastructure that adds new value to infrastructure by converging telecommunications and AI, focusing on technology development, global standardization, and ultra-collaboration."
© The Asia Business Daily(www.asiae.co.kr). All rights reserved.

