SK Telecom said on the 1st it has commercialized "Topda," a solution that automatically inspects for anomalies in telecom infrastructure by analyzing footage taken by work vehicles with artificial intelligence (AI).
Topda uses AI cameras installed on SK Telecom work vehicles to detect 11 types of inspection elements, including construction sites, heavy equipment, and defects in line facilities. Because it uses driving footage from existing vehicles dispatched for handling complaints or repairs, there is no need to add separate inspection personnel or vehicles.
SK Telecom trained 22,000 field photos over the past five years, boosting detection accuracy from 65% in 2024 to 88% this year. In a pilot run with six vehicles in the Seoul metropolitan area from May to Aug., at least 50% of key wireline sections' facilities were inspected once. The company plans to expand the inspection scope to 70% in 2027 and over 90% in 2028.
Topda is powered by SK Telecom's in-house AI video analytics platform VISTA. Vision AI recognizes facilities in the footage, context AI determines whether action is actually needed, and spatial AI tracks the location and changes of the inspection targets. Faces and license plates are automatically blurred in the vehicle immediately upon capture.
SK Telecom plans to apply Topda to inspections of wireline equipment inside tunnels by the end of this year and to expand its use to other industrial sites such as gas, power, and oil pipelines.