SK Telecom, South Korea’s largest wireless telecom operator, has developed an inspection system that checks the safety of cell towers using drones and automated image analysis.
The company’s new AI model has been developed by in-house experts and has the capability of determining the status of bolts and nuts by analyzing the images taken by drones.
SK Telecom explains that the cell towers across South Korea equipped with transmitting and receiving antennas can reach a height of up to 75 meters. These cell towers require frequent maintenance to prevent accidents caused by deterioration, such as corrosion or loosening of bolts and nuts, therefore specialized personnel have been climbing up the towers to visually check their condition.
But now with an intelligent safety inspection system in place, not only can SK Telecom prevent accidents due to deteriorating cell towers, but it can also ensure the safety of workers. Moreover, the company can also automatically identify defects by analyzing images taken by drones.
Previously, safety inspectors had to study around 100 images to complete the inspection of one cell tower. With the adoption of the new AI analysis model, SK Telecom has been able to reduce the time required for the process by 95%, while increasing the reliability and consistency of the analysis results.
The company says it will enhance the system even further by adding inspection items such as wind pressure safety/inclination. It will also look to improve the AI model and link the application with the safety management system.
As Park Myung-soon, SKT’s vice president and head of Infra DT Office, says: “By building an intelligent safety inspection system that can complement the existing visual inspection, we have secured greater safety for workers. We will continue to make efforts to achieve AI transformation of our telecommunication networks, while focusing on developing our field workers into experts who can develop and operate AI.”
In addition to drone-based cell tower inspections, the telecom company is actively applying AI to other areas of its network, including equipment error/anomaly detection, power cost reduction, and work completion inspection.
