Seoul to Use AI to Enhance Public Safety on Subways

Seoul recently unveiled its plans for developing an artificial intelligence-based system to monitor and track unusual behavior among subway passengers, including instances of violence, with the goal of completing the system by the end of the year.

The announcement from Seoul Metro and the Seoul Digital Foundation comes in the wake of a series of incidents that have heightened public concern about safety on the subway, including a stabbing incident on Line No. 2 in August that saw two passengers injured, and a rush provoked by a false alarm about a stabbing on the same subway line that injured 18 passengers in September.

The project’s objective is to create a real-time monitoring and detection system for identifying abnormal behavior among subway passengers by utilizing AI-powered “image captioning” technology. This technology translates images or video footage into textual descriptions.

The generated text can then be forwarded to subway officials for immediate action should an incident take place.

The two organizations signed a memorandum of understanding on Oct. 6 for the project and have set the goal of completing the design and technical testing phase by December.

They plan to choose a station on Line No. 5 for initial testing. After successful initial tests, the system will be expanded to cover all stations and subway cars by 2025.

Seoul Metro said that the project is expected to bolster the response capabilities of authorities during critical situations by fostering collaboration between the police and subway security personnel.

“The AI-powered system for detecting and tracking abnormal behavior will evolve into a technology that enables a more effective response to incidents targeting members of the public,” said Baek Ho, CEO of Seoul Metro.

“We are committed to establishing a secure subway environment by harnessing the infrastructure of Seoul Metro and the technical expertise of the Seoul Digital Foundation.”