The Korea Transportation Safety Authority (hereafter TS) announced that it is working to create a safer bus operating environment for the public by using AI video analysis technology to warn city bus drivers against risky driving behaviors such as signal violations and drowsy driving.

From July to October this year, TS carried out the “AI-based City Bus Safety Monitoring Pilot Project,” aimed at improving risky driving behaviors and reducing accidents.

According to TS, the analysis of risky driving behaviors such as drowsy driving, signal violations, failure to maintain a safe distance, central line encroachment, pedestrian collision risks, use of mobile phones, and smoking among city bus drivers who participated in this project showed that accident rates decreased by 71.2 percent (0.424 → 0.122) compared to the same period last year (July–October 2023).

In particular, the analysis of actual driving records indicated that the number of risky driving behaviors in city buses decreased significantly, with signal violations dropping by 79.4 percent, illegal U-turns by 87.2 percent, and central line encroachment by 38.5 percent. This data was based on 500 city buses from 13 bus companies, normalized per 1,000 kilometers of driving distance.

Furthermore, the number of internal risky behaviors also fell sharply, with drowsy driving reduced by 99.0 percent, smoking by 37.0 percent, and lack of forward attention by 88.5 percent.

In addition, among the buses participating in the pilot project, ADAS (Advanced Driver Assistance System) warning incidents also decreased, with pedestrian collision alerts reduced by 20.5 percent and safe distance alerts by 31.4 percent.

The “AI-based City Bus Safety Monitoring Pilot Project” is an accident-prevention initiative that uses AI video analysis equipment to monitor bus drivers’ risky driving behaviors in real time and provides audible alerts to warn them immediately.

The participating buses were equipped with AI cameras both inside and outside the vehicle. These cameras, combined with video analysis equipment, monitored risky driving behaviors in real time. During driving, when risky behaviors such as signal violations, central line encroachment, use of mobile phones, drowsy driving, and lack of forward attention occurred, the AI terminal inside the vehicle immediately issued an alert.

After driving, data on risky driving behaviors such as speeding, sudden acceleration or deceleration, and near-miss incidents, along with accident video footage, were uploaded to the safety management platform. These records were then provided to operators and used in driver retraining programs.

Moving forward, TS plans to expand the use of AI video analysis technology for safety monitoring beyond city buses, extending to chartered buses and other transportation sectors, in collaboration with the Ministry of Land, Infrastructure and Transport, local governments, and bus operators.

To this end, TS signed MOUs with the Korea Bus Transport Business Association on March 15, A.I.MATICS on March 28, and LG Electronics on August 22.

Meanwhile, in November 2024, TS received the Minister’s Award from the Ministry of the Interior and Safety in the “Digital Governance Innovation” category of the 2024 Government Innovation Excellence Awards for its achievements in bus driver safety management using AI video analysis technology.

TS Chairman Jeong Yong-sik stated, “Through this project, city bus drivers were able to self-correct their risky driving behaviors such as traffic violations and unsafe driving habits, leading to meaningful improvements. Going forward, we will do our best to continue utilizing cutting-edge technologies like AI to create a safer public transportation environment for all citizens.”

Source: Breaknews, columnist Jin Byung-hyu