Background

The Korea Transportation Safety Authority (KTSA) has been working to reduce traffic accidents nationwide. To validate the effectiveness of AI safety solutions, KTSA conducted a pilot project targeting city buses in urban areas.

By adopting AI systems capable of real-time detection of drowsy driving, distraction, mobile phone use, and other risky behaviors, KTSA sought to verify how such solutions could enhance safety outcomes compared to existing programs.

In June 2024, a pilot was conducted across 13 bus operators nationwide, covering 500 buses. The project deployed an AI-driven monitoring system that tracked driver behavior in real time, improving attentiveness and safety indicators while reducing risks.

Solution

A.I.Matics’ Roadscope 9 (2-channel) devices were installed on each vehicle, designed to detect and improve risky driving behaviors through real-time AI alerts and behavioral data collection.

• Real-time monitoring of driver behavior and forward view

• Detection and alerts for drowsy, distracted, and smoking behaviors

• ADAS event detection (unsafe distance, red-light violations, etc.)

• Driver behavior data–based scoring

• LTE-based cloud upload and integrated platform visualization

• Reporting and feedback programs provided for KTSA and operators

Results

The pilot showed that drivers’ behaviors improved in real time, enhancing not only accident prevention but also operational stability. Results were featured on KBS News and acknowledged by public organizations as both credible and effective.

• Accidents reduced by 55%

• Driver safety scores improved by 51% (June: 50.9 → November: 76.9)

• Safety awareness increased by 87%

• Mobile phone use reduced by 99%

• Drowsy driving decreased by 80%

• Dangerous lane changes reduced by 32%

Based on these results, KTSA plans to expand AI safety solutions nationwide in 2025, establishing AI-based driver monitoring as a standard across public bus fleets.