New AI-powered system to help detect motorists flouting traffic rules like driving in bus lanes
System will use footage captured by cameras on buses, and get rolled out from 4th quarter of 2026.
Singapore's Land Transport Authority (LTA) is set to implement an AI-powered system to detect motorists disregarding traffic rules, such as driving in bus lanes. The system will analyze footage captured by cameras on buses and is scheduled to be rolled out in stages starting in the fourth quarter of 2026. During trials, the system achieved an average accuracy rate of around 90 percent.
The LTA is collaborating with public transport operators to deploy the technology on all bus routes by 2030. The system can identify errant motorists who encroach into bus lanes, park illegally on roads, and other infrastructural issues. During trials, the AI system detected a daily average of seven vehicles encroaching into bus lanes on four buses covering 126 kilometers of bus routes.
In the past three years, the LTA issued an average of 36,800 summonses for bus lane encroachments. LTA Deputy Director for Innovation and Analytics Germain Tay explained that the system can also identify road defects and non-compliant roadworks. The LTA has been issuing safety reminders via audio announcements on buses since August 2025, which has been well-received.
The authority is also planning to launch an automated safety announcement feature in 2027. The LTA's Bus Safety Tripartite Taskforce completed its review in 2025, and its recommendations are being implemented. The taskforce aims to have 99 percent of bus drivers complete the Bus Captain Drive Safe Refresher Programme by October, ensuring all drivers have completed the course by then.
Additionally, driving times for bus drivers on routes exceeding two hours have been reduced to minimize fatigue and enhance safety. The LTA also holds a bus safety roadshow to raise public safety awareness, scheduled from August 21 to August 22.
Written by urgent.news from Straits Times's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
