The Land Transport Authority (LTA) of Singapore is exploring innovative ways to enhance road safety and traffic management by utilizing the cameras installed on the country’s fleet of 6,000 public buses. In a recent call for proposals, the LTA aims to develop a video analytics system that can detect various traffic offenses and infrastructural defects, thereby improving operational efficiency and response times.
The proposed system is designed to identify a range of traffic violations, including vehicles that occupy bus lanes during restricted hours, stop in yellow boxes, fail to yield to buses at priority boxes, or are parked illegally. The technology would need to capture essential details such as the offending vehicle’s model, color, type, and license plate number, as outlined in the tender documents released on October 7 via the government procurement portal, GeBIZ.
In addition to monitoring traffic offenses, the LTA envisions using bus cameras to identify structural defects on roads and surrounding areas. This initiative aims to streamline the current enforcement processes, which are often labor-intensive and reliant on the deployment of enforcement officers and mobile cameras stationed at hotspots for illegal parking and other violations.
Currently, video footage from bus cameras is recorded only during service hours and must be retrieved manually for incident investigations. The LTA’s new system would leverage artificial intelligence (AI) to automate the detection of offenses and defects, providing timely alerts to LTA officers within 24 to 48 hours. The system is expected to operate with minimal involvement from bus captains, ensuring that bus operations remain unaffected.
Experts in AI have welcomed this initiative but also pointed out potential challenges. Professor Pradeep Varakantham from Singapore Management University noted that differentiating between visual clutter—caused by the constant movement of people and vehicles—and actual offenses could pose difficulties. He suggested that additional laser sensors might be necessary to enhance image clarity for accurate detection.
Motorists have expressed optimism about the proposed system. Sam Yew, a regional manager and experienced driver, described the initiative as a “great idea” that could significantly improve road safety and compliance with traffic regulations. However, he also raised concerns about the possibility of AI misidentifying offenses, which could lead to an increase in appeals and additional workload for the LTA. Yew emphasized the importance of implementing stringent data protection measures to safeguard against unauthorized use of the collected data.
As the LTA evaluates the feasibility and reliability of the proposed solutions, the authority remains committed to leveraging technology to enhance traffic management and road safety in Singapore. The next steps will involve live demonstrations from shortlisted companies and a proof-of-concept test lasting four months, paving the way for a potential rollout of this innovative system.