Amap has announced the international expansion of its traffic light countdown feature, with plans for a gradual rollout across more countries and regions worldwide.
Launched by Alibaba-owned Amap in May 2022, the traffic light countdown is a product feature that utilizes AI technology to simulate and calculate real-time changes in intersection signal lights. In practical use, whether users are navigating while driving, cycling, walking, or are in cruise or driving modes, they can see a countdown timer for red lights within the Amap app when approaching supported intersections. When traffic at an intersection is heavy, the app will also indicate how many red light cycles are expected before passing through.
To date, the Amap traffic light countdown has been fully implemented across Mainland China, Hong Kong, Macao, and Taiwan, covering nearly 500,000 signalized intersections. This has effectively improved traffic efficiency and safety at these junctions.
In the global navigation market, providing accurate traffic light countdowns is a widely recognized technical challenge. Due to varying standards for traffic signal systems and differing levels of data accessibility from infrastructure across countries, major overseas navigation products have not offered large-scale, precise traffic light countdown services. Amap's pioneering traffic light prediction technology addresses this by using a large AI model to perform comprehensive calculations based on historical traffic data, real-time vehicle flow information, and intersection signal patterns. This enables the simulation of a countdown without relying on direct hardware integration with the traffic lights.
Currently, Amap's traffic light prediction technology has undergone multiple iterations. The Visual Spatio-Temporal Model (VSTM), developed based on spatial intelligence, incorporates visual temporal perception into spatio-temporal calculations. This allows the system to directly perceive real-time dynamics at intersections, moving beyond reliance on traditional spatio-temporal sequence features, thereby enabling more accurate understanding and prediction of traffic changes at complex junctions.
Comments