Technology R&D
VLM in Action: Real-Time Traffic Congestion Detection During Heavy Rain
Heavy rainfall can quickly reduce road visibility, slow traffic, and increase the risk of congestion.
HiPower is applying Vision-Language Models (VLM) to real-world traffic monitoring, enabling existing CCTV infrastructure to move beyond passive surveillance toward intelligent, automated road-condition analysis.
Through ViDA.ai × VLM, live highway images can be analyzed for multiple conditions simultaneously, including weather, traffic density, traffic flow, road conditions, and environmental abnormalities.
Beyond traffic congestion, the same VLM architecture can be extended to detect heavy rain, dense fog, flooding, rockfalls, smoke, and other road hazards, turning existing cameras into intelligent sensing nodes for smarter city operations.
From seeing traffic to understanding what is happening.