Multi-lane free-flow (MLFF) tolling depends on a deceptively simple task: reliably identifying vehicles travelling at highway speeds without requiring them to slow down or pass through a physical toll booth. A new camera platform from e-con Systems illustrates how edge AI is being used to bring more of this processing directly to the roadside.

The company has launched TrafficSenz, a 5MP automatic number plate recognition (ANPR) camera that combines image capture, infrared illumination, vehicle triggering and AI processing in a single roadside unit. It is designed to capture vehicles travelling at speeds of up to 300 km/h, with the aim of reducing missed or unusable plate reads that can subsequently create exceptions in tolling transactions.

A global-shutter Sony Pregius S sensor helps avoid the geometric distortion associated with rolling shutters when photographing fast-moving vehicles. The system also uses automatic exposure bracketing, capturing multiple exposures to deal with difficult high-contrast situations such as headlights and reflective number plates.
Night-time operation is supported by synchronised 850 nm pulsed infrared illumination. According to the manufacturer, the system can capture readable plates at distances of up to 50 metres, although the actual performance depends on factors including vehicle speed, lens configuration and environmental conditions.

An important part of the concept is edge processing. TrafficSenz can perform ANPR, vehicle classification and make-and-model recognition locally rather than sending all imagery to a remote platform first. It can also detect approaching vehicles directly from the image and trigger capture without separate external hardware.

This approach could simplify roadside installations while reducing the amount of raw video that needs to be transmitted across the network. Instead, the camera can send processed results and relevant evidence to higher-level tolling or traffic-management systems. Support for ONVIF also facilitates integration with wider ITS architectures.

The development reflects a broader trend in intelligent transport infrastructure: cameras are evolving from passive recording devices into distributed computing nodes. For tolling, that means moving closer to a model where detection, interpretation and initial transaction validation happen at the roadside, while central platforms concentrate on management, evidence and revenue processes.

The original article: here.

STONE, Tom. e-con Systems launches edge-AI ANPR camera for free-flow tolling. Traffic Technology Today [online]. 2026_10_01 [cited 2026-10-06]. Available: https://www.traffictechnologytoday.com/news/tolling/e-con-systems-launches-edge-ai-anpr-camera-for-free-flow-tolling.html

Picture: chat gpt