ARGOS mobile

ARGOS mobil is an AI-based, mobile traffic counting system for the precise recording of traffic flows. It uses artificial intelligence directly on the device (edge ​​computing) to recognize a total of 13 classes, including pedestrians and specific bicycle classes. Its key feature is the high counting accuracy at intersections and in complex traffic situations.

The system was developed for very complex traffic situations, which occur especially in cities with increasing mixed traffic, to quickly provide reliable data to mobility and traffic planners.

Argos mobil can capture all traffic flows within a 47 m diagonal with its wide-angle camera. This includes 10-lane roads but especially intersections, roundabouts, and mixed traffic. For larger scenes, such as intersections with 10 lanes, 2 Argos systems can be deployed mobile.

Argos 1 can distinguish between pedestrians, cyclists, cars, cars with trailers, vans with and without windows, trucks, trucks with trailers, articulated lorries, and busses.

In Argos 2, the class set (see Argos 1) is further expanded to include cargo bikes, bikes with trailers, and scooters.

Traffic counts are required at regular intervals in many countries, including Germany. They form the basis for mobility planning, tenders, and traffic regulations.

The data is needed for mobility and urban development, traffic light phase control, noise mapping, infrastructure and capacity planning, road condition assessment, safety analyzes, and site analyzes.

These are very important data for every city, municipality, the state, and the federal government. The highest data quality and most complex situations can be captured with Argos mobile. For simpler cross-sections, a side radar measurement system (SDR) is sufficient.

Customers have 90 days after a measurement to modify the polygons and measurement fields, thereby ensuring the accuracy of their measurement, which is a problem with many systems. If the zone was set up incorrectly, the data may be wrong. But not with Argos mobile. The user has everything in their hands and can thus also determine the data quality. No trained personnel is needed for the setup anymore. The mobility planner can conveniently make the settings later from the office or home office.

There are two battery cases: a small one for 24 hours and a large one that lasts 3 to 7 days depending on the battery.

The system is attached to existing infrastructure and can be extended up to 6 meters.

Data processing is done on-board (directly in the camera). A connection (WiFi or LTE) is only needed for configuration, live viewing, or data transfer to the cloud platform myTrafficData2.

Yes, 100%. Since the AI processes the objects in real-time and only stores anonymized metadata (numbers, classes, paths), no personal data is transmitted by default.

The storage of video material is optional. For a purely statistical count, this is not necessary. If videos are stored for validation, this will be done locally on an encrypted storage and can only be accessed by the user.

In comparative tests at the Federal Government's Mobility Lab (including drone footage and other camera manufacturers), ARGOS mobil achieves an extremely high accuracy of > 95% for a DTV at a T-intersection with over 70,000 vehicles. Only the drone footage was slightly better. Tests in cities like Berlin have shown accuracies of over 95% at intersections with a DTV greater than 100,000.

The Argos 1 requires passive light, such as street lighting nearby. Then it achieves the same results as during the day.

The second generation will be delivered with improved nite vision. This can then also deliver similar results without passive light as in daylight. Does the capture also work at nite?