AI Modules

Based on modern machine learning algorithms, we offer several solutions for multiple applications in smart mobility.

Contact Us

Traffic data augmentation

The most common types of traffic data come from roadside sensors, on the one hand, and floating car data on the other. 

Roadside sensors produce a reliable and precise point measurement of the traffic situation at one specific location. In contrast, floating car data covers a much larger region, but cannot produce information about absolute counts and the multimodal nature of traffic. 

With the traffic augmentation module, we offer the best of both worlds. The fusion of roadside equipment data, such as traffic cameras, with floating car data offers a powerful solution to accurately estimate:

- Traffic counts on all parts of the road network (even where we do not have equipment installed)
- Classification of vehicle flows in multiple modalities

Depending on the requirements and the use case, several data sources can be used to provide an accurate measure of the traffic on any road network.

Equipment defect detection & classification

This module monitors and analyses the data coming from roadside equipment to determine whether it is functioning properly.

A common use case consists of ANPR (Automatic Number Plate Recognition) cameras producing less recognitions over time due to their lens/faceplate gradually becoming dirtier

Another use case is to identify sudden anomalies in data, e.g., from an inductive counting loop, that can be a precursor for the equipment breaking down. Several monitoring modules, analyzing both real-time and historical data, were developed to monitor roadside equipment and produce alerts when their behavior suddenly or gradually changes in time.