**Set of multimodal hourly counting data by displacement mode from thermal sensors.**
The City of Paris collects vehicle counting data by:
* **travel modes** (Trottinettes, Trottinettes + Bicycles (when the distinction between these two modes of travel is not implemented on the sensor), Bikes, 2 motorised wheels, Light vehicles ‘3.5 tonnes, Heavy vehicles’ 3.5 tonnes, Buses & coaches),
* **traffic queue types** (Corona-piste, Bicycle paths, General traffic routes),
* **traffic direction**.
This data is built using a **artificial intelligence algorithm** that analyses images from thermal cameras installed in public space.
** Images from thermal cameras do not identify faces or license plates. The data thus collected does not present personal or individual data.**
No image is transferred or stored on computer servers, the analysis being carried out as close as possible to the thermal camera. Only counting data is transmitted.
This dataset is powered by the metering performed by the sensors and the dataset describing the trajectories of the counting sites [** Multimodal counting – Counting Sites and Trajectories**](https://parisdata.opendatasoft.com/explore/dataset/comptage-multimodal-sites-et-trajectoires-de-comptage/information/?disjunctive.voie&disjunctive.sens&disjunctive.trajectoire)
**The number of sensors and their ability to distinguish the type of vehicles (e.g. scooters and bicycles) can change over time.**
**
**
**Accuracy on the content of the “Trajectory” field:**
It is a string designating the detection start area (input) and the detection output zone (Enter > Output)
**A trajectory is characterised by a direction of circulation and by the traffic line taken by the vehicle entering and exiting.**
Example: A bike that can in the detection area of the thermal camera enter the bike path and exit the general traffic lane.
**More details can be found in the package leaflet** of the dataset.