How can soft modes of transport be integrated into the city traffic so that citizen feel even more likely to use it?
Traffic lights ensure the flow of traffic and the security of road users. At the city scale, the duration of green and red lights are organised so that vehicles can move easily and roads are not congested. Currently, the regulation of traffic lights is mainly adapted to motorised traffic, that is, cars, trucks and buses. With more and more trips carried out by foot, bike, electric scooters or other sustainable modes of transportation, their presence should be integrated in the management of city traffic.
more trips by bike in Paris
between 2015 and 2020
In this context, Leonard (VINCI Group’s foresight and innovation platform), Evesa, LACROIX Group, the City of Paris and the startup CyClope.ai got together to adapt the traffic light system to these new uses. This project explores technical solutions to take into account all modes of transportation, both sustainable and motorised, in decision algorithms that control waiting times at traffic lights. Firstly, the experiment seeks to validate the technical feasibility of detecting pedestrians, cyclists and electric scooter users in real me through image analysis. The goal is to develop an effective algorithm, capable of recognising users of sustainable mobility in order to extract real- time information on their presence at pedestrian crossings. Secondly, the project aims to deploy the algorithm at a traffic intersection to adapt the lights in real time.
Cyclope.ai develops video analytics softwares, based on cutting-edge technology using artificial intelligence and Deep Learning for road infrastructure operators. It builds industrial solutions to help traffic actors to optimise flow management in real time.
Based on images collected by cameras at a traffic intersection, an algorithm using artificial intelligence detects pedestrians, cyclists, motorists and other forms of mobility. The traffic light regulation is then adapted according to this information in real time. The switch to a green light depends on the density of people waiting on the pavement, the number of cyclists and electric scooters on the roadway and stationary at the traffic light. As a result, the presence of sustainable mobility users waiting is taken into account in the management of the traffic lights of that specific intersection. Moreover, the data generated from the video analysis constitutes valuable information on traffic composition at this junction.