Using data intelligence to line up environmental challenges with urban traffic management



It is no news: mobility is in a large part responsible for carbon and other pollutants emissions affecting the air quality in cities and metropolises. Facing this critical assessment of the situation, it is becoming essential to shift our current mobility system to a more virtuous one. What tools can be used to reach this goal?


Urban mobility: a transforming complex system

The urban network was conceived for the use of individual cars and up to now was not encouraging the use of softer and environmental-friendly modes of transportation. Comfortable and safe, the individual car is still perceived as the most practical and fastest way to move around. This prevalence given to the individual car in cities is shown through pedestrian ways sometimes dangerous and bike routes still rare, frequently non-existent.

The main challenge for today’s urban mobility is to coordinate and integrate many uses on a single road infrastructure: individual vehicles, delivery vehicles, taxis, public transportation, and micro mobility.


To decarbonize the urban mobility and decrease its environmental impact on the air quality, the voluntarist policies are focusing their strategy on soft and active modes such as biking or walking. On the public transportation side new layouts of the road infrastructure or updates of the traffic signal control are deployed to help the buses traffic and decrease the traveled time: rsuch as bus corridors or prioritized traffic signals at intersections.


Regarding the individual vehicles, several measures are taken to build low emissions areas within cities, as well as measures to decrease average speeds such as enforcing 30 km/h in cities’ streets (effective in Lyon this year, for example). The parking offer has also to be redesigned to encourage the modal shift. Solo-driving is also an issue to tackle: to curb its dominant use for commuting to work and to support carpooling development, reserved lanes for carpooling are implemented for example in majorurban ringroads or highways.

What about traffic management?

Even if the priority is not given to the individual car anymore, it is more about controlling traffic to avoid traffic jam and long queues, which generate stop-and-go phenomena producing carbon and pollutant emissions. However, control strategies are rarely centralized, except for large metropolises, and they even more rarely adaptative. For example, many crossroads’ controllers work as a single unit, without coordination (except for the green waves) and with a fixed signal plan: the signal durations are set locally. For the more advanced controllers, they have access to preestablished libraries of signal plans which are, for the most part, activated in advance according to fixed hours : morning or evening peak times for example. Nonetheless, these strategies aren’t enough to avoid traffic jams.


On the other hand, some implementations work with adaptative strategies: the traffic light plan changes according to the real traffic flow coming upstream at a given crossroad. This kind of traffic management is possible thanks to sensors and traffic counting devices measuring the traffic volumes and density and sending this information to a smart traffic management center that optimizes traffic lights signals in real time by sending back instructions to the local controllers.


Another issue to tackle is also globalized strategies, on the scale of a neighborhood for example, which imply the real-time synchronization of crossroads controllers.


Transitioning traffic management to the era of data intelligence

The main challenge is to transform a traffic management center software solution that is simply remotely monitoring field equipment (traffic light controllers, counting devices, etc.) into a smart tool able to integrate and manage multiple data flows as incoming sources. It requires also to integrate powerful algorithms to compute this data in real-time before sending back the new orders to the crossroads’ controllers. All of this should, of course, be done in a very short time, which is now boosted thanks to the improvement of broadband or wireless communications bandwidth and technologies.


It is also critical to upgrade traffic counting devices to measure all possible types of flows: private vehicles, delivery vehicles, heavy trucks, bikes, etc. It is not only about detecting the vehicle presence anymore but also counting a traffic volume with the best accuracy possible. Adding more measuring devices within cities is also required. New artificial intelligence technologies applied to the processing of images and videos seem to be today a must have required by local authorities to complete traditional counting devices.


The integration of third-party data flows and in particular data coming from connected vehicles (also called FCD – Floating Car Data), plays an essential role as well in enhanced traffic management operations: this type of data is completing and enriching data sources coming from traditional counting devices. Their combined processing enables to inform the traffic management center in real-time, allowing the adjustment of field equipment almost instantly.


It is then about improving existing traffic management center solutions with new technological bricks of data intelligence (data science technologies / machine learning / artificial intelligence) to complete existing data collected with counting devices (loops, cameras, radars, etc.), enable short-term traffic conditions or travel times predictions by combining historical data and real-time data.


The challenge is really to optimize urban traffic management and control strategies to fit as close as possible with real-time traffic flows and act efficiently to accelerate the transition towards a smoother, safer, and low carbon urban mobility.


Wish to discover how Neovya is using data to optimize our roads?
Talk with an expert ➜ Book a meeting!