Intelligent street lighting - Datacity

Implemented

Intelligent street lighting

Home > Challenges > Intelligent street lighting

Backgrounder


How to provide intelligent lighting in the streets of Paris by analysing urban travel data ?

In Paris and other cities, street lights stay on at night even though they could be dimmed when there are fewer people about (to reduce their carbon footprints and light pollution) or brightened for specific events (e.g. around stadiums on match nights). This intelligent street lighting system in particular needs to cater to soft transport (walking, cycling, etc.).

 

One approach could involve fitting sensors and meters on street lights but trials so far have shown that these systems require heavy investment and furnish comparatively unreliable results. This challenge will consider other approaches, such as using geo-statistics from a mobile operator.

Corporates

Startups

  • The problem

  • The business opportunities for the stakeholders

  • Use cases and experimentation field

  • Results

How to provide intelligent lighting in the streets of Paris by analysing urban travel data ?

This challenge will interest any company that operates urban infrastructure – such as EDF (France’s legacy electricity utility) or Bouygues Energies & Services (which manages and maintains street lights in Paris via its subsidiary Evesa) – and wants to sharpen its competitive edge by providing a wider choice of urban services and connected objects including upgradable dynamic lighting.

 

Cities such as Paris are broaching these issues to reduce light pollution and power consumption. This challenge should provide the city with a new night-time map including soft transport.

 

SFR is keen on a solution to manage street furniture in general, and street lighting in particular, that taps into the geo-statistics expertise it has developed by using radio-mobile data.

The existing maps showing activity on the streets of Paris at night are only based on noise produced by motor vehicles, whereas light grading also needs to accommodate soft transport (walking and cycling, principally).

 

The experiment field covers a selection of Parisian neighbourhoods and other places such as the area skirting Jean Bouin stadium, the esplanade in front of Notre Dame cathedral, and Place de la Nation. In these neighbourhoods, the goal will be to plot out time-based lighting grading scenarios catering to different points in a same group of streets, using the analysis of urban travel data. This map’s scenarios will for example need to encompass variations on a weekly and seasonal basis, as well as one-off events (White Night festivals, etc.).

 

In order to pursue the experimentation, the selected startup have access to the following data sets:

 

– EVESA will supply georeferenced data pertaining to street lighting facilities in the experiment area. The selection will be representative of Parisian streets and sights

– SFR will provide travel-related data from mobile devices (technical data from GSM antennas) for a representative sample (approx. 30% of the population), covering uninterrupted periods (24/7) in France (locals and foreigners)

Solution:

 

Quantmetry teamed up with Dataiku to modelize urban travels at night, crossing anonymized and aggregated data from SFR mobile network, and urban streets and travel data coming from the open database of Paris city. They demonstrated the possibility to graduate street lighting and developed a web app, enabling the visualization of off-peak time where the lights can be dimmed, as well as cost and energy savings related to the dimming.

 

Impact:
–    Savings ranging from 3% to 10% on the city annual electricity bill for street lighting (representing on average 1 million of euros per year)
–    Scalable to the 199 000 street lights in Paris and to any street light furniture in the world, without installing any specific sensors or equipment and using mobile data