Artificial intelligence for recognition of electrical equipment.


Arnaud Tasset

Arnaud Tasset ENEDIS

Séverine Mulatier

Séverine Mulatier ENEDIS

Olivier Girard

Olivier Girard ENEDIS

In a word

Dataposte is a mobile application, using artificial intelligence, which identifies electrical equipment based on photos, thereby revolutionising the collection of technical data.

The project

Only 15% of the characteristics of electrical substations are known, and existing data collection systems are unsatisfactory.

Dataposte is a genuine technological innovation providing semi-automated collection of data, for anyone with or without specialist knowledge of electrical equipment.

This mobile application was developed by Enedis in agile mode, in collaboration with regional offices, corporate specialists, the information systems division DSI, the Digital Directorate and external partners (the engineering and science institute Centrale SupElec, and collection service providers).

It offers numerous advantages: traceability of electrical equipment, increased reliability of assets-related data, customer and community satisfaction, focused investment, enhanced maintenance programme performance results.

The application uses an image recognition algorithm entirely developed in-house at Enedis. It is currently being used in Ile-de-France and will be rolled-out to the rest of Enedis in 2020. Its deployment is the subject of ongoing work by SEI (island energy systems).

The scope of the project is currently shifting towards more extensive recognition of electrical equipment.

Why is this a game changer ?

A cross-cutting and collaborative project
that delivered an application used by Enedis personnel and collection service providers in the field.
On a nationwide scale
with more than 5 million data points to be collected or verified across approximately 500,000 electrical substations.
The 1st image recognition algorithm
to be developed by Enedis, with an 85% recognition rate for electrical equipment, and a data collection time that is 10 times faster.
Enhanced reliability of downtimes
and greater asset protection.
A unique image library
containing all the photos of Company electrical equipment.
A technological breakthrough
with the implementation of a data collection solution that can be harnessed for all the Group’s technical specialisms (nuclear power, hydropower, etc.).

EDF Pulse Internal Awards Calendar


6 January 2020

Launch of the 7th edition of the EDF Pulse Internal Awards.

31 January 2020

Registration closes. 330 files received.

11 March – 13 March 2020

Selection of the 24 finalist projects by the selection committees.

23 March 2020

Announcement of the 24 finalist teams

2 November – 1st December 2020

Employees vote for their favorite project.

12 January - 13 January 2021

Finalists coaching.

13 January 2021

Auditions before the members of the Group's Executive Committee. Awards ceremony.

13 January 2021

Revelation of the 7 winners!