A Data Challenge to rethink inter-company cooperation and encourage the emergence of innovative solutions


In November 2021, TotalEnergies and Air Liquide revealed the ranking of their co-organized data challenge. For six months, over 200 employees from both companies tackled complex issues related to wind power production predictions. The initiative was a resounding success and precedes the launch of a new data challenge organized by a greater number of French industrial players, and open to external participants in the first quarter of 2022.

The need to learn how to collaborate on data issues is part of the ambition described in the Manifesto on Artificial Intelligence for Industry, kicked off in 2019, and which now has around fifteen signatory members. “TotalEnergies and Air Liquide are two of the co-signatories of the AI Manifesto, which encourages us to pool our efforts when it comes to AI issues to provide answers to the need for supremacy in this sector,” highlights Sébastien Gourvenec, Program Manager of R&D for Machine Learning and Artificial Intelligence at TotalEnergies, and co-organizer of the data challenge, alongside Athanasios Kontopoulos, Manager of Computational & Data Science Scientific in the R&D department at Air Liquide, and International Fellow.

The aim of this challenge was to reveal innovative approaches to data processing to provide answers to the challenge of predicting the production of wind farms. A real case example pertaining to the energy transition and renewable energies — a sector to which both TotalEnergies and Air Liquide are strongly committed. “We wanted to remind people of the important role of data science in creating innovative solutions that solve the challenges of the energy transition,” Sébastien and Athanasios explained. “We have identified interesting methods and now it’s up to us to study them in greater depth to see how they can be applied to our real-case scenarios.”

The experience was also an opportunity to run a collaborative project based on data with another industrial player, to share best practices and to showcase and expand the Data Communities at TotalEnergies and Air Liquide. “This project confirms the importance of identifying and developing talents, especially in the data sciences field, to support our innovation strategies and highlight the opportunities that this type of open innovation can bring,” Athanasios points out.

The successful collaboration between two members of the AI Manifesto, major players in the Paris-Saclay ecosystem, has paved the way for a new data challenge project, involving some of the other French industrial players who also signed the manifesto. “A Data Challenge is designed to encourage collaboration, creative mindsets and to offer new Machine Learning/AI approaches to solve a given problem. It encourages emulations, stimulates talent and sometimes results in highly promising solutions. We’re ready to embark on the second data adventure!” Sébastien and Athanasios conclude.


Key figures

    Phase 1

    • Individual competition, open to all TotalEnergies/Air Liquide employees.
    • From May 10 to June 28
    • 210 participants
    • 2,710 contributions

      Phase 2

      • Competition in pairs from TotalEnergies-Air Liquide, exclusively for the best candidates selected after phase 1.
      • From July 5 to October 4
      • 376 contributions