Researcher profile: Georges Hébrail

With a career experience alternating between positions of researcher and manager, Georges Hébrail has been fascinated with “data” for over 15 years. His credo: facilitating the ownership, mastery and processing of data to face the digital revolution and boost the performance of the company.

I began my career outside EDF's R&D in the unappreciated field of “management IT” where we were already processing a lot of data. This is when I really understood the growing importance of data analysis to improve the company’s efficiency. After my doctoral thesis on aggregates in databases – a precursor of decision-making informatics and business intelligence, I designed an environment for analysis and management of research projects for EDF's R&D; this proved to be the occasion of clearing the unexplored field for “text mining”1. I later perfected my expertise by conducting work on statistics and data analysis, leading me quite naturally to Big Data.

Without any hesitation I would say the “Courboscope” which we designed jointly with my colleague Corinne Muller. The software, intended to analyse and classify load curves (clustering), is still in use 15 years later, in particular at our Sales division!

Work on big data is booming and progressing very quickly. We therefore need to keep up-to-date on the latest technologies for data collection, storage and analysis on the one hand, and on the other hand build a big data vision for the R&D and the EDF Group. Accordingly, we produce usage case studies and proofs of concept focused on electricity generation (monitoring large-scale power plants via sensors), on distribution (processing of metering data, smart grids...), on supply (new offers, CRM...), and upstream/downstream optimisation. On a more scientific level, I am researching the analytical processing of personal data as related to privacy protection, and the time and spatial disaggregation of power consumption data.

In my field of expertise, collaboration with EDF’s operational units is essential, otherwise we have no access to the data and therefore no possible results on feasibility! Collaboration with academic labs in computer science and applied mathematics is also crucial in order to take advantage of the latest methodology breakthroughs on analytical data processing. We are evolving towards new, more open collaborative modes, working jointly with academic labs, other industrial groups (e.g. Orange, La Poste or Veolia) and with startups. I have experimented with several of these realms, firstly by holding a number of management positions for a few years, and later for 8 years as professor at Télécom ParisTech where I also founded and headed a joint lab on business intelligence with EDF's R&D, called the BILab. Today, I still keep strong ties with academic research as a doctoral thesis supervisor.