The SOGRID SmartGrid project, in which my research group at Ecole Polytechnique has been a partner, is – after three exciting years of hard, gratifying, and at times frustrating, work (as “working with the real world” sometimes can be) – coming to its end.
This has been an, in majority, industrial project – funded by ADEME, the French environmental end energy management agency.
The 27 M€ project launched in 2013, and with the ambitious goal to develop a comprehensive, embedded, unified, IPv6, PLC-based, end-to-end communication system for the SmartGrid – the evolution of the electricity distribution system which will, among other things, enable integration of renewable energy, decentralized production, electric vehicles, etc., as well as increase the reliability of the electricity supply.
Part of the goal obviously was to develop the underlying protocols and technologies. Another, more important perhaps, part was to bring it from “intellectual idea” to industrialized chip (SoC), prototype/demonstrator deployment on the actual power grid – and international standard.
In my research group at Ecole Polytechnique we developed the routing protocol used, and the associated technology – in close collaboration, of course, with the industrial partners and in particular on this topic with Cédric Lavenu from EDF R&D.
One of the interesting highlights of participating in the SOGRID project was, for me as an academic, to have an occasion to attend a very different kind of events, speaking to very different audiences from that which I meet in usual “academic conferences”. One example of this was attending European Utility Week in Vienna in November 2015 – essentially a “trade show” with some technical sessions, which was a fantastic opportunity to exchange with engineers “from the field”.
|Cedric and I gave what I think was a well received talk at European Utility Week 2015 on the whole SOGRID communications stack – from the physical channel to the topology management and networking layer.||
|At the same event, Damien Dufresne from ENEDIS and Olivier Genest from Trialog – both also participating in SOGRID – gave a very thought provoking talk about SmartGrid security.|
|To be a little more specific, the two major contributions of Ecole Polytechnique were:
LOADng has been ratified as the routing protocol for G3-PLC (G.9903) recommendation, thus checking the box “contribute to international standardization” as was part of the objectives of the project.
Building the actual demonstrator/prototype, justifying “in the field” that the SOGRID approach (IPv6-over-PLC) was viable, was really a core objective for the project. As an academic, it was an interesting occasion to get an insight into practical engineering of mission-critical (& potentially dangerous – we’re talking an energy distribution grid, so 20KV) systems, as practiced by the industrial partners in the project.
To this end, a System-On-a-Chip, affectionately called ST-COMET, was developed encompassing the complete stack, from signal processing and modem, up to and including routing and IPv6.
Equipment was built, tested, qualified, and installed on the grid. From SOGRID-fuled power meters, over data concentrators in transformer stations, and to various captors
So, in summary, how did we do?
- The SOGRID consortium contributed to the ITU G.9903 standard for smart grid communications. I and my team held the pen on writing the “routing specification”.
- My team and I also developed a prototype implementation of the routing protocol, which was available to our industrial partners as a reference implementation for interoperability testing, etc.
- A chip was built – that’s really cool, a complete communications protocol start in a low-power piece of silicon.
- Equipment was also built and installed on part of ENEDIS’ grid in Toulouse, France. This is also really cool, as we’re going to exploit that for a while for getting experimental performance data (notably: are we able to satisfy application requirements? I dare say “yes” but wait for the paper to come out…)
- The bet was: is it possible to build an end-to-end IPv6 based, multi-hop PLC SmartGrid network? Can we develop the technology, algorithms, etc., necessary? For my research group, of course, the key question was “will the routing protocol we have developed (LOADng) actually be able to maintain the network connected? The answer to all of that was a resounding YES.
SOGRID has concluded – at least officially. I, modestly, think that we did something really awesome. I know that we learned a lot – at the intersection between industrially minded academics, and industrials with a strong sense for the necessity of basic research…and, who are all profoundly curious as to what “venturing into the unknown” might bring.
It was hard work. It was frustrating, at times. We cried. We laughed. But we got great results, both academically and industrially.
Connections were made, friendships built, and foundations for future collaborations were laid.
Thank you, Xavier, Damien, Alain, Gilles, Thibault, Olivier, Samuel, Stefano, Cedric, Cedric, Sonia, Pierre-Louis, Matthieu, Amaury, Hassan, Marc, Alexandre, Omar, Romain, Jiazi, Benoit, Nicolas, Jerome, Vincent, …