[Open Data] The National Building Reference System (RNB) – the creation of a national database

The National Building Reference System (RNB)

Walid: Hello everyone, welcome to this new episode of Projets Libres. Today, we’re going to continue the series on open data and repositories. I’m delighted because today, we’re going to talk about a subject that I discovered completely by chance, I think, on LinkedIn. It’s called the national building reference system, the RNB. I threw a bottle into the sea and said “hello, would you be interested in coming and talking about RNB on the podcast?”, and here we are. So I’m delighted, I have with me Elisabeth Ville, who is a co-entrepreneur of the RNB, and Félix Veith, who is also a co-entrepreneur of the RNB. With the two of us, we will talk about RNB: you will be able to explain to us what it is, in what context it intervenes, what were the discussions that led to RNB, etc. So there you have it. Felix, Elisabeth, welcome. Thank you so much for being here with us. Welcome to the Projets Libres podcast.

Elisabeth : Thank you, we’re delighted to be here.

Felix : Thank you very much.

Presentation of the guests

Walid: Well, to start with, before we talk about RNB, we’re going to start by talking about you. Please, and I’m going to ask you to introduce yourself, tell us who you are, what your background is and how you fell into the digital commons. Elisabeth, do you want to start?

Elisabeth : Yes, so we arrived at the same time with Félix a little over four years ago on this project through a program called Entrepreneurs of General Interest. So it’s a program where at the time, they were looking for lawyers. And I’m a lawyer by training, even trained as a lawyer initially. Absolutely nothing to do with free data. And over the years of my career, I first innovated for jurists and lawyers. And by doing innovation for jurists and lawyers, I came to innovation in the public service. And as a result, we were put on the RNB file at that time. So that’s how I got there.

Walid: And you, Felix?

Félix: So, I have a profile of a general engineer by training. Following a few experiences, I was a data scientist on energy saving topics. And when I saw the opportunity provided by the Interministerial Digital Directorate (DINUM) of the Entrepreneurs of General Interest programme, I thought it was an opportunity to make my skills available directly for the general interest. And that’s how we ended up on the RNB project, which at the time was called BATID. And at the time, I didn’t necessarily have a lot of knowledge about the digital commons. In the end, I had used them without knowing that it was the case. And it was as the project progressed that we realized the existence of this concept and all the potential behind it in terms of its ability to act and transform the digital universe.

What is RNB?

Walid: There are a lot of things that I discovered while preparing and that we will be able to address and it’s really exciting. Can you start by explaining to us what RNB is?

Félix : So the RNB is the national reference system for buildings.

It is therefore generally called RNB or R&B if you want to be a little humorous, and it is a digital public service that lists all the buildings in the territory and associates them with a unique and perennial identifier. And in a way, once it is assigned to the building, this pivotal identifier, which is called RNB ID, simplifies the cross-referencing and exchange of building data between the various public administrations, local authorities, private actors and citizens.

Felix Veith

The environment before RNB

Walid: Very good. You say that you arrived on the project 4 years ago. What made it decide to create this database? What was the environment like before RNB?

Elisabeth : So… The idea of the reference framework came to ADEME, the Agency for Ecological Transition, since it is the agency that is responsible for conducting studies on the environmental impact, in particular public policies. And so the building department, which is a department dedicated to impact specifically for the building sector, realized that without a reference system, they had a lot of difficulty, especially in steering public policies, in conducting sufficiently qualitative studies. They were really missing a stitch. And so that’s where there was this idea of creating the repository and this need. And what was quite interesting is that when we arrived in September 2021, we started the project with a lot of user research. We asked ourselves in fact, is a major project like this, the creation of a national repository that will have millions of lines, relevant? Does it meet a real need? And that’s when we realised that ADEME had initiated the need for a public policy for renovation through this prism, but that in reality it was a need that went far beyond that and that there were a lot of different sectors that were suffering precisely because they didn’t have an identifier. So typically, if we think about emergency services, we will also have a lot of things around building insurance, tax collection, census. Anyway, there are going to be many, many public policies that will be affected. And so, before the identifier arrives, in the end, what happens is that as nature abhors a vacuum, everyone will create their own databases, their own repositories of buildings in their field, but suddenly they don’t benefit at all from the information that the others have. So typically, there will be a base of buildings for schools, for sports facilities, for each community, etc. And so all this creates a lot of redundancy, a lot of silos, very little data sharing, which means that everyone has data in their corner and not very qualitative.

Walid: People don’t necessarily talk to each other, you know.

Felix : That’s right. In fact, they don’t talk to each other because, culturally, they are not used to sharing this type of data. However, they can see that in fact, they have difficulty maintaining, perhaps even sometimes, fairly large databases and information on them. And also, it’s not easy to talk to each other, because in the end, we talk about “yes, I have a database with millions of lines, and then, suddenly, there’s another database with buildings too, but they may not necessarily have the same definition. And then, to cross them, how do I do it? How do you talk about it?” And as a result, there is a technical barrier that means that it is not interoperable, and so they don’t necessarily do it. Well, they didn’t.

The birth of RNB

Walid: And now, the ADEME that is initiating this project, it is going to see the DINUM to talk to it about this project. How does it work? Because you’re coming through the Interministerial Digital Directorate?

Félix : At the time, the DINUM’s Entrepreneurs of General Interest program worked in such a way that there was a recruitment of digital profiles, so digital lawyers, developers, data scientists and designers. Overall, this is the type of profile. There was a recruitment campaign of “do you want to work for 10 to 12 months on projects of general interest within the State?” So we applied through that, without knowing the projects that were there, and on the other hand, they are doing another kind of campaign, this time, looking for projects, and so the administrations, for example ADEME, were encouraged to propose projects if they had problems that they wanted to address, with this contribution of resources from the DINUM. And that’s how the DINUM, it was Romuald Caumont, who was at the time deputy in the building department, who noticed the existence of reference systems sometimes for buildings or housing in other countries, he said to himself: “hey, in France, we realize, and as Elisabeth said earlier, that it is a problem for us not to have any, Isn’t this an opportunity to dig deeper into this question?” And so, we finally arrived at this title, by doing a bit of a post-baccalaureate admission, I don’t know, we made project wishes, we both found ourselves on it, and quite quickly, there was ADEME, but finally, there was also the CSTB , which is the Scientific and Technical Center for Building, who, because they also had these problems and they were working on a database, as a result, of buildings. There was an interest in immediately being these different actors around the table.

Walid: In a previous episode, we talked about a somewhat related subject called the national access point and Maxime Siret, who is the current manager, said that in fact, when you create a project like this and incubate it within the DINUM, it’s for a given period and there are metrics. We define metrics that will make it possible to know if the project is viable and if so, then to be able to transfer it to the ministry in charge. How did it go for you, precisely, when you were offered this project? I guess you’re being offered this project, how is it going?

Elisabeth: So for us, it was a little different, since the Entrepreneurs of General Interest program was originally for ten months only. That’s how it worked at that time. It doesn’t work like that anymore, there’s no longer that time deadline. In any case, it was a very short time. And the idea was to do a kind of proof of concept very quickly to be able to start a digital project.

Then we had the particularity of creating a state startup, and that’s the episode you’re talking about on board, which is a little different, since EIGs, general interest entrepreneurs, are individuals, so people like Félix and me, who arrive in the administration and try to dig into a problem, whereas the state startup is an entire team that develops to advance a product and which must then be reintegrated into the administration when it is sustainable. So these are two devices that are super complementary, and we were able to do both. And it’s true that it was a real chance for us to start with EIG, to be completely integrated into the ADEME and CSTB teams, so to go and see everyone, to be able to ask lots of questions. Whereas then, when we launched the state-owned startup, we were able to develop a product with a tight team, to have a lot of autonomy in the way we launch this digital product. And so, it was really two parts of the project that were very important and very complementary.

Elisabeth Ville

Félix: And maybe even to complete, finally, this transition, in fact, it was at the end of the EIG program. There was suddenly this need that was carried, that was significant and justified. And then it was: how do we manage to get it carried within the administration? And the Beta.gouv method, the approach you mentioned, with iterations on digital services, lent itself quite well to it. And in parallel to that, there was also an executive at the IGN who changed his posture to put more open source and created a Beta.gouv service incubator, which is called the Fabrique des géo-communs, while it was very recently renamed the Fabrique de la données territoriale, but in this case it finally made it possible to have a new player who was lending… in which Fabrique des jeux communs, it’s a little bit what we wanted to build at the time. And so we started this construction with the actors mentioned above and the IGN, which joined the process. So it was very timely, especially since the state startup also made it possible, as Elisabeth said, to create a team and no longer just be two people, but potentially more, to carry out this digital service.

The main stages in the creation of RNB

Walid: From the moment you start working on the subject, what I understand is that there is a first phase of understanding the subject, understanding the actors who are involved, understanding the scope of the subject, because I guess that when you start, it’s a big thing, in the end. What are the major steps that have occurred between the time you start working on this subject and, let’s say, the first public version of the RNB?

Elisabeth: The biggest step, in the end, was, and that, I think, we didn’t necessarily expect with Félix, because we were there in a digital program. I think we were thinking that very quickly, we were going to have a product, we wanted to launch something, a version 1, etc.

And in fact, as we said earlier, when we did our research, we realized that most of the players did not have the same definition of what a building is. And so that was a real key step where we said to ourselves: but in fact, we will never be able to build a digital product if we are not clear about the object that we have to determine in the product. So, at that time, we launched workshops with experts in the sector, a kind of big call within the framework of the CNIG, it’s the National Council for Geolocated Information. It is a public entity that serves as a forum for geolocated data. And so, as a result, we created a specific working group on buildings and we said: here we are, we are new. But in any case, we want to succeed in converging on this subject. So, all the experts, for once, we need you to discuss what a building is, how we are going to define it.

Elisabeth Ville

So there, we did a lot of workshops at that time. We brought in definitions that existed in different codes, definitions that could exist abroad on different similar standards. Then there were a lot of other things that we brought back to try to get to the end, to a definition of the built object that everyone agreed on. Knowing that by doing this, we realized that in fact, this reference project had been desired for 15 years, but that it didn’t come to fruition because in the end, it was really too difficult to agree. And I think there was also a real opportunity effect because we were dedicated to it. And in fact, it’s very difficult for people who have 50 things to do at the same time, to work on a project like that, partially with someone else who also does that, a little in another department, etc. There was a real need for real resources to be put on this.

And then, I think we do so, and I think it’s interesting in hindsight to see it like that, but as we were quite young and novices, there was really this freshness, I think, which also helped to smooth out the debate. Where before, there were still a lot of… chapel war, let’s say, between administrations and sectors, etc. And we arrived with a bit of this novelty and neutrality that meant that we were able to conduct discussions perhaps with fewer preconceptions. And we managed to come up with a definition in six months, I would say, more or less. And then, beyond the definition as such, it really allowed us to federate a group of people in many different sectors, administrations, communities, private sectors, who said to themselves “Ah, but this reference project, it’s interesting”. And so, when we released the product, we already had resources to rely on because they had participated in the creation of the definition and they felt hyper involved and actors in this thing. So, it also played a big role afterwards. But I don’t know if at that time, we realized it. We were really trying to make sure that we had a basis for discussion for everyone.

And the image that was often used at that time was a bit like the Tower of Babel. You see, all the administrations that try to work together, that don’t speak the same language, and therefore they can’t build something common. So our goal was really that, to first find semantic elements that actually allowed everyone to work together.

Foreign countries that have served as models

Walid: I was wondering, I don’t know this subject at all, what foreign countries have you looked at where it had already been put in place? What served as a model for you? I don’t even know if you have had contact with people in these other countries.

Félix: We had contacts, in particular with whom we exchanged a lot, they were always available to share information with us: it is the Swiss who have a national repository of housing, buildings and addresses . There is a lot of information that is shared openly, which finally makes it possible to realize that it exists, to see what definition they have adopted, what scope of action, how it is managed and by whom. So that was a very interesting and rich exchange.

Moreover, what was also interesting was to see that each time, in each country, the need to create the reference framework already appears at times that may be different, and especially for needs that are different. For example, in Switzerland, they have been setting up a reference system for more than 20 years. And at the time, it was for a population census story, to come and declare yourself and say “I’m in such and such a place and in such and such a building”. And so, as a result, it’s managed by the equivalent of INSEE in their country.

Felix Veith

Quite a few exchanges with the Swiss. It is also in Denmark that there is a concept of basic data with certain basic data that must be disseminated, including that of the building, with another entry prism which is, if I am not talking nonsense, the petitioner. As a result, each person, when they file their tax return or something like that, must bring information about the building. And then there are other countries. The one that served as a little bit of an example was in England, where it had just been studied and with an initial prism, the same, of energy renovation. So, that’s a little bit what also clicked at ADEME. And then, there are still other benchmarks. If I say the Makani Number, but each time, I am wrong about which country.

Elisabeth : It’s Dubai.

Félix: in Dubai. Same thing, but here, the problem is rather multilingual, buildings that are more recent, with different levels, etc. So not the same problem, but ultimately a need to share information on what we are talking about, what object, and therefore the need for an identifier on buildings.

Walid: ok, super interesting. I think I’ve already seen things about the Swiss in railway conferences [MATSim at SBB: Using and contributing to the open-source transport simulation for advanced passenger demand modeling] where they were talking about just that and the fact that they’ve been lucky enough to have repositories for a long time, which allows them to calculate the evolution of the population, movements, etc. I’ll link that in the transcript.

The definition of a building

Walid: I never thought it would be complicated to agree on the concept of building. By the way, I don’t know, do you have it, the notion of building? What is a building?

Elisabeth : So, we have our definition. You see, that’s not the definition when you’re going to sign your contract to sell your house. Well, you know what I mean. This is a definition that is specific to the reference system, but which has made it possible to divide up the park anyway.

So, in definition, in the sense of the reference framework, it’s a construction that is underground and/or above ground — in this case, we, for the moment, still mainly identify what is above ground — and which aims to be permanent. So this whole notion, there have been major debates on the building part, light or not. And then, the objective is for a building to be something that can house humans or human activities. So overall, this notion of serving an objective must really be for people or for activities. And then, we also say that a building necessarily has at least access from the outside, since we don’t want to make a repository without a link to the addresses.

Elisabeth Ville

And so that’s also a whole subject of how do we create a repository that dialogues with the other repositories that already exist. So really like Russian dolls. We have the plot, on this plot which is identified in the cadastre, we can access it by one or more addresses. And on this plot, we have one or more buildings which then contain one or more dwellings. That’s how we look at RNB, that’s why we put this aspect in the definition.

And the last part of the definition is that as far as possible, one building is distinct from another when it is impossible to move between them. And so that’s a little bit more delicate, because it involves a lot of knowledge of the field to know how to distinguish a large complex versus terraced houses, which in fact are each of the buildings. But if you take aerial shots, you have the impression that they are glued, and so you could say that it was only one or in large blocks of buildings or things like that. So that’s the part of the definition that is perhaps a little more difficult to implement. And there are always real questions. At the moment, we are asking a lot of questions about the definition, about large hospital complexes where you clearly have a lot of different buildings, which have different construction dates, different materials, etc. But in fact, everything communicates because there are tunnels or walkways between buildings. And so, do we consider one, several? So, there you go, obviously, these are still things that we continue to think about, but it still allowed us to initialize the first version in a fairly qualitative way. And overall, this definition is broad enough for everyone to find their way around and find their use case in it. And that was the goal.

Felix: That’s right, yes. And then this definition, it is completed by an appendix, finally, which is both… which is intended to be a little detailed while giving, as the definition gives, enough flexibility, but to try to orient and explain the different possible cases. We even made some illustrations on the website to differentiate it, is it a building or not a building, can it be one or several, why, and really tend towards something shared by all. Even though we know it, on some of them, we typically work on large housing estates, to see if there is a specificity around large housing estates and can we draw from them rules of the game that are a bit common between the hospital, a university or things like that that will make it possible to explain in this case and in such a way that we manage to have an identifier for each building object that ultimately corresponds to the different uses.

Walid: Just for my understanding, when you start the project and you see the number of actors. What is your feeling when you start to have the number of bases, the number of actors? Is it scary? Are you saying that potentially it won’t work? Or are you just super excited and you don’t ask yourself too many questions and you go for it?

Elisabeth: Actually, what I liked about this project and what I like in general is that most of the people around us told us “you’re never going to make it” because it’s too big, because there are too many people, there are too many actors and everything. I love it when it starts like that because you can only raise the bar. That is to say, at worst, if we don’t succeed, no one expected us to succeed anyway, so why not. But then, that was a bit the state of mind at the beginning. And so, I think that when we realized that we were already able to federate everyone around the definition and have a kind of group of experts who really wanted to go for it, that was still very encouraging.

Afterwards, of course, there is still a side, when you don’t have a product, and you imagine what the product will become and the flows, in particular, when you do all the conceptual part, in fact, you say to yourself, we’re going to have to take into account the building permits, we’re going to have to take into account the new addresses, There will have to be municipalities, private and public. Now, you’re like, “Be careful, how are we going to make this thing happen?” But overall personally, I don’t know about you Félix, but we were rather optimistic and we managed to be, even if there were inevitably more or less simple moments.

Felix: Yes, that’s right. And then, I think there was a bit of a combo of saying: yes, indeed, there is a plurality of actors that is enormous, and therefore, it can be a bit scary. We say to ourselves: “wow, how are we going to manage to get everyone to agree and everyone around the table?” But also, this plurality of actors is also a plurality of expertise and people who ultimately have things to bring to us, especially us who are not experts in this subject. And so, that was also reassuring because in fact the approach we had, I think, which was perhaps a little naïve, but the fact of arriving rather in a position in fact where we need these levels, help at different levels… We received help, different opinions, and that’s very interesting. So in fact, there are plenty of actors, but above all many motivated actors who shared this need and wanted to give it a door to contribute a little bit. And so, they did it. In the end, it was stimulating from the beginning to be well surrounded.

Walid: There is this whole first phase which is not technical, of discussion to reach an agreement. You start developing the product from the moment you agree on the definition. When do you start developing the product? At the beginning, there are only two of you. Besides, I don’t know how many of you are now. We’ll talk about it later. When do you make a team and start working on the product itself?

Félix: So, we make a team…

Walid: It makes you smile, I see (laughs).

Félix: Frankly, we’ve never had this phase of building a team, what happens afterwards. We didn’t know that this would be the case at the beginning. So, it was, as well, on a personal level, a very enriching experience. And we basically started to build a team after a year. Overall, from the moment that, in fact, we formalized the fact that we were going to work in state startup mode and therefore, suddenly, that there were these different administrations — suddenly, ADEME, CSTB, IGN and DINUM — that agreed and said, “Ok, we’re going to allocate a budget and suddenly, we’re going to be able to recruit people for a given time to work on this.”

There, a team was formed and, as a result, the first technical bricks began. Before, what we had done technically, in the end, was mainly to study the different existing bases, to see what the potential of each of them was, how it was going to be compatible or not. But the purely development part, we might have liked to start before but we realised that the two of us, with all the work that had to be done, hadn’t done it at all. I think it’s good to have pushed back a little, not to get into the product and technical aspect too early.

That’s when the team grew: we went from two to four. We started recruiting. Little by little, we have grown to eight people today. We were a little more at one point.

Elisabeth: What’s funny is that suddenly, as we didn’t have a product, we still did a lot of hyper visual stuff. Because in fact, people have a hard time imagining something purely conceptual. And so, as a result, I remember, we were making mock-ups of what it could look like. We made lots of schemes with flows, lots of actors. And as a result, when the product was released, it’s true that, especially for — I’m thinking a lot about deployment — it was very hard to deploy an idea, whereas from the moment the product was released, it was really much simpler and more concrete for all the players. So, for sure, there was a phase where we were a little scrappy about what we wanted to present and what it was going to look like.

Walid: All this first phase, you don’t communicate outside, outside the people concerned? You don’t communicate outside, to the public in general?

Elisabeth : yes, for once, we have quite quickly… In any case, we have always been hyper-transparent, because it is one of the principles of the DINUM. And so, overall, we’ve always done it quite quickly, we’ve done newsletters. Sometimes I think to myself, but there was nothing to put in it. But actually, yes, but compared to what we do today and everything we develop, we did a lot of newsletters, and as a result, we tried to communicate a lot as much as possible in order to create a kind of effect, an emulsion, you know what I mean? To succeed in making a sauce set. And it’s true that there’s often this side, “oh but what’s going on? What’s this thing? I can participate! I want to be part of it!” Finally, the fear of missing something. And at the same time, it’s true that in the public service — but I think it’s true in the private sector as well — there are always major innovation projects. And in fact, people see a lot of innovation projects that will be disruptive innovations, that will change their daily lives, etc. And which, in the end, do not necessarily succeed. So, there is also a little fatigue, I think, of being constantly involved in projects that start and ultimately don’t come to fruition, etc. So for us, it was important to communicate a lot, precisely, to show that we are there, it is moving forward. It’s concrete, things are happening. And so, quite quickly, we were rather very transparent. And then, the communication, it really took on more importance when we had people in the team who, in addition, got into these subjects specifically. And that’s one of the things that we put a lot of emphasis on.

Working with the public staff involved in the project

Walid: Still on these stages of creation, were there any constraints? Were there any legal constraints at the beginning? Were there any particular things about contracting with the ministries, with the people from the different ministries who worked or the different administrations who worked with you? I don’t realize it at all. Is it fair, well, let’s go, let’s go, then there are people who want to, they work with us? Or do you still have to contract things? It can’t be done just any old way.

Elisabeth: No, so we were very lucky.

We arrived in 2021, and that was after the Lemaire law of 2017, which provides for a legal framework for the circulation of data within the public sector, which is ultra-favourable. So before, the administrations had to enter into a lot of agreements, even among themselves, to exchange data, etc. And all this has been simplified a lot. So this does not mean that there is no agreement. However, sometimes people are reassured by the agreement, so they do it anyway. But overall, we were in a fairly favourable exchange framework. And what’s more, we still have the particularity of working on a subject that is not sensitive data because we link it to no information, neither personal nor fiscal.

Elisabeth Ville

You see, we really identify a physical object in space, and that’s it. It was also at this time that there was the paradigm shift at the IGN, where they stopped marketing their data to ensure that the data was available publicly, free of charge. We were a little bit after the rise of open data, which was very useful to us precisely because we did not have these constraints and limits on the way we work with other administrations and other actors. So, overall, it was still a fairly free framework, I think.

The indicators that are monitored to measure the success of the project

Walid: On the PAN, what he was saying — Maxime — was that: “For us, it was simple because in fact, it was a regulatory provision. So we couldn’t be told that the PAN no longer exists.” But the question I was asking myself about this was you, when you started, so it wasn’t a given that it was going to work. Once the state-owned startup was set up, what could make you say to yourself: “ok, it’s okay, in fact, we achieved the objectives”. Did you have any target levels? How could you measure the success of what you were doing?

Felix: Absolutely. We did not have, and I think we may have the opportunity to talk about it initially, a legal basis for the standard. So there was already a challenge to reassure and convince different actors to work with us, and then to work together on an object that is not yet official. So there was still a big exercise of conviction to be done on that side. But as you point out, what made it possible to convince us was to build indicators around our data, and in particular indicators around reuse: quite quickly being able to show that, in fact, the repository, the first indicators we had, was to say: a repository, the objective, is for it to be present in as many databases as possible. So we have to succeed in creating the snowball and pushing it, so make sure that there are different actors who adopt it, and then that they realize that it will be adopted by different actors. And so, there is always an entry cost for the first one, which is more substantial, but it is those who have the most needs who are willing to make the effort.

So there are really indicators around the number of players who join the database. And then there are aspects around demonstrating that it’s reused, so there’s a lot of downloading, using APIs, which make it possible to make information about buildings available in different formats. So that’s a lot of things that we continue to follow. And there, the current dynamic, it will also tend towards “Ok, very well, we have a stock of buildings, but we have to be able to capture the flow and also to correct.” Because we have built a database of buildings, we have tried to do the best of what existed from various existing databases, But there are bound to be initial mistakes: addresses that can be bad, a building that doesn’t have exactly the expected shape, or at least that needs to be shifted, things like that. This one has been destroyed, this one has just been built. And for all that, we try to follow the dynamics of contributors.

The contribution dynamic is something that we put in place a short time ago. We’ve been thinking about it for a very long time, it’s what we’ve been talking about with a lot of players since the beginning. Contributions started to arrive this year, especially during the summer, when we conducted an experiment by saying, for a given time, we will try to measure, to monitor what is happening. And during the summer, we have had more than 40,000 contributions since the launch. We have well over 100,000. I am not going to get ahead of myself. But on the other hand, it’s an indicator that is very important and that shows: are we capable of moving towards something? As a result, our objective is to better represent the reality on the ground as well as possible.

Felix Veith

Walid: ok. Because behind the indicators, there are financing issues too, I suppose. We have to show that what we do deserves to be funded.

Elisabeth: In addition, in the current context, financing is really the sinews of war. So what we are trying to see is what investment expenditure is avoided thanks to RNB? As initially, we could see that there were a lot of redundancies. Does RNB minimize these redundancies a lot and people start from something that exists? So, they’re not spending the money that they would have spent on creating their own built base, on launching tools, on things like that. So that, we are starting to be able to identify it. And then, we also have the whole game: does it even generate gains? And that’s more difficult because in fact, there are plenty of gains that are linked to RNB, but it’s never really direct. That is to say that it is not by having the RNB that I was able, I don’t know, to target a million euros of renovation much more precisely. It was the RNB that allowed me to cross-reference data from other databases, which in turn allowed me to create a management tool that made me win… Anyway, it’s always a lot more by rebound.

So it’s clear that the monetary value of RNB is a little harder to assess. And at the same time, I find that it is more and more audible for most administrations to realize that, in particular, if they do not do so, they are ultimately losing value in relation to all the information they could collect. And to understand the value of the information they hold. It’s just that it’s not always easy to calculate in a way… in euros, in time saved, especially since we are still very young. So, there are also many local authorities, on the other hand, who use us, who tell us: “we perceive that this will save us time and quality. But on the other hand, at six months of use, we can’t tell you exactly what it is yet.”

Walid: Are you still attached to the DINUM, for the moment?

Elisabeth  : Yes, we are still attached to the DINUM.

Walid: And tomorrow, if you leave the DINUM, you will be attached to one or more ministries?

Elisabeth: At the IGN, then.

Walid: To the IGN, directly.

Félix: In this case, the state startup format that we have, I don’t know if it’s particular, but in general, it’s a promising administration and supported by the DINUM that is always involved. And then, suddenly, there are four players today. So, what we were saying, the ADEME, the CSTB, the IGN and also the DAGLN, the General Directorate for Housing and Nature Development. And as a result, with these different players, what we have discussed a lot is where to sustain this service. And as a result, where it will be made permanent, it will be at the IGN, so it will be carried out within the IGN.

The level of progress of RNB

Walid: If now, precisely, we have started to talk about it, we are talking about the progress of the RNB, today, where are you in relation to the development plan? There, we talked about the fact that you had a V1 of a product, that you were able to open up to contribute, etc. Where are you in relation to what you have set?

Félix: Compared to what we have set, we are very happy to have the reference system that is distributed quite massively, and above all, there are uses. As a result, there are mapping tools, APIs, ways to download it.

And what we’re working on a lot right now is everything around the contribution dynamic and how we create trust around that. That is to say, today, there are many players who are inclined to test, but as a result, their questions regularly are to say if it is open by default, how am I guaranteed that there is a certain level of quality? Because the high historical approach is to say it’s closed, I do it myself. On the other hand, I know that I potentially do it well where I have the capacity to act, but on the other hand, I’m always going to be late, I’m not going to be able to benefit from what’s elsewhere, etc. So, there is a whole challenge around creating the beautiful operating methods that make it possible to guarantee…

Felix Veith

Félix : governance, you know.

Felix: Yes, that’s it. So, it’s all the governance and its technical implementation. So, that’s the big challenge we currently have and which is reflected in the technical developments of the project. Because because we set up a good monitoring system, backtracking, contribution modes that make it easier, whether it’s a result, there are very few errors, etc. So on the advancement, I would say that’s it, it’s mainly what we’ve done and the big issues ahead. And around that, there are still advances that validate our work and also allow us to ensure it for tomorrow.

And so, maybe Elisabeth will talk about it better than I do, but there’s the fact that, quite recently, we entered the law, through the bill on the simplification of urban planning and housing law. We were able to be integrated into it, which allows us to have a legal basis for the project. That’s very rich, and soon, it will be published in the Official Journal, which reassures the players about the existence of the RNB. So that’s still very strong. Even so, we will specify by decree elements of governance and implementation.

There’s also the fact, because there were a lot of initial concerns that were very good, you’re launching something, but how are we guaranteed that it will last over time. And so the fact that the project is sustainable also gives us some perspective, even though we must of course succeed in obtaining the associated funding. So yes, I think about the levels of advancement, I think that’s it. After that, there may be other things that I don’t know.

The stages of RNB sustainability

Walid: So, Elisabeth, can you tell us about the steps for the sustainability of the RNB? That is to say, when is there this need for sustainability and what are the stages of the process? Where are you in the process of making RNB sustainable? Can you enlighten us a little, please?

Elisabeth: So, state-owned startups are always a way to create new projects. And so, indeed, quite quickly, from the moment we had more and more users, we realized that we needed to be much more solid than a state startup beta version. So, for that, there were several steps.

The first step was to appoint an administration that would take over, since we have several sponsors. So, that was done this year, in 2025. it is the IGN that will take over the project. This makes it possible to designate the producer of the data in a certain way in a reliable way and it is a big step to know who is carrying.

Then we have a whole component of formalization through legislative vehicles. This year and last year, we have carried out a lot of work to try to find a law in which it is relevant to talk about RNB. We tried with the project to simplify economic life but we were told that in fact it was not in line enough with the text and it is true that it is a very broad text so we can try it but it did not work. Then we had a more recent text, which was the bill on the simplification of urban planning law, which was carried by a deputy. In particular, a range of tools was planned to help local authorities with their urban planning procedures. And since there were tools that were listed, we thought we could add the RNB to this list of tools since that’s what we’re doing.

And as a result, we went to ask senators, with the help of our sponsors, so the administrations. We went to see senators who were going to vote on the law and we have several senators who said to themselves, “Ok, that’s relevant, we’re going to make an amendment to this law to add RNB to the list of tools provided for in the law.” So that was supported by senators. Then, the government, and therefore the Minister of Housing, also said: “we agree, it’s relevant, we’re adding it to the text”. So this amendment was voted on by the Senate first. It was supported by the Senate, but also by the government. And then, it was voted on by the joint committee of the National Assembly. So the law was passed and we were in it. So we were very happy.

But suddenly, as you understand from the way I’m speaking, there is a but: on other articles of this law, the Constitutional Council has been seized. The Constitutional Council in France can be referred to it if there are 60 deputies or 60 senators who say “be careful, this law has been voted but we think it is not in accordance with the Constitution”. So that’s where, sorry, I’m going to be a lawyer, but suddenly it’s called the hierarchy of norms where the Constitution is the highest standard and therefore it’s the one that must be respected and all the laws that come after must always be in accordance with the Constitution. We start from the principle, in a way, that it is always compliant, because otherwise, it would be hellish to check all the laws that pass.

But suddenly, there is a criterion which is that if 60 deputies or senators think that it is not compliant, they can refer the matter to the Constitutional Council. So there, there was a referral on other articles, not on the RNB article, different articles on a different subject of the law. But once the Constitutional Council is seized, it has the right to look at the entire law. not only the articles on which it was seized. So, it can’t self-grasp, but once it’s grasped, it can look at everything. And in this specific case, he looked at the whole law and the decision of the Constitutional Council came out a few weeks ago: they decided that in the end, the law was going to be reduced in its scope. So, they excluded 13 articles from this law that had initially been voted on and in the 13 articles, there was the RNB article. They say that in the end, it extended the subject of the law too much and that the law, at the base, had a more restricted vision. And so that’s an article of the Constitution, Article 45, which says that we can’t finally put in a law something that is too far from the initial project.

Elisabeth Ville

It is supposed to be, to prevent an article on social security from being put in a law that talks about housing, for example. As a result, the link was perhaps less extrapolated but they still considered that it was too far from the initial project and for this reason they excluded a number of articles including ours. Obviously it was a big disappointment because we had done all this work to convince people of the interest of the RNB.

After that it always sounds a bit impressive but in the end it’s not at all a decision on the merits, they don’t come to say that the RNB is not in accordance with the constitution, they just say: “find another vehicle because this one was not the right one”. So at the same time, we have to find other texts so we’re starting from scratch and at the same time we’re already very happy to have been able to raise awareness among senators, to have obtained a consensus between the legislator and the government at a time when if you listen to this podcast in 2027, it may be different, But in 2025, currently, it’s difficult to have this kind of consensus. So that’s already a real victory.

And it shows us that RNB is still important. Now, in addition, with the calendar and the presidential elections that are going to happen, etc., there will not be many laws where we will be able to insert. So for the moment, it’s a bit on pause from that point of view.

But on the other hand, in 2026, we will probably be presented as an addition to the reference data of the public data service, so that’s another completely another vehicle, where basically it’s a list of data that is said to be reference, so it’s the SIRET number of companies, the address… Anyway, these are data that everyone uses and therefore RNB should be included in this list of reference data in 2026, which will still give it an important official strength. But on the other hand, it’s not going to be enough to make it mandatory, for example, in all procedures, for that we’ll need the law we tried to have this year.

We will try again, but as we will have the fact of being a reference data, it will already be a more important weight. So there you have it, it was a step that we would have liked to take in 2025 but it will be for later and in the meantime I admit that the most important thing for us has always been to be good enough for the product as such to be considered a reference. And that, in real life, this year we really managed to transform this trial in the sense that we went from 15 bases that integrate us to more than 60 and we are really known and used and it helps people in their job on a daily basis and in the end that was the most important thing so for that we are still happy with the year that is ending but we can’t wait to To see.

Walid: You are collateral damage…

Elisabeth: Exactly. But as a result, this is also the limit of the state start-up. That is to say, we are at the service of the state, like all civil servants and people who work in the public sector. And that means being dependent on the vagaries of public life. And you have to put your ego aside in those moments to tell yourself that it’s not personal, obviously, that the work you’re doing is still relevant, but just that you’ll have to find a better vehicle, a better timing at another time.

The first wave of contributions and governance issues

Walid: Super interesting. I would like to come back to the governance part, because in a common, governance is very important. So what I understood is that governance is something that you are working on. And so, I guess these first contributions, they allow you to understand how the tool is going to be used, by whom. What do you know, by the way, about these first 40,000 contributions? Who made them? Over how long? How did it go?

Félix: First of all, on this subject of governance, it’s true that there is something, I think, that is very strong, that has a lot of value, and that, suddenly, we notice, but throughout the project, and even, finally, in the construction of the law, I think it works, is to say: “we have to leave room for experimentation, To carry out tests to understand, in the end, where the need really is, what the different actors can do”.

So, that’s what we’ve been doing since the beginning on a lot of aspects. We weren’t going to change something that works pretty well. So, that’s what we did on the contribution to the repository part. Contribution, that is: who will be the actors who will be able to make changes. And as a result, in this sense, we hope to consolidate the repository and make it richer and more useful to other players.

For that, there are different contributions, there were many more than we could have expected in a test phase. That was super interesting. In the end, they were very varied actors. So there were actors from the administration, entities, it could be agencies of the administration, it could be the DGFIP, the IGN, a whole bunch of actors. There are a lot of local players, so municipalities, metropolises, social landlords, in short, who have an interest in their stock, their building perimeter, real estate companies, and also a lot of citizens. So, generally, more profiles, I would say, maybe OSM contributors (OpenStreetMap) [see the episode on OSM], or in any case, who generally, maybe have this culture of commonality. And who, as a result, understood the interest, and therefore got caught up in the game, and brought a lot of information.

So that was for the actor’s profile, so suddenly, very varied. Then, what we asked ourselves as a question was: were these contributions qualitative? Because it was still a big concern on the part of the different actors, and including us, it’s to say to ourselves, very well, we seem to be kissing bears, we hope that it will go well. And for the moment, the tests have shown that there are extremely few errors in the contributions.

And if there was an error, it was mainly due to a misunderstanding, either of the definition, or of the actions that are induced, that we can do on our interface. There were about 4% of errors in the contributions. However, we realized that these were errors that were due to: “oh yes, in fact, “if an object, by mistake, was not a building, it should not be changed to “demolished”, but rather it should be deactivated”. And so typically, these are notions that may not be obvious at first, and that have led me to mistakes. I cut it badly, because I didn’t really understand how I split a building in two, even though you can walk around in it. Sometimes, it can be a misunderstanding of the definition, but it was still very rich.

Of course, they were first contributors, they are also the ones who have the most interest and who followed the project, perhaps, although there were more than a hundred contributors, so there were people we didn’t know. So, very interesting to think that it was quite qualitative from the beginning. And yet, this is not enough to determine tomorrow: is it completely open? Who is allowed to contribute? Are there actors who, when they contribute, are more valuable than others? Does it somehow certify the data, maybe even block it for a given period of time? Because suddenly, a building doesn’t change every four mornings either. So these are questions that are still open, because suddenly, we absolutely have to authenticate ourselves with something like FranceConnect or ProConnect. So these are all questions that we are building. And as a result, we will build with the actors. So to have, at the very least, an open contribution with a minimum guarantee of trust and quality behind it.

The RNB Data License

Walid: Before I forget, because I don’t think I wrote it down in my frame, under what license are the data in the database?

Elisabeth: So, we use the Etalab open license, which is in a way a translation of the law that says how administrative documents can and should be shared in an open way. So in the end, it’s a license that just helps you understand how it’s shared and how it can be reused.

Walid: And by the way, when this governance, when you know more about it and you have managed to set rules, they will be available on the RNB website?

Félix: Completely, on the RNB website, most probably also on the project’s GitHub , which is open, because all the code in the approach, especially in the BetaGouv approach, everything is done, all the products are built in an open way.

And finally, and perhaps above all, we spoke earlier about the CNIGE, so the National Council for Geolocated Information: we have come up with a definition of the building, but the objective is to arrive at a building standard that will reference in the long term what the data will be, etc. And there will potentially be governance elements in there as well. The goal is also that once we have a built standard, therefore built with the different actors, it is stamped by the CNIGE and it has the possibility of being cited in articles of law, by decrees, which would make it possible to link this governance work to data, to a legal basis.

Felix Veith

Walid: And here, we’re talking about a building identifier, but is the database intended to store tomorrow, I don’t know, other data that is a little standard associated with buildings or not? Or is this database just used to make an identifier?

Félix: It really helps, yes. A priori and above all, to help identify which building we are talking about. So, in the end, if we attach addresses to it, it’s to understand which addresses are attached to this building. If there is a form, a geometry attached to it, it is to understand which building we are talking about in an overall grouping that can be complex. And so, finally, what is the status of the building? That’s information that is attached.

For the rest, it’s not in our mission, because there’s already a lot to do. And above all, there are already a lot of players who have, let’s say, today, in the absence of a reference system, tried to cross-reference building data. This gives a statistical truth, although in some cases, it cannot always be true. They do the exercise behind the: the repository is in a lot of databases, I aggregate this building data, I build services around it, web services, services or others. There are other actors who do it. This is not our mission. I think it’s rather, we’re going to make it easier to cross-reference this information and give a much easier basis to access the pooling of this information, but it’s not in our mission, in any case, to be a little bit the database that assigns the building.

The major players who will use RNB data

Walid: Precisely, here, we are talking about different actors. Can you briefly mention the major databases or the major players who will use RNB data?

Elisabeth: I think we’ll put in the notes the link to precisely, we have a dedicated page on the site that lists them, but overall we already have all the bases that are linked to our sponsors. And so, these are the ones that had the most obvious interest. Typically, within ADEME, we will have the DPE database, the OPRAT database. There is a new IPPER tool that is being built, which is supported by ADEME.

Walid: What is OPERAT?

Felix: OPER is the Observatory of Energy Performance, Renovation and Actions in the Tertiary Sector, which aims to ensure that tertiary sector players, some of whom are obliged to share their energy consumption and especially their reduction target over the next few years, with ADEME. And so for that, we need to know what perimeter is built for each of the actors, what we are talking about. And initially, this database, there was not yet the national reference system for buildings, so they had created another ID that was dedicated to that, but which did not allow the information to be cross-checked. And so, since the time we’ve been working with ADEME on this, there has been an understanding of the interest of putting the RNB ID in it. To cross-check. In fact, it’s being done.

Elisabeth: So, on the ADEME side, that’s right. Then, on the CSTB side, there is the CSTB’s BDNB , which is a database that is really very useful because it allows us to retrieve all this aggregated data, which we don’t do. So, it’s very complementary. On the IGN side, there is the TOPO database which also rebroadcasts the identifiers of the RNB. Then, we have a lot of different cases that fall more into the housing policy part of the DGALN, which is our other sponsor. Here, for example, we will have things on the vacancy of housing with Zero Vacant Housing, the LOVAC of CEREMA. We’re also going to have the basics…

Walid: sorry, what is LOVAC?

Elisabeth : LOVAC is the basis of Cerema which is created as part of national plans against vacant housing. This is to help local authorities identify vacant housing.

Félix: Overall, we are becoming a bit like the kings of acronyms.

Walid: Yes, I’m not very familiar with all the acronyms of the administration (laughs).

Felix : Don’t hesitate to ask questions.

Elisabeth : After that, you have the link with the land files, so all the housing that is also done through CEREMA. And then, we’re going to have a lot of themes. For example, social housing: there, we have the Banque des Territoires , which has made an application a priority for social housing, but also, quite simply, in the SDES database of the statistics services. It is the directory of social landlords, therefore the RPLS. We are in the directory of co-ownerships, of the ANAH. And then, we’re going to have the basics maybe a little more by theme.

For example, Accès libre is another state-owned start-up that provides information on the accessibility of housing for people with disabilities. We’re going to have themes, for example, national education, which are built, which manage — so we’re going to be in that base. The same goes for sports equipment. Data Tourisme is the specific buildings in the tourism sector. So you see, sector by sector, we are in certain bases.

And then, the other big panel we’re in is the bases of the communities. So there, it’s going to be both big cities like Lyon, Marseille, Nantes… We are also in Toulouse. Anyway, we’re in a lot of very big bases. Here, it’s more people who have GIS, resources to process this data. And then, we will also be in smaller municipalities that often create their own built base, and which start from the RNB to initialize them.

So here, we have Lannion-Trégor in the Côtes-d’Armor, we will have others. And it’s true that this is an aspect of the project that is really close to my heart, especially on the local part, which is to say that we are giving a level of data to all the communities, which has a kind of qualitative basis. And that, in the end, it’s not the big communities with big means that can have data on their buildings, it’s also small communities. And finally, that’s also what public service and open data is all about: it’s about succeeding in flattening the minimum level a little, and not having so much heterogeneity across the territory. And that’s very important for us too.

Elisabeth Ville

So there you have it, that’s the kind of reusers we have today. As a result, we know that we are reused by about 40 datasets. But we have more… I don’t know how many millions of API calls and tens of thousands of downloads. I think we’re at 14,000 downloads and maybe 14 million API calls, something like that. And so, reasonably, we can estimate that we are probably also integrated into many other datasets. It’s just that, as we’re in open data, we can’t trace it, and as a result, we can’t know exactly where we are.

So that’s a bit of the frustration of open data, because we’d like to know everywhere we’re being used, what use case, what it’s for. But for example, it often happens that… Typically, last time, there was Action Logement who told us “Oh yes, I’ve integrated you!”, or someone from the MAIF who came in and said “Yes, we use it!”. And so, you’re like, “Really? But it’s great!”, and we’re delighted. But it’s true that there’s the small frustrating part, especially on the impact monitoring, which is not easy, because the metrics we have, in the end, are usage metrics, a lot, but not use case metrics. In fact, we don’t know what people do with it. And so that can be a little bit more difficult. But we try to do our best to follow via webinars or typically what you’ve seen on LinkedIn, you see. When you find us, there are plenty of people who find us like that and contact us.

Félix: But I think it’s extremely positive for us, because it testifies a little bit to the change of scale where, historically, in fact, we’re going to push, look for users, their needs, approach them, try to take the step so that they understand the interest of the project and adopt the framework. So now, if finally today, there are some who — as a result, we learn elsewhere, and fortuitous — that it’s starting to be used, it’s a very good signal.

Walid: But that’s also the benefit of open data.

Felix : Exactly.

Elisabeth: Exactly.

Félix : It’s to promote reuse without having to worry about having to push it to everyone.

Walid: So, I would like to add a question. How do people react, for example the OpenStreetMap contributor who finds RNB interesting, who participated in this experimentation phase? Was it already clear that it was an experimental phase for some time? And how can he project himself in relation to the fact that there are people, producers who will potentially have more weight or have exclusivity over the data? Have you had any feedback from people who have reacted in one way or another to this?

Elisabeth: Absolutely. So, we launched an experiment over the summer because we had to frame the project in a time frame to make sure that we are at a kind of point where we say to ourselves “OK, in two months already, what do we have, what?”. So, there were the two months of this summer. In the end, we left the tools open, it’s still running.

But suddenly, the idea has always been to test, to see the results. We saw that there weren’t many mistakes in the way people contributed. And then, to transfer these results to the CNIG that we have already mentioned. So during the CNIG there are people from the official administration, local authorities, there are also people who are simply interested.

And that’s where we see, I find that this is what is fascinating in this project, all the tension that there is between the reference on the one hand, so it means a form of truth, and that the truth what it implies for some is not necessarily what it implies for others. In particular, for example, everything that is going to be the official component, so the local authorities, the central administrations, etc., etc., perceive the truth by the fact that the data has been stamped by an official actor. And so for them, this is ultimately how the truth emerges. Where we, in fact, realize that the truth is held much more widely and that it can also be in the hands of the average citizen or the OSM contributor who knows the field extremely well and who is very interested in it and who has information that does not go through official channels, In fact. And so, suddenly, this is where there is a challenge to succeed in ensuring that, on the one hand, officials feel reassured about this data, which is a reference, but on the other hand, to preserve the value of the common as such through much broader contributions.

Elisabeth Ville

And so, at the moment, it’s all the debates we’re having on data governance and, in fact, there are plenty of opinions, not everyone agrees and I can’t tell you again today exactly where it’s going to fall, but we feel that there are really these two things that are in tension and we don’t have any examples in terms of data that are reference, which are also commons. For example, the BAN [national address database], it is the municipalities that are designated as the official source of truth. Now we’re trying to create something new, are we going to succeed? It’s a real question, or will the administration with its conception of truth have to take over because otherwise, it won’t be in the official processes. These are questions that are really ongoing. And if listeners are interested, they are obviously welcome to give their opinion. We have a channel on the geo-commons forum which is dedicated to RNB governance issues where the different actors exchange and everyone is welcome to give their opinion. So, I really invite you to go and look at this if you’re interested. But that’s really the challenge of 2026. It’s about finding your rules and managing to resolve this tension and see if you can do it.

The initial construction of the repository

Walid: One of the other points I’d like to address — because, to do a bit of computer science, I think it’s a point on which you must have racked your brains — is the initial construction of the repository. I would like to know how you managed to build this repository and especially to cross-reference the data. Was it complicated? Did it take you any time? How long did it take you? And how did you proceed? Because having done this kind of thing already, it’s still quite a headache to do, I have the impression.

Félix: It’s a big job and we were lucky, for that, to have actors who had already done a big part of the work. It took us a while to talk about it, but we were lucky that the CSTB, the Centre scientifique du bâtiment, was working on a national database at the building level and was doing a huge amount of work on the division of the building, with a definition that was quite close to what we came up with. We were able to base ourselves on that. It was very rich, especially since it was the only player that documented its cutting process in open source.

Because, in fact, if we pull a division out of our hat, it’s much more difficult to get it accepted, rather than when we’re able to explain why there were these different choices. So that was very rich for us to start with.

And so, for a quick entry into how they did it, they are mainly based on two bases — maybe three — they based themselves on the IGN’s TOPO database, which lists the buildings, and documents that are the land files, which are therefore data disseminated by CEREMA, and which are initially data from the DGFIP, So which are accessible to certain rights holders, but which have made it possible, as a result, with different meshes of aggregation of plots and to know who they belong to, etc., to divide up the park. And in this sense, to divide up the built stock, and to get closer to something that is the definition we wanted.

So we were able to base ourselves mainly on this work to initialize, even though we had to make some adjustments, because the perimeter and the vision of what a building was or not, it was not exactly the same. So there was that, especially since there is also the great importance of saying that we produce data that is sustainable, that is traceable. And so, if we start and then tomorrow we say to ourselves that we are wrong, we have to put things back, we have to change everything, it’s complicated. So we took the time to initialize from that.

And once we had the technical infrastructure that allowed it, that we were quite comfortable with a first level of data model on that, there was also, perhaps already ready at the time we wanted to cut it down, there was the definition of the building, but also how we referenced them, what type of ID we used.

And so we ended up with an ID which is a nano-ID, so a string of 12 characters with an alphabet that is meant to be easily readable by humans, but that we can still exchange. It must not be too long, but there must be no risk of collision when they are generated. But for all that, they must be easily generable to push it everywhere.

So yes, it was a big project, but honestly, once again, we were lucky to have actors who had pre-chewed the work and who were also at our disposal to help us on these subjects.

Elisabeth: And also, perhaps, a mention for the RNB team which is still… Finally, we’re lucky to be with, you know, within the team, people who are super passionate, super invested.

The RNB team

Walid: What is his profile? What profile do you have in the team?

Félix : So, there’s Paul [Etienney] and Francis [Chabouis] who are devs. And then there are Félix and I with a bit of our specialties, and we still have this cross-functional aspect. Then there’s Léo [Paul-Goffic], who is a general interest entrepreneur in another class, who joined us this year, who helped us a lot this year also for the contribution part.

And then, on the deployment side, we have Inès [Chaïbi], Émilien [Guillemot] and Audrey who each do a deployment branch. So, there is a central administration branch, a community branch and a software publisher branch. And there was Lucia [Tahara], who is no longer here at the moment, but who was there for a long time, who was our coach.

And so, put together, all these people, it still makes a great team with lots of very complementary skills. And it’s true that that’s also one of the real elements, in my opinion, of the success of the RNB and the speed at which we’ve been able to do this: it’s to have people who have completely different jobs and who are very good at their own field, and who manage to work with people who don’t do the same thing as them at all. And that, I admit, I find that it’s really a very enriching source in a project. And it’s something that is not necessarily easy to set up within the administration, because often, there is the deployment team and the tech team, and they are on 15 products at the same time, each, and as a result, they are not in very frequent exchange. And we, on the other hand, have teams with everyone who talks and who has their own skills, and who manages to discuss this all the time. And so that helped a lot to perpetually get user feedback, to code the new technical feature, the test, well, you know, and to have this iterative thing. Without this whole team, I think we couldn’t get there so quickly.

Walid: Having watched, there’s something I find… There, I said to myself, for the search, it’s going to be great, it’s the history of identifiers with merges, with mergers and everything. That’s when I said to myself: if you want to retrace what happened in a given area over years, this is the ideal tool.

Felix: Exactly. And that, we’re happy with that, but it forces us to do it and to tell ourselves that we promise that it’s sustainable. And so now we have to do it. And to tell himself that it has to hold. So, we’d be happy if in 50 years, it could still build on that.

The future and challenges of RNB

Walid: I would like us to talk quickly about the future of RNB. We’ve already talked about it a little — well, even quite a bit — but if there are additional things you want to say about the future of RNB, how do you see the future? What are the challenges ahead of you? What would you like to bring to the attention of the listeners of Projets Libres on this subject?

Elisabeth: I think that one of the big issues is, as we said, on governance, that we manage to land on clear rules and that it is very readable for everyone. That way, we’ll get out of this stage today where we have very motivated and very expert people who use it. But we would like it to be much more democratized and better with rules.

We would also like the entire transition phase within the IGN to go as smoothly as possible. It’s really an exemplary administration in terms of geographical data, with a great network, passionate, hyper-talented people. So, we want this change, to become one of the IGN’s flagship services, to work well.

And that’s always a challenge, because suddenly, the state startup mode that is retransferred to the administration, it’s not necessarily easy, since we’re going to pass it on to people who haven’t worked on the product since the beginning, like us, etc. But that would be a real victory.

And then, being a reference data, I think that would really be the culmination of this project. And then, in the long run, I think we’ve done something called a “futuro-spective” for the RNB in 30 years. And we said to ourselves: “if we are used in all the databases, and people can cross-reference them and say, in such and such a place, we must renovate as a priority because in fact, this is where there are thermal sieves, because we have seen in the DPEs that it does not work”. These are really the use cases that we are starting to have. But in fact, these are things that will take years before we can see an impact all the time.

And so, I say to myself, but the day when the RNB, you see, it’s not really a question anymore. It’s embedded everywhere and used just like the address, in fact. That would be fantastic.

So then, there are other things that could be even more impressive, which would be to succeed in making a kind of total fusion between address, cadastre, local, built data. To make sure that everything is hyper up-to-date, hyper linked, permanently, etc. That’s a bit of a dream, because inevitably, it’s not easy. But I think for reusers, it would be really great. It would make their lives much easier.

Elisabeth Ville

So we still keep this objective in mind. And overall, we’d be really happy about that. I don’t know if Felix, you, did you have other dreams?

Félix: Other dreams, but already these, they’re cool. But yes, that it’s so widespread that we say to ourselves: a building number is like a social security card number. To say to oneself “I have a social security number and I have a building number”, that it is also widespread, I would say, in usage. That would be, I think, a form of implicit success.

Elisabeth : After that, it will never be a mainstream thing, I think, after all.

Felix: Yes, of course. And on the other hand, I come back, I think that the network between the different basic databases, which are the land registry, the address, the premises: if we want to build a geographical data base which, in France, can make it possible to do a lot of things with a good level of efficiency and a maintenance mechanism that is collaborative and qualitative, We will still have succeeded in something that is quite strong.

The European level

Walid: Before we leave, I’ll come back for a second to the European level, is there a desire at the European level to have an equivalent of RNB everywhere?

Elisabeth: Not explicitly. There is no specifically targeted project like that. But on the other hand, it is true that, if only because of what we explained earlier about high-value data, there is still a level of harmonization, or at least the expected quality of data by theme, which is quite clear.

Afterwards, it’s sure that when you see how difficult it is to set up at the level of a single country, to want the same thing everywhere, in the same form… I don’t know to what extent this is something that is being carried out at the European level. In any case, what’s funny is that we questioned our neighbors a lot at the beginning. And now, we have people who come to question us about how we did it.

We had people in Germany and Belgium who came to see us and said, “How do you do it?”, etc. So that’s always quite nice to think that we can inspire other projects elsewhere on how to do it. In any case, we’re happy if it can create a little innovation elsewhere too.

Walid: And all the actors who have spoken to you or who have come to talk to you, are they public actors? Because what we were discussing in the national access point is that depending on the country, this thing was implemented sometimes, it was private, in fact.

Elisabeth: You’re absolutely right, indeed. It’s not always the same entry point. What is certain is that at the European level, in fact, the definition is compatible with the Inspire directive. So, I think that in fact, all those who comply with the Inspire directive at the European level, they may have a definition that is not so far away. And as a result, perhaps a way of managing a repository that could be similar.

The actors who came to see us… in fact, for example, it was in Belgium, in the end, it was a consulting firm that was commissioned for the Brussels-Capital Region.

Walid: Brussels-Capital Region, certainly.

Elisabeth: Yes, exactly. So there it was a public sector. And I think that in Germany, this was also the case. However, the final portage, even if it emanates from a need for a public actor, it can be carried out by a private actor. In this case, in England, it was in fact initially the initiative of the public, but today, it is a private operator that manages the repository — at least the technical part. But what I want to say is that it’s completely possible.

Walid: ok thank you. That’s something I find super interesting, because from our French prism, we immediately imagine that it’s a public actor. But in fact, in reality, depending on the country, it can be a private actor, who can charge a fee to the different people who access… In short, there can be many different ways to do it.

Conclusion

Well listen, we’re coming to the moment of conclusion. We’ve been talking for an hour and a quarter already, it’s not bad. I would like to ask you two quick questions before I leave you with a final word.

The first question: what message would you like to convey to the general public? Even if we’ve been talking to people who are, for you, the general public, for a quarter of an hour, but what synthetic message would you like to convey to people like me?

Elisabeth: I think the most important message is around the value of the commons. We are in a society where, more and more, we feel that the tools we are offered are closed, the value is held by a very limited number of actors, to which we, as citizens, contribute massively. And I’m not sure that everyone has that notion in mind, necessarily. In any case, until I worked in this field, I hadn’t necessarily understood it.

And as a result, the common, as a way of increasing everyone’s global knowledge and benefiting from it in return, is really a system that has immense value, and that will allow us not to perpetually feed machines over which we have no control in the end.

I would encourage everyone to take an interest in: what is a commons, how does it work, how to participate in it at their own level? Not necessarily geomatic commons like ours, but commons in general. I think that’s the message I’d like to get across.

Walid: Felix, a message?

Félix: I think I agree with what Elisabeth said.

I think that the fact of having joint initiatives… I’m thinking, I don’t know, of Wikipedia or OpenStreetMap — I think that these are initiatives that are very strong, which, behind them, have the desire to build data on which we are autonomous, we are not dependent on different actors, we control the data that is shared. And I think that’s very important in this day and age. To be interested, even in small concrete things, in these issues, we put our finger on a lot of very interesting aspects.

Felix Veith

So I would tend to say: go and make commons, I think that can be a message. And then, perhaps, go to the rnb.beta.gouv.fr website and let us know if you find the service interesting or if there are any errors on your home.

Walid: Second question: what message would you like to convey to future contributors to the RNB?

Elisabeth : I think the most important thing is to read the definition of a building before contributing. Because we have people who tell us: “Ah, but for me, it didn’t correspond to what they had in my accounting database, so I changed because I wanted it to work for me.”

In that moment, you say to yourself “ah, oops! Well, we’ll do it again…”

But overall, it’s a definition that is well understood. But I think it’s interesting to realize that an object can be perceived in many ways. So, for all our contributors: go see the definition.

Félix: And then, come and adopt the RNB. I think it’s an opportunity to benefit from a database that will be very rich, that is interoperable. And in return, you’ll consolidate it yourself. That’s also a little enriching thing: to think that I have made my contribution. And as a result, we build a community of contributors to have building data that is as close as possible to the field. So let’s get started.

Walid: I must admit that while preparing with Audrey, when she told me the volume of discussion around the notion of building, I fell out of my chair a little. I didn’t imagine that there could be so much discussion around this notion of building. But it’s true that when you think about it…

Félix : And people are all right, too. It’s not a question of saying that we’re right more than the other, that’s not it. It’s more that we have agreed on a definition, so we have to respect it. But on the other hand, it’s really very interesting to see that such a banal object, and that we use every day, is perceived in many different ways.

Walid: Before I leave, I’ll leave you both with a final word. That’s it, if you want to get a message across to the listeners of Projets Libres. Elisabeth, do you want to start?

Elisabeth: I’m not sure I have anything inspiring to say…

Walid: That’s the trick question.

Elisabeth: … No, but I think that…

Walid: It’s always written in the plot, but every time people say “ah, I don’t know what to say”…

Elisabeth : No, it’s in relation to everything we’ve been saying to each other for the past hour. I think that if we were able to inspire either people who work in the public sector, or even in the private sector, which in many ways resembles the public sector in terms of having processes, ways of doing things… Sometimes it feels like there’s only one way to get a result. And I think that the RNB is a good example of a step aside, while still being hyper-integrated within public structures.

So, I think that’s what’s interesting. And if I can encourage people to go and see this in their own organizations, to say to themselves “but could we question? Could I do it differently?”

Me, my sentence — come on, I’ll give you a sentence, not a boat, but an inspirational sentence to finish. It was in Marguerite Yourcenar’s Memoirs of Hadrian, who said: “Even where I was innovating, I liked to feel like a continuator above all.” And I think that’s a great phrase we had during our EIG promotion. I’ve always resonated a lot with that phrase. So, I’ll leave on that.

Walid: Felix, last word?

Felix: I won’t do better than Elisabeth on that, of course. But in any case, I wanted to thank you, because I think that open source project initiatives would have a very strong interest in being highlighted. Because it really means caring about issues of general interest, and developing our capacity to act. So, thank you for highlighting them through a podcast, in particular.

Walid: It turns out that through the podcast, I can take all my questions, my curiosity, ask them to people for an hour and a half. And I must admit that I often say to myself “ah, but what is this? That looks pretty good… Wait, I’ll contact people.”

Look, thank you both very much for taking the time and coming to talk about this. I’m very happy. It also echoes what we said before about the national access point. We were also able to shed another light on state-owned startups, which is a subject that I didn’t know at all before making these two episodes, and which is, I think, very interesting for everyone, because it’s public money, it concerns us all.

For listeners, listen, as usual, share on social networks, talk about it around you. Maybe also listen to the episode on OpenStreetMap too to start if you want.

And then, don’t hesitate to go to the site, look at the different social networks, share on the ones you prefer. Of course, I have a preference for the Fediverse, but all the social networks that are on the site, you can use them, it will be with great pleasure. And I think that in 2026, we will continue to do episodes around the digital commons and this type of initiative.

Listen, Elisabeth, Felix, thank you. Looking forward to meeting you in person one day. And then, who knows, maybe we’ll talk again in a little while so that you can give us some information on governance, where you stand.

Felix: Absolutely, with pleasure.

Elisabeth: Thank you, see you soon.

Walid: See you again.

Episode production

• Remote check-in on November 3 and December 15, 2025
• Plot: Walid Nouh, thank you:
  • Audrey Houssais for the first presentation of the RNB
  • Sébastien Dinot for his feedback and additional questions
• Editing: Walid Nouh
• Transcript: Walid Nouh

This article has been automatically translated from the original language into English.

License

This podcast is released under the CC BY-SA 4.0 license or later