Interview with UAVIA’s CEO, Pierre Vilpoux, on our collaboration on the Total Energies and CNES AUSEA (Airborne Ultra-light Spectrometer for Environmental Application) project

UAVIA is investing heavily in autonomous monitoring and quantification of greenhouse gases and toxic emissions using drones and other intelligent robots.


UAVIA is a Deep Tech software publisher;

What we do is provide a software infrastructure that allows people like Total Energies to use drones quite extensively on all industrial sites, without a driver, in the simplest, most secure way possible, to process and put intelligence into the data.


In other words, we put an intelligence embedded inside drones that replaces the pilot. There are means of communication, there is on-board artificial intelligence so that we can give high-level orders from a unified interface, whatever the drone, to tell the drone to “go and measure this”, “go and observe that”, “go and clear up any doubts”;

And the drone receives the command, interprets it and decides on his own how he’s going to carry out the mission by autonomous navigation, taking maximum account of safety issues, on sites that are often reputed to be dangerous.


For example, AUSEA is an important use for Total Energies: when you deploy drones, you send remote pilots with a machine to the site; (Remote pilots are people who have been trained in a special profession that is not the Energy business), and who arrive to carry out a mission; The person in question is a drone specialist, but not a specialist in the field, and doesn’t know what to measure;

So being able to free ourselves from that means that people who operate remotely can operate remotely, don’t need to go to an offshore platform, can be behind their desks; They know exactly what they’re looking for, they are data specialists, emissions data specialists, and they give high-level orders and as a result the operation becomes very simple to assemble:

A drone can be left on station; It has our on-board intelligence; He receives the order from here, from the dome if need be, he carries out his mission and the data is transmitted back in real time with the interpretation that is made;


Qui dit autonomisation dit passage à l’échelle. We work on the principle that an industry that emits polluting emissions must have the means to understand what it is emitting precisely, so that it can take action and monitor the progress that is being made. From the moment we want to systematise these operations, we need to move towards autonomous and intelligent solutions;

We’re adding intelligence to the capture of the data; It’s much more comprehensive, thanks to on-board intelligence; we put into it than if we weren’t there and deploying these solutions is much simpler and much cheaper. So yes, we think this really is the future for this type of measure;


The story is quite interesting: UAVIA and Total Energies began working together in 2018 on issues that did not involve emissions; We didn’t know about the project at the outset; In fact, we put intelligence into drone operations firstly for HSE issues, crisis management, maintenance, inspection of installations, and then to remove doubts about safety;

For us, safety was already part of an environmental issue, meaning that you operate sites that are dangerous and which, if they are damaged, can lead to environmental disasters. Donc on a commencé à faire ça et nos technologies ont été prouvées. At that point, a cross-functional approach was adopted at Total Energies; We were introduced to the AUSEA team and told “perhaps the technology you bring to drone operations could be of interest for this project”

And that’s the case, i.e. CNRS and Total Energies were collaborating on a sensor, and this sensor was designed to be fitted to drones with complex environments and pilots, to take measurements on site and process them off-line. From the moment we started working together, we realised that we had a perfect fit with the CNRS: in other words, their sensor was attached to a drone; we added our on-board intelligence, we processed their algorithms inside the drone, we accelerated performance and processing speed;

Which means that, to keep things fairly simple, the measurements that were originally about 2 to 4 hertz so 2 to 4 measurements per second, we’ve managed to optimise them to 24 hertz. This means that between two measurement points on an oil platform, we’ve gone from about 1.50 metres to 20 cm. What does it do? This allows the drone to operate at a faster speed, so operations take less time, are not interrupted by the drone’s autonomy, and avoid missing emission peaks;

We can have methane emissions that are identified over a metre and at that point, we can have exhaustive measurements. And the acceleration also means we can loop back on the drone’s autonomous navigation. In other words, our on-board intelligence processes the data, the drone has the concentration measurements and, depending on what it measures, it will be adapting its autonomous navigation algorithms to make sure it measures everything and doesn’t fly into places where nothing is happening;. And all this is aimed at improving performance, data exhaustiveness for more accurate volume sizing, and ease of use;


C’est un enjeu comme je le disais de performance et de qualité des mesures. In other words, today, I hope that most of the industrial companies that generate methane and CO2 emissions have activities in this area;

Today, Total Energies is not content with what already exists, and has gone in search of more advanced performance through R&D with partners, with the CNRS and with us; As soon as we achieve these results, what we are doing is also investing massively on our side, alongside what Total Energies is financing; To turn these issues into technological standards that can be propagated even beyond pollutants such as CO2 methane, so beyond greenhouse gases, to other pollutants. So that the industry as a whole can adopt this kind of process, standardise it and so that it becomes, I would say, common market practice.

And that, from my point of view, is extremely virtuous because once again, owe can’t imagine an industrialist only using data that would come, we’ll say, from satellite vision, we’ll say from a global policeman to taking belated action on its own emissions. So from the moment that he leads the movement towards the development of solutions that enable him to understand more precisely what he is doing, that’s how we set in motion, I would say, a virtuous process towards the minimisation of emissions;


What is obvious to me is that any industry that does not give itself the means to monitor its own emissions with great precision will not be moving towards reducing them.

I don’t know if you’ve heard of Climate Trace, which involves satellites that observe the Earth and make it possible to determine the size of site emissions; And then potentially slam a manufacturer or a country and say: you emit so much;

So we’ve established the facts? Everyone says “when you look at a problem you’ve already solved it”? Except that no, he who looks at the problem cannot solve it. And so the sovereign approach of a manufacturer who invests in R&D to give itself the means to be precise and systematic on its pollutant measurements, in any case, from my point of view, it’s a sign that it’s doing this for something and therefore probably to reduce its emissions;

I didn’t know anything about the energy production business; I’m a telecoms and AI specialist; The subject engages us, it engages my whole team. It’s become the majority of our investments, we want to focus on that subject. Today, contributing to that, as a technology company, we believe that we are doing our bit to help move towards cleaner energy sources. To think that before our solutions, including drone measurements, a manufacturer or entity that measures could be missing or not be exhaustive in its emissions, for me, that means that we’re improving the situation quite significantly.


For me, this project is very differentiating for us because it uses all our technologies, I’d say to the maximum of their performance.

It’s always exciting to think that we’re innovating, that we’re pushing back technological barriers and that we’re doing so in the service of a subject, a use, that makes sense for all our teams, for all our shareholders, and potentially for all our customers; It’s different because, in my opinion To date, no-one else has implemented what we’re doing with the CNRS and Total Energies;

I think that Total Energies is going to deploy it globally on all its sites, and we are going to make progress with our technology to ensure that measurements of this type, including with partners such as the CNRS, become market standards; So that the industry at large can appropriate these technologies, each with its own particularities (not everyone emits the same thing, there are different forms of pollutants, there are different sensors); And wewe’re going to try to standardise our technologies, always being as high performance as possible so that it becomes a global standard, a standard adopted by all industries.


Non pas du tout. J’ai cité les sites offshore parce que ce sont les plus complexes. À partir du moment où on arrive à faire des choses efficaces sur des sites offshore, en gros on sait le faire partout. There’s one thing that’s perhaps a little simpler offshore, and that’s the regulatory side; Since we’re often outside territorial waters, or not directly, local authorities are probably less careful; But in reality, a company like Total Energies has HSE procedures that are often much stricter than what a regulator would impose;

So all in all, we manage to facilitate operations on offshore sites but all production and processing sites are either greenhouse gas emission sites or sites that present SEVESO-type risks of pollutant emissions in the event of an industrial accident.

So the aim of our technologies is to measure and continuously monitor greenhouse gas emissions; And then it’s also about being able on other issues to make more spot measurements, but automatic, very reactive, in the event of an industrial problem in particular so that manufacturers or security forces, firefighters, can make decisions.

Today, when you have an industrial accident like the one that happened recently, including in France, you have models for the spread of toxic pollutants; Firefighters have access to this information and have to make decisions based on modelling, potentially evacuating areas to protect people; Having the ability to check these predictive models in near real time to confirm or refute what is happening is another key challenge that we will be tackling at UAVIA using the same technologies;


A company like ours is gradually expanding, finding specific outlets and one day, there’s a subject that you realise is general, topical, urgent, has significant traction and that your technology brings something special to the table; That’s exactly what we’re doing;

So we decided at the beginning of the year with our investors to put in place a a major four-year investment plan to move as quickly as possible and as far as possible towards the widespread use of these solutions.

And our aspiration is simple: to become the world leader in these technologies;