Pierre-Philippe Mathieu is heading the Φ-lab Explore Office (within the Earth Observation Future Systems Department) at the European Space Agency (ESA) in ESRIN (Frascati, Italy). Pierre-Philippe spent 20+ years working in the field of environmental modelling, weather risk management and remote sensing. He has a degree in mechanical engineering, an M.Sc from University of Liege (Belgium), a Ph.D. in oceanography from the University of Louvain-la-Neuve (Belgium) and a Management degree from the University of Reading Business School (UK).
At the Web Summit 2021, in Lisbon, Jorge Pimenta, who is Innovation Director at Instituto Pedro Nunes (https://www.ipn.pt/) and director of the ESA Space Solutions Portugal (http://space.ipn.pt/), had the opportunity to meet Pierre-Philippe and have a conversation about Earth Observation Technology and how it can interact with startups, by understanding their role in sustainability and environmental changes.
Jorge Pimenta (JP): Hello Pierre-Philippe! Thanks a lot for being here. Could you give us a description of your job at ESA?
Pierre-Philippe Mathieu (PPM): Hello Jorge, many thanks for your invitation, I have the privilege to lead a team of researchers at ESA, in the Φ-lab, called the “Φ-lab Explore Office” (see https://philab.phi.esa.int), and the mandate of this outfit is basically to scan the horizon for new technologies (mainly digital), understand their potential value for space applications, in particular Earth Observation (EO), and then create a series of use cases to better understand the strengths and weaknesses. So, it’s a kind of innovation lab mainly focusing on emerging digital technologies (see strategy https://bit.ly/2VJRSrf and LinkedIn https://www.linkedin.com/in/ppmathieu/).
Our main focus is on AI because we identify this technology as the most impactful and scalable approach to the big data issue we have in EO. As you know, in Earth Observation, Europe is in a very good position, because Europe is driving a programme called Copernicus, that builds a family of satellites, called the Sentinels, and third-party missions that monitor our planet on a regular basis with high-resolution – order of 10 meters – across the whole electromagnetic spectrum. We end up with a dataset of 15+ PB (Petabyte) on the state of our planet, the ocean, the land, atmosphere and we have to mine this big data to extract “information”. That’s where the knowledge comes from, but also the business value, so you can turn this data into information services. To enable such process, you need new techniques like AI that can really extract the patterns, correlations, etc. on an automatic and scalable way, and then you critically also need “domain experts” to understand these patterns and their associated physical basis, because you can have a lot of spurious correlations.
So, what we do in the lab is explore this very exciting field of AI4EO: we try to transfer technology from the AI world – the mathematicians that are doing the theory of AI, the application to medicine, etc. – and we transfer this to the field of Earth Observation. And that’s why it is important to meet new people, new AI talents, outside the space sector that can disrupt our business, by bringing their knowledge to our field.
JP: That’s really interesting, Pierre-Philippe. You see that a lot of startups are also moving in that space and a lot of things are appearing there. How do you see startups interacting with big organisations, like ESA, and how can this Earth Observation knowledge and information be integrated in everyday products with startups?
PPM: This is, in fact, one of the challenges we are tackling. We feel that there is a lot of potential out there for innovators and entrepreneurs to create a new generation of services with high commercial value and the problem is that they are not aware of these datasets. When I go to an AI conference and I say that we have open data (because we have an open-data policy) with a lot of data on the state of the planet, that everybody depends on the environment and that they can create services with AI that could provide solutions to address some of the biggest problems of mankind, the AI community gets really excited! There is a way with this technology to actually do good for the planet (so called “AI4Good”).
The idea would be to bring the AI talents to our world and, to do that, we have Space Solutions, we have ESA BICs (Business Incubation Centers: https://www.esa.int/Applications/Business_Incubation/ESA_Business_Incubation_Centres12), we have programmes like in InCubed (https://incubed.phi.esa.int/) that has a strong commercial angle, where entrepreneurs can pitch and then have two-stage proposals to be supported by ESA at 50%. There are also EO programmes where we have open calls for innovative ideas. There is a variety of tracks for people to interact with us. They could also just have a look at our data and try to integrate them directly into their processes, because the data is out there on the web (google sentinel hub). There are some specificities of these data, so you need a bit of training. In particular, radar data is not as obvious to use as optical data.
The potential is there and what we try to do is to have the meeting of this demand – from people who have problems and need your spatial solutions – and supply – from people who have the solution and the technology and can apply them. The lab is trying to work at this “intersection” and that’s “where the magic happens!”: when you put different communities together around the same problem trying to find new solutions for these problems in a very interactive cycle and rapid prototyping cycle.
One example is that InCubed created what we call an AI4EO Solution Factory in Germany with DFKI (German Research Center for Artificial Intelligence), which is the leader in AI in Germany, and they create an environment by which solution providers, like startups or experts in AI, can meet big corporates who also bring data to the table and their problems to shape this kind of new solution.
JP: Where do you feel are the main challenges, as we move to this new era of sustainability, where environmental changes are happening every day? Do you see that startups have a role and an opportunity in that?
PPM: They have more than a role, they might have the keys of the castle. Startups are the “engine of innovation”. There is this adage “that you cannot manage what you cannot measure”. This comes from the business side but is true in all aspects of life, your own fitness/health and etc. The data we provide in the system we put in place is a monitoring system so it’s really giving the “health” of our planet, but you need the doctor, you need someone who can interpret these data and provide the real information to the people. This is the only way we can measure our track to sustainability. There lies this whole debate about the Sustainable Development Goals (SDGs) which, for the first time, integrates environment at the heart of it. People in the United Nations realised you cannot have a healthy economy without having a healthy environment, so they integrated key aspects of our global environment but didn’t really know how to measure it accurately and consistently at the global scale. So, now, there is a debate on how to measure this globally, because the issue is that you can always measure locally, but the way of measuring in one nation might be different from another nation. By using the unique vantage point of space, satellites have this ability to measure consistently and globally, so they certainly play a key role in that monitoring.
The startup would come with the tools to actually transform these data into the information you need to manage our planet. One aspect that is very close to my heart is the footprint we have. We are not aware of the real-time footprint we have when we buy goods, when we do actions, when we make policies, so we need a technique by which we become conscious of our impact – which we do locally, but the impact is global. Using all the data – from space, IoT, drones, open data on the web – and aggregating them and extracting this real information is certainly one of the critical impacts for moving towards sustainability and raising our global consciousness about what we are doing as a species. There is a whole business there because doing good for the environment is doing good business – it’s a whole space that a startup can use to create services that today do not exist. The elements are there, the technology is there but, again, we need this intersection and this feedback loop between the people who will be using the information and the startups.
One example is the Digital Twin Earth (DTE). This is a big initiative of the European Commission, and ESA is playing a role into this, to build the most comprehensive view of our planet, integrating all type of data but also predictive capability by using models. The idea there is like using a virtual laboratory: you have only one planet, there is no “planet B”, but you could do experiments in a virtual planet to quantify the impact of your policies and actions. For example, if I stop deforestation what’s the impact on climate, if I do this management of the land what’s the impact on climate and water, etc. So, this would be a laboratory for knowledge, for science, but also for creating new environmental services.
JP: ESA is actually a well-known agency but do you feel that ESA’s role in connecting to startups is well known by startups or do you still need an effort to reach them?
PPM: It’s always an ongoing effort, that’s why we’re here [at the Web Summit]. The ecosystem of startups is also growing extremely fast. As you know, today, people want to have an impact on mankind, they want to be in their own business and they want to lead. These are the values of the new generations, so the number is really increasing and we need to connect with this new generation with many efforts, by coming to events like this and stimulating entrepreneurship. Building this connection is like a relationship – it’s long-term, it’s continuous and it’s investing on both sides. We have a different track of activities but, of course, if you have suggestions, they are warmly welcome!
Events like this one hit me, because AI is a very agnostic technology so, when you have a startup doing annotations for medical research or a startup doing blockchain for cryptocurrency enabling a digital asset to be exchanged, all this is highly relevant for our business. I would say 80% of what I’ve seen today is relevant, but nobody knows it can be transferred and applied unless you have this conversation, so enhancing this dialogue and conversation is key.
JP: We are just finishing our interview. Do you want to leave a message for startups? What would your message be?
PPM: My message is that I hope that entrepreneurs here could realise there is a big opportunity to put their talent at the service of the planet, by simply looking at how they could implement their technology with space technologies, data from space, to build this awareness about our impact on the environment and measuring it but also making it greener. The message is: good environment is good economy and good business, and they have the tools, talent and energy to make it happen and we can provide data, coaching and services to help them achieve that. Data is out there, make the most of it.
ESA BUSINESS INCUBATION CENTRES: THE LARGEST SINGLE INCUBATION NETWORK IN EUROPE
ESA Business Incubation Centres (ESA BICs) are the largest network of space incubators in Europe. The main objective of ESA BICs is to support entrepreneurs with a space based business idea, thereby creating and growing clusters of space related start-ups across Europe.
All ESA BIC are managed by local champions who connect their ESA BICs to their local industry, universities, research organisations, government, and investor communities, while also maintaining strong regional/national links. This makes the ESA BICs very well embedded in the local communities as well as the place of choice for all space related innovation and business.