Transcriber: Octavio Delvecchio Reviewer: Renée Arias (Applause) So we all like natural aquatic environments, like a river, for example, with transparent water, green vegetation, birds singing, or a wild beach washed by blue waves, or the ocean full of fishes and corals or the crystal clear ground water. Don't worry, I'm not asking you to imagine the holidays of your dreams. In fact, the scenario is going to change drastically because I'm going to show you this much less nice picture. This is a natural aquatic environment, but the water was dramatically contaminated. So now I have a question for you. Do you know which is the link between the two scenarios? Us. All of us, we are the link. But why? Because human beings are very inefficient machines from a waste management point of view. For example, when we eat, we are assimilating just a part of the food. All the rest is wasted in form of solid and liquid excrements producing wastewater. More than 50 cubic meter per year of wastewaters are produced on average per year of yarn. But not only. All the industries that need to use water in their production processes are producing huge amount of wastewaters. So now just think to the huge quantity of wastewaters that can be produced by the 8 billion people on this planet. But what are exactly these wastewaters and where do they end up? Wastewaters are basically water at the beginning, taken from natural water sources for the production chain of... for being used in the production chain of many anthropic activities for all our exigencies, more in general. At the end of the production chain, this water is not anymore only water, it's water plus contaminants. Consider that in our waste there are also contaminants like antibiotics, hormones, pesticides, viruses, colorants, and many, many other toxic compounds. So basically, before giving back to the environment, the wastewaters, they pass through what is called the wastewater treatment plant. Is it the end of the story? No. Unfortunately no. Treating the water we are contaminating is very expensive, energy consuming and a huge source of gas emissions. Just consider that wastewater treatment plants are responsible of about 3-5% of the total anthropogenic greenhouse gas emissions. But not only. The problem is that together with our waste, we are also losing very precious elements that we would reuse also as energy sources. For example, nitrogen. Nitrogen is an essential component in the proteins of our body. But nitrogen is also a key ingredient in fertilizers. Just think that in the standard way of producing one kilo of nitrogen for fertilizers, you need to burn about one kilo of fossil fuels. And for this producing also carbon dioxide. And we are wasting the nitrogen. Another example. Phosphorus. Phosphorus is a fundamental component of the molecules in our body that are carrying the energy. Phosphorus is also a key ingredient for constructing car batteries, for example. Consider that the natural sources of phosphorus on our planet are limited. It's going to finish. So we really must think to the fact that we are wasting it. All these aspects cannot be ignored in an era where it's so important to find new ways for renewable energies. So how can we at the same time remove contaminants and recover the precious elements and the chemical energy in the water we are wasting every day? How can we reshape the wastewater cycle in the logic of a circular economy? And here we are arriving to what am I working on as a researcher at Inria. I am exploring and studying new biotechnologies for treating wastewaters in order to find more efficient solutions in terms of gas emissions, energy, costs, and also providing renewable energy sources. But before arriving at the core of what I'm working on, why biotechnologies? You must know that in nature a micro world is active, made of microorganisms, and when they are all together, they can do very beautiful things for the environment. That's why we want to exploit and amplify their potential, growing them in an environment that we can technically control called a "bioreactor". In particular, I am studying bioprocesses involving microalgae and bacteria. We can grow them in wastewaters because they are able to use the contaminants in our waste and to transform them in other molecules. But wait one moment. Who are these guys? Microalgae are very fascinating microorganisms. They can accumulate nitrogen and phosphorus in the cell, for example. It means that when they are growing on wastewaters, they can accumulate the nitrogen and the phosphorus in our waste to transform them in other molecules, valuable molecules, that we can further extract for producing biofuels, bioplastics, biofertilizers, other byproducts, so entering in this logic of circular economy. If you want, microalgae are like microscopic plants growing in the water. They can perform photosynthesis. That means that they can capture the energy from the sun to transform the carbon dioxide in biological materials. Like all the plants, in consequence of photosynthesis, microalgae can produce oxygen that they can provide to other microorganisms to grow, like bacteria, instead of supplying it artificially. That means saving lots of energy and money. Bacteria, apart of consuming oxygen, can provide lots of benefits to microalgae because they can provide vitamins for growth, breaking down complex molecules for algae, and also provide them with carbon dioxide for photosynthesis. So growing hundreds of different species all together in a bioreactor, we have created an environment where billions of possible interactions are possible. And understanding these billions of possible interactions is actually where things become very, very complicated. So how we can understand the dynamics of this micro world of interactive microorganism? How we can control it? And here, we finally arrive to what I am working on as a researcher at Inria. I am developing mathematical representations of these ecosystems to see what is happening but in a computer. Mathematical models are very powerful. You can see the invisible thanks to them. They can really help you in understanding what is happening behind the scene of this world of microalgae and bacteria, and how to direct it for giving clean water back to the environment and for recovering the precious resources we are wasting every day. These models are mathematical monsters, very difficult to master. Adjusting the output of these models according to the reality is very challenging and requires real data. Experimental measurements. That's why in my research group at Inria, we are a joint team, together with biologists and microbiologists from the Laboratory of Oceanography of Villefranche. So developing a mathematical representation of the reality seems a very exciting job. And indeed it is. But sometimes, this micro world of interactive microorganisms, it's so integrated that its complexity is pushing us really at the limits of our understanding, of our knowledge. And sometimes, indeed, we really do not understand. Anyway, giving up is never the favorite choice of a researcher. We must perseverate for weeks, months, sometimes for years, until we find a way that can work. And this is actually what I am struggling with in my job. I am trying to find a new way. Instead of being stuck by the lack of knowledge in biological systems, I am trying to explore a new numerical approach that is involving artificial intelligence. Artificial intelligence nowadays is widely used for plenty of applications like image recognition or cars that can drive alone. But still, it was not deeply used to explore the complexity of biological systems. And that's why I want to involve in this story of wastewater treatment, artificial neural networks, because I do believe that lots of key informations are hidden in the data we collect. And artificial neural networks can learn directly from the data. For example, we often observe phenomena of inhibition or limitation of the activity of these microorganisms, and we don't know why, but artificial neural networks can find in the data the reason of the problem, the information we are missing, because they are mathematically designed to have a learning process similar to a human brain, even if, of course, it's an artificial brain. So artificial intelligence can really give a very important contribution for making more efficient this synergistic collaboration between our friends, microalgae and bacteria, to transform our waste into resources, to reduce gas emissions, and of course, to give clean water back to the environment. So I want to use artificial intelligence for climate change mitigation, for reducing the energy need and for giving clean water back to the environment. But all this, all this is just the top of the iceberg. Artificial intelligence has really the potential to give a very fundamental contribution to plenty of environmental and ecological efforts, because these algorithms can process big data sets from a huge network of environmental sensors. They can track deforestation, air and water quality, or the health of an ecosystem. For example, artificial intelligence can detect the biodiversity just from the noises recorded in a forest. So, in conclusion, by exploiting the power of artificial intelligence, we can really gain deeper insights into environmental challenges and work towards a more sustainable and ecologically balanced future. Merci. (Applause)