 Our next presentation is by Alberto Massa from 3StartConnect on counting fish from forests. Welcome Alberto. Welcome everybody. My name is Alberto Massa. I'm a software and hardware expert in artificial intelligence, prediction and IoT devices. My expertise took me to work with the biggest in AI like Google, Amazon AWS and Intel. I've been working with the most brilliant mind in artificial intelligence around the world with a special collaboration of Eric Cambria, professor at Nayante College University of Singapore. He was listed in the best 10 brain in the artificial intelligence feature in Forbes. So why am I speaking to you today? Today I'm going to give you an insight into our bridging the digital divide can contribute to a sustainable future toward a digital planet for all. Thanks for watching. The project we have talked about today is the Resilient Rivers Mobile App. It's one of a kind app that will allow to counting fish from forest for food security and will be a game changer in many areas where fisheries data collection over many years has been quite a challenge. The digital transformation live from technology that I'm going to present to you relates to four key points. Multidisciplinary practice counting fish from forest with more accessible digital resource for stakeholders. Innovation for empowering accessible digital participation for user at the front line of aquatic food security. FAO fisheries and forestry, the Resilient Rivers multidisciplinary project aims to bridge fishery and forestry management sector through a more holistic approach to watershed management through collaborative practice and bringing data from many dimensions together. Data knowledge gaps, how we can address data deficiencies, freshwater ecosystem with new way of collecting data related to habitat condition and food security. Over the past 20 years or so fishing data collection in Zambia as in some other regions of the world has added sufficient resources to implement a full picture of the state of ecosystem resources. Data tends to be sporadic and extrapolated between surveys which are sometimes only partial data sets. One of the reasons for this is due to the limited resources that the fishery officers have to collect data. Often data has been collected on paper but there are few computers available to digitalize this data set so aggregating this data also become a problem. So a solution to tackle these issues is through a mobile app. The app needs to be a set of clearly defined user friendly tools to enable data collection to provide more information for improving management decision about fisheries and forestry resources. Not only will be very easy to use but also easy to maintain due to the low cost in the mobile tech. We strategically handle all the big data costs. One of the big challenges in the cloud industry by the app identifying the minimum necessary data and compressing it before it's uploaded to be analyzed. We are also exploring ways to incentivize data collection so that not just for fishery extension offices but also stakeholders can play and activate the role and participate in resource management. It's very important to frame the time of the app in terms of sustainable development goals like through the social element of the app to empower stakeholders to build a community of practice for responsible production and consumption or empowering social interaction within a food security and ecosystem-based management context. The stakeholder app enables participation in data collection from downstream stakeholders in value chains, particularly women, trading and processing commodities who frequently have access to more data than primary producers. Verifiable data set. This asset can be processed both by using machine learning or verify using manual analysis of size, count and species, using photogrammetry and traditionally taxonomy for inmate data. Knowledge gaps. Collecting species-level data versus cages enable monitoring of invasive species. Cases enable water quality, monitoring and ensign from increased data collection frequency. Leap from technology, innovation to bridge technology gaps enabling a greater level of inclusion and low-cost data collection without large-scale investment in computers. We are at the end of the presentation and I am very grateful for your interest on our ResilientRiverMobile app. Thank you for your attention. Thank you very much, Alberto. It's great to see you working on ResilientRiverBasins. As you can see, I'm not an actor. I think you did very well. It's good to see people working on river systems, something that definitely needs global attention and often is left out of the mix when we're talking about aquatic system. I'm just wondering when you're trying to develop these apps that are going to be used by people on river systems in Zambia, for example. There are many challenges. Yeah, it must be very challenging even just to roll out some type of system and get feedback. Have you found any other groups doing this where maybe we could, for example, set up some type of community around how the apps are being taken up? What's working? What isn't working? Have you just relied on your experience of building new apps? Yeah, I've relied on it. As I mentioned, I was working with Google. I was working with Amazon and AWS in many startups. And the problem also in our world, in a normal situation, is to involve the people. If it's not a social app or social dating app, the people are not really involved in to give the personal time to collect data, for example, in the fishery. So I just watching all the previous presentation that are all very nice and very clever solution. The one that is close to us in our app is the one of the tuna scope. The tuna scope, they just give a software as a service, the SAS. Why? Because they took something that already is going in this whole year to recognize the tail of the tuna by eyes. And so they take practice in a set of collection of all the different quality of the tuna and give a nice tools. So the people, of course, they just change. It's a game changer from a normal person that are going in extinction because the young people, they don't even know how to recognize a tail. We got a very good tools that we call in US software as a service. So we give something and we collect data. Yeah. It is exchange. My vision, my vision for the app and involve the people is about, OK, we got the officer, the fishery officer that they got to do their job. So they go around, they take a movie, they take a video movie. We convert all the information and we got a data set. But this is just a part. We want to evolve also the woman. The last in the chain is the one that they go in the market, take the fish, go home, cook. And this is where we need to have the security in this case. So our app is, OK, we recognize the fish, but also we want to give to this woman, to these people also in the market, the one that they got to sell to the end user. In between between the fish and the last woman that they got to cook. We want to give this app that help them to recognize also the quality, because I just was trying with some data feature. I have the problem that when the fish was sick, so he got like a bite or normal disease. So one is that we didn't recognize well the fish. So I had to create a library for all the diseases. And then I say, OK, I give you a solution for all these women for all these people. They, they take a picture of the fish. They straight away know if he's good in good shape. The size is not an old one. If you got some some spot that can be a, I don't know. I'm not in the insology, but I got some philaria, I say, like my dog can have it. And so they already recognize that the fish is good. Then also, based in my research, they can also, when they open the fish, for example, a salmon, I understand that the salmon, if it's coming from a wild environment or they come from the archery, there's a huge difference. They got the line of the fat that is bigger or hurt. So it's also another quality of fish. So this is the tools that we want to give to involve this person. About all, also we want to give a reward because some people think, OK, I got to work for the FAO to collect data or whatever. But I'm doing, nobody want to work for free. So we give you something that is already good for you. But if you collect more data, I can give you a reward. The rewards can be, we can ask Google to give a free YouTube movie or AWS something from Amazon. There is a way or just I send you a pair of pen to cook a better cook. Or I give you free recipe. So we need all the data can come by themselves. And we cannot ask the people to help us to make a better world because they got already something to do in their spare time. But if it can give you something reward, this data collection will resolve many more gaps that we have in this situation. Another problem that we see is about the connection. Not all the people can have a good connection. So what we're going to do, we're going to do on the edge, a small library where we already recognize. For example, if we are talking about Zambia in a lake, so we probably have 10 pieces of fish, 20 or whatever. So we just reduce a small library on the edge in a way that I recognize straight away the fish. I recognize if it's a sick or not for the food security, so we know that it's okay. And then once I got the connection, we just synchronize all the data with our server or compress it. So we try to move in this way. Thank you very much. Elvete, it reminds me of the talk we had yesterday from Scott Nizum, who was also trying to bridge the gap between data that comes from the fish or all the way to data that comes from the consumer and link those two. And you can see in the future from your story that potentially in the future, those relationships will start to link up in real time. I have these fish coming in and the lady knows or the man knows that I'm coming down to, we can do this. It's going to take us a while, but that conversation is already visible in the distance and hopefully we can get there. Matt, have you got one question by any chance? Yeah, I do. I think what's interesting coming from Alberso's knowledge is a lot of the software that we're using across every presentation that we've seen has not been born in our domain at all. Typically photogrammetry, point tracking, SIF, SLAM, all of these things come from technologies such as aerospace, missile tracking systems from the 80s, and then the AI algorithms for the rebuilding camera positioning, camera tracking, and 3D modeling, texture recognition. All of these things are like third hand tools. And in the most recent development in AI and machine learning, I think, are typically in marketing online for scraping data in the world. And we're really using these algorithms as almost fourth hand technologies and applying them to a real world scenario, whereas typically they'd be based on 2D images of what people bought from a shop. Alberto, from your experience, where do you see AI going and what's the trajectory in the next, say, 10 years? In our review, what we are talking today. Well, no, just from your knowledge of, you know, in the normal domain. There is one word, prediction. Injustice prediction. So, as you say, we have a lot of scrapping tools around with the artificial intelligence. That's what we're doing. For example, once in a project that I was following Prada is one of the biggest brand in the world and the fashion. They start using the prediction by scrapping around Instagram. They just was to see how the people will dress. In order to give a product, they do already like it. For example, me, I use the Polo. I use the Polo or jeans, special color. So they was just coming out with something that you already like, because was the common user in the young people or in the middle-aged guy. You know, so they was targeting the people. For me that I'm already 50, they say, okay, I use Polo also because I'm getting fat. So the shirt is a little bit. Also, that one was a data set to see a big round face Polo. This is fat. So there is a million people like this more in USA. So let's focus on selling Polo for big size guy. So this is the evolution that is coming, the prediction. Like we are doing also in the factory in 4.0. We use this kind of guys. Now this is a demo. This is the small one that we put. And we know if electric motor is going to die. So we keep a maintenance before this happened. So we don't break a production in, I don't know, in a chain of pasta, the factory of pasta or whatever. So this is the, because the future also, and now I'm talking about the green planet. So I can see a future. We are working on the smart city. We can work also in the smart forest. This gap that we use to ask these people, the end user to help us to collect data can be a normal. A small camera IoT device that can last five years with a small solar panel in the marketplace where all the fish pass. Or we can put all in the forest. Also for animal in the in the forest, not only in the water, in the water, the challenging as a, I think it was Amanda, maybe. Amanda that you got the camera in the water. So they come in the water. You know, it can be a gas can be dirty in a quite a very soon. So, but can be a solution for that. And with the camera and in our, the things network for, for green planet. We can record, we can take all these data collection and we can have 100% cover the situation of the green planet. Yeah. I think it's really important for us as people that are interested in ecosystems resources to bear in mind that the main financial sectors as using AI and deep learning and neural networks to predict the future now. And that's going to have an impact on consumption of resources on production of greenhouse gases and everything that links that so they're already, they're already predicting the future because they're interested in financial investment, which relates back to us right here right now. Discussing fish, actually. That is the job of fisheries management is prediction. So your whole role in fisheries management is to try to see the future and hopefully make the right decisions for the fisheries so they're doing the right thing before you get into trouble. So thank you very much, Alberto. That was a very insightful talk and hopefully the way the waves of data coming in will make us better informed to make those predictions wisely.