 Okay, welcome, we are at the middle, almost the last session. Well, the topic today is like, well, I want to motivate you to create H projects. That's my idea. To be fun a little bit, to show something different, something practical and push you to start creating something. The topic is building smart farms with H competing system using Kubernetes. I work as a DevOps engineer. My other work is to be a professor in the public university of Guatemala. I am from Guatemala. So I am going to show you like, a specific kind of use cases for Latin America, but you can apply it to another countries in the world, right? Because we are like, from a lot of countries in this room, right? So this talk has three parts. The theory part, the technical or technology part, and the practice, let's say. In the theory, because maybe you are a researcher, you are not a sysadmin guy or something similar. So what is H computing? Well, or H computing refers more to process the information near to the edge. So let's say that you have the data in this room and I have a cluster here where I am staying right now and I am processing the data here, not in the cloud. I am closer to you. So the data is closer, right? That refers to H computing, how you are processing or where you are processing the data. IoT is some concept that you have here and it's like mentioned every time that you are talking about H. So when you connect that device to the internet, it has some kind of smart software or something that collects data or something similar is IoT. IoT refers more like a small devices that perform a single task. H computing, let's say that IoT devices use more like microcontrollers. I know that you have here Arduino. So Arduino is a microcontroller. An edge device use a microprocessor. What kind of device like a Raspberry Pi like I have here? Like here are three Raspberry Pi's. This thing uses microprocessors. What kind of microprocessors? ARM, right? H devices because has microprocessors can perform multiple tasks are more powerful, right? Because are more powerful consumes more energy. Both IoT is different than H computing. H computing refers more, where are you transforming or processing the data? You are near to the source of the data. These concepts are complementary. Maybe you can get that data, process the data locally and then you can move that data to the cloud, right? If it's your use case in a specific or maybe you can distribute that data in different locations. We can say that maybe H computing is the next level of evolution of IoT, but in reality it's complementary. Or who knows, maybe all the people is going to start to forget IoT concept and all the people is going to add up the H computing term right now. Like cloud computing. A lot of people mentioned the cloud, the cloud, but years before of that nobody knows what the cloud was. So what are smart farms? So here is my beautiful chicken here, the farm. I was reading an article and I saw a similar image but because of the copyright I was like, I can use it. So I was trying to design my own kind of farm and the chicken and that kind of thing. I think that this image represents a lot of countries. For example, in my country it's more agriculture. But maybe in other countries are more like chickens, cows and different animals, right? So the demo for this topic is going to more focus on more on plants, not in animals, right? But what is a smart farm? A smart farm, well it's a smart, we can think about artificial intelligence or something that use some kind of intelligence. So smart farming refers, he's mentioning that the farms are using modern information, communication technologies. And using this kind of technologies, they are like increasing the quality, the quantity of the products and our automizing processes to create these things, right? They are optimizing the human labor, maybe automating the things using artificial intelligence, automation using robots, automation using machines, right? So when you are combining your farms with technology, in special with software, you are creating something smart without the human intervention, right? But what kind of technologies we can use or we can find in the smart farms? The sensors, let's say that here I have like my protoboard, a pretty basic sensor, like just to measure weather. If you remember when you were studying engineering, that's the first project of measuring weather, the humidity, temperature, that sensor. So sensors that measure movement, weather, humidity, et cetera, technologies like some software or maybe IT platforms like some, maybe somebody is using like something from AWS, something from Azure and that kind of platforms. I refer to that part in a specific. Do you need connectivity? Like the phones, Laura, let's say that in this talk, I was like trying to present something like, but because of the internet connection that I don't know in this room, I can connect my phone to the internet. So that the reason to don't present my live demo here in the session, right? But I have a video. So don't worry. Technologies like GPS, right? Let's say for example, Waze, we use Waze, we use Google Maps and that kind of things. Robotics and we are analyzing data, right? That kind of technologies form the smart farm. Also big data and similar things. So maybe here's a graph about how maybe a smart farm looks. I don't know if you have seen this video from Mars that there's a guy and a cow in another planet and something like that. Well, here's a kind of reference. The drones are pretty popular right now. Maybe we can use a tablet to connect to the system that is in the farm, near to the source of the data in the plants, who knows. So here's a concept that there's a farm cycle that is based on IoT. They are like, let's say simple steps or parts of the cycle, like observation, that you perform that observation using sensors, diagnostic, with that information you are going to identify any deficiencies or needs in your system, maybe with plants, maybe with animals. So you have to take decisions, maybe using machine learning and an action and then you are going to repeat the cycle, right? Some specific use cases could be like precision farming, refers more to, let's say, how many water do you need for a plant and, et cetera, precision livestock farming are more referral about how many food or nutrition quantity for animals. Automation in smart green houses, like that kind of small houses, some of us, we have experiments at house trying to put some water in some plant using a raspberry pi or something similar, that kind of climate houses and agricultural drones, maybe for perform irrigation across like analyzing images when the drone is flying and something like that. In my country, when they are like checking plants or that kind of things, they perform a lot of irrigation using drones in Guatemala for the coffee plants and that kind of things. Also cacao and that kind of things. Here's some numbers in Latin America. For example, say that Brazil has the largest agriculture and food exporter, their presentation on billion of dollars, followed by Argentina, also Mexico is present. I think that these stats are from 2022, are pretty recent and Chile and Ecuador and Peru. So that's a lot of money that countries in Latin America are doing, like when processing plants, food animals and that kind of thing. So there are big numbers. So we have like a big market to implement smart things there. So for example, here says in the 2018 and the 14% of the labor in Latin America represent that kind of percentage in Bolivia, Ecuador, Guatemala, Honduras. A lot of people is working on that, right? I represent a lot of money for our countries. But this thing is not limited just to Latin America because we have South America, Europe, Africa, Asia in all the continents. It's like just numbers and specific in Latin America but applies to the whole part of the world. So this represents, some people is saying that this is the third green revolution that refers more like, let's say that, because we are using a smart farming revolution, that means that we are using less fertilizer, we are optimizing the quantity. So we have more efficiency for that. So we are, let's say creating safety food for us. We are using less water for that. So we are optimizing that treatments and the input. So we are healthier for our world, right? So we are optimizing that thing. So that refers, this is another concept that I was like pretty, pretty interesting on that because here at the Clown Native Computing Foundation we have a technical group that refers more to a specific environmental sustainability that they talk about how to use the energy in a better way, how Clown Native can tackle that part of things using less energy, be green with the world, right? So when you are optimizing the processes with plants, with animals, we are using less energy. We are creating safety food for us, safety plants, safety animals, bread, et cetera. So we are trying to be green for the world, right? So that's an important thing that this thing is mixed with smart farms, the better use of energy. So the technical part is that, for example, K-3S is a lightweight Kubernetes that you can use. It's designed to use like, not all the components of the regular Kubernetes cluster, just the minimal use. Of course, it's using less energy than the regular Kubernetes. You can run it on a Raspberry Pi, right? You can also use a solar panel to power your Raspberry Pi just using that part to give them energy. So Kubernetes can give you that power also to scale your applications to manage your farm, and that things scale to a thousand of nodes, right? ARM is one of the processors that is more like, balanced, let's say, about power and energy consumption, right? Intel has a lot of power, but a lot of energy consumption. That's the reason that right now, also in the cloud providers, they are promoting the use of ARM processors because of that balance. Better for the world, better for your bucket, right? For, let's say that home use cases scenarios, you can use a Raspberry Pi, but some people use this Raspberry Pi for a real world scenario, right? I think that is a version that the NASA is using outside in the space, right? But you can use a Raspberry Pi. They are like similar kind of devices. What do you need to create your system? Do you need like micro SD cards, put an operating system or your Raspberry Pi, just a small switch like this one, pretty small. Also you can just mask and tape to put all the things together. And your internet connection. You can use your smartphone as an access point. That's the way I use to show sometimes a demo on real time. Some challenges, how you reduce your cost, your power supply, the sensors, how to maintain your images. So with all of these pieces, you have to create a smart phone to measure temperature and humidity for plants. That's an example that I have. The demo part is the last part. So let's create your system. With a Raspberry Pi like this, you can install Ubuntu, maybe Raspbian, that is an operating system based on Debian, a micro SD card, a small switch, maybe an starter kit with sensors to use as simple as you can. A cheaper and easier. Then you have to install your Ubuntu operating system, configure your Raspberry Pi, install K3S with some simple commands. Well, this one is a pretty complicated, more for the specific use cases, but you can run it on one line, right? Then you have to install additional software like simulate load balancers, like Metal LB for example, and some storage using Longhorn and another software like KNET for serverless functions on your Raspberry Pi to save some power. But in this use cases scenario for a small smart farm, just to measure the weather and humidity, we have to install MySQL and Grafana, just that. You have to run your client or your device to send the metrics of our temperature and humidity and how this looks in the device. Well, I have my mini cluster here, the switch, all the connections, a kind of adapter for my Raspberry Pi, the sensor, pretty basic, right? Something simpler, you don't have to do complex things for that. How our system is designed. Let me show you the architecture. You have two devices, maybe one, that you can use as a cluster and a device to capture that metrics or that information. So I have like Grafana deployment and MySQL deployment, some REST API that gets, well, let's say that my H device that has a Python script is capturing the humidity and temperature and is calling an API REST from my cluster, right? So this is going to get information, is going to store information in MySQL and Grafana is to show you the data that is being recorded there, right? And then you have a laptop to check the metrics, maybe you can, let's say the farmer, can have like a tablet to check the dashboard in real time and you can maybe create an alert to say, oh my God, the temperature is a lot, something is not working and I have to perform some optimization for the weather, right? So in that way, you can start your pretty basic experiment there. And here is a video about how that kind of system can work. And let me see. Okay, I think that is on, okay. Here is the dashboard. It's showing the information, right? Showing the information about humidity, temperature and well, the temperature on Fahrenheit and Celsius and humidity, right? So if I put the finger on the sensor, you are going to see that the graph is going to be up, right? And then when I remove my finger, it's going to down the temperature, right? Because the finger is a little bit warm and that thing, right? So let me, okay, you can see in this part how the graphic goes down a little bit. And here's the part of my SQL. And here's the part of my SQL. Just a basic query, right? The number of device return me the temperature on Celsius, on Fahrenheit and humidity. Please graph that on Grafana and that's it, right? A simple basic age system that you can do at your home, right? That's the basic example that I have here, something that you can implement. Some, okay, some code that I can show you. For example, the REST API to get the information. The Python script, pretty basic. Like, let's read from this print from the Raspberry Pi, then call my REST API, right? Then my service is going to get the information and just perform an insert on the database, right? And then Grafana, here is the query. Please show me that, right? And that's it, a simple basic age system that you can use, you can save farmers. You can push them to be more productive with this kind of systems just to measure that information or to see that information in their own smartphone, right? Pretty cheap, without using a lot of money. So Latin America has a lot of potential for that because we don't have a lot of money. So we need this kind of systems, right? So let's return to the presentation. It's loading, okay. So here's the link for the slides. And basically that's it. I think that we have like four minutes for questions. Well, this is the important link, right? The slides and then all the links inside are there. So, well, there's a GitHub repository with all the source code, the containers and everything. Resources that I just said, please remember that environment sustainability is important. They are promoting to be green using cloud-nated technologies. The book I wrote about JH Computing, there are a lot of examples there. I also use the examples there to create my own example and my contact, right? I don't know if you have questions to start implementing this at home. Well, at least in Guatemala, no. So I think that depends on the country. I think that, for example, in the US, I think that they are exploding that kind of opportunity, right? But in Guatemala, we have like, get some support outside Guatemala, maybe with US or some non-lucrative organizations can help that third world countries to start implementing that kind of things. I think that that's the way it works in Latin America, I think, yeah, so. Another question, yeah. I have here, there was some statistics in some site, but I couldn't screenshot the thing because you have it to pay. But I was looking that in some countries, they are like Brazil. They are starting to use these kind of technologies, Brazil, Mexico, and Argentina. For example, in Guatemala, they are technical schools that are dedicated for technology as some to agriculture. I think that there's like some kind of potential in the schools, let's say, because in the university, the government, that kind of things could be complicated, and the money, and that kind of things. Yeah, so it's, let's say that Latin America is a potential place to start moving smart farms, right? Okay, let's say that it depends in the use case, right? Because for example, if you don't have like, also you can use an old computer, right? So it depends. For example, if you don't have that old computer, maybe you can, and you don't have space to put the computer, because like, let's say that a regular computer is like this, and a Raspberry Pi is a small place. So maybe the advantage that you can have with a Raspberry Pi is the size, the energy that a Raspberry Pi can consume. And the other thing is that the Raspberry Pi has a lot of kids of everything, that you can add up sensors and everything, but you have to, also Raspberry Pi has some production versions that you can use and put outside the, in your house. Ah. For example, in use cases that maybe you can scale the thing, and you can automate deployments or upgrade the information. So it has a lot of good things of using Kubernetes there. I think that is time, because of the sign up there. So thank you very much. I will be there for the other questions.