Hello. It's super nice to be speaking to you today. I don't know if it's today or early evening for you. For me. It's evening al almost uh but I'm going to be sharing with you what I love. So I'm very hyped about this.My name is Ekaterina Tamboli, but people usually call me Katia or even shorter cut because it's, it's faster, it's easier. Um And I have been working with uh space technology or analyzing satellite images for the last two years of my life and the previous five years of my life, I had just worked on um using data science for social and environmental problems. I was having a couple of doubts on whether uh on which is the best format to present today. And I think that speaking of only about um what my company does, it's uh not the best way to go forward. Speaking only about me personally, it's also not the best way for us to go forward. So what I decided to do is share with you my story and my journey on how I became a space tech entrepreneur and what happened before that and what's going to happen uh in the future, I don't know if any of you are um technical people because as women in tech, we might be working in the tech, tech part of things.

So we might be working on the business part. Hr marketing, you name it? Uh Hi. Yeah. Um and then for me, I'm both um a businesswoman in tech and I'm also very technical so I can code uh even though nowadays I don't have that much time. So the way I decided to structure this uh following talk is um the following. So I'm going to tell you a little bit about my personal story. Then I'll tell you a little bit about um technicalities and analyzing satellite images. And then I'll tell you how I managed to combine both so that I can have a job that really makes me happy. Uh That also scares me a lot because it's um it's a job of responsibility. Uh But it's also a joy ride. So um what I'm going to do is I'm going to start with the um personal story first. Uh So what happened is I studied computer science in, at university. And then for my master's, I did natural language processing, which is part of the um field of artificial intelligence. So by education, I'm a artificial intelligence person. And then over the last 10 years of my professional life, I uh also became uh quite good in data science in general. So analyzing all sorts of data. And then at the end of 2017, I found myself at the British Interplanetary Society um, at a talk during which I was hoping, I was hoping the event and the preparations of the event. And what happened is I was a little bit cold.

So I went outside because the ac was blowing really cold air. And there I started chatting to the security guy who had been working at the British Interplanetary Society for um almost 20 years. And I was telling him, hey, you know what, in 10 years from now, I'm going to be working in the space space sector, which is growing now as you know, SpaceX, SpaceX is happening. Uh A lot of other companies are building their own uh launchers or spaceships, whether they should want to put something in orbit or go to other planets. This is a different uh story. There is also a lot of exploration on space mining are sending probes to mine precious materials from asteroids or the moon. Uh And I was telling him about this and I was really excited. And then the only thing he told me is why wait for 10 years. Why don't you start it uh next year? And this is actually what happened in 2018, I found my co-founders and we decided to create a space tech company to solve an environmental problem. So I wanted to, to start with this because as business people, regardless of whether you're a business owner or an employee, we always feel happier when we have a purpose or when we have a dream. Uh and we want to make this dream happen.

And as soon as we reach this level of dreaming, we start dreaming bigger. So don't get discouraged. If, if someone's telling you, you cannot do this, this is impossible. Just read a lot, see what is possible and what is not possible and work with um what you get on your hands based on your knowledge to start with the dream. And my dream was dream a little, a little bit of a of space and with knowledge. So as I told you, I'm very technical, I am no data science, artificial intelligence programming, of course, programming is the tool to allow me to um do my artificial intelligence projects for example. And this is how it started. So this is the personal part of the story. The next thing I'm going to speak about is the more technical part of the story since we only have 20 minutes, I didn't want us to um go deeper in technical details and tell you which programming language to use so that it's easier for you to to analyze satellite images and how to do this.

What I want. I wanted to make sure is that you get to understand how analysis of satellite images works and then as soon as you grasp this, so as soon as you receive the knowledge, then it will be very easy to implement it in whatever project you fancy and with with whatever tools that you have in at hand.

So here the science is quite important because we need to also to understand the natural conditions of the world so that we understand what satellite images are telling us about. So satellite images are just another form of data like numbers, data or just textual data. And what happens is we have, I hope that you will see my mouse. I don't think you see my mouse. OK. So we have the sun which is emitting radiation. This is the correct physics terms uh term for this and what is radiation, it can be thermal energy, it can be the visible light. So the visible light is part of all the radiation that is emitted from the sun. And then this radiation bounces off the um the surface of the earth and is reflected back into space. This is true for the visible light. So the green blue and the red colors that we see. And it's also true for the invisible radiation that our human eyes are not equipped to see. And what happens then is as soon as this radiation bounces off the surface of the earth, then it is just read and received by instruments which are mounted on satellite images, uh satellites and the satellites are orbiting earth all the time.

Uh This is the data that we work with. And it's important for us to know where this data comes from. Oh And now we know this is radiation that is being reflected of surfaces of on the ground and the surfaces can be rooftops of houses. It can be, they can be the ocean, they can be glaciers or ice caps, they can be uh foliage of, of trees and vegetation. They can be water bodies, they can be uh the rose as well. So anything that is on the earth is reflecting back some kind of radiation into space. And then if we want to zoom these a little bit more, uh this is where we come to my case study is. So what I'm uh what I have been exploring for the last two years is um understanding what's the state of the vegetation on earth by analyzing satellite images. So what happens if it is if we zoom in a little bit more? So we see only one tree instead of the whole earth from space and then we see each leaf of the tree. And what happens is every single leaf is going to reflect back the radiation that is um emitted by the sun. And in the specific cases of leaves, they are green because they absorb, they have a lot of chlorophyll, which is the substance that allows vegetation to um absorb nutrients or to transform nutrients from the sunlight to into nutrients for the tree so that it can grow.

So what happens in in there is um speaking, in terms of uh chemistry is chlorophyll is the substance that reflects mainly the red lights that we see and the invisible near infrared radiation. So if we are smart, smart about it, what we'll do is we're going to take one satellite image which is taken from earth. And if we want to analyze only the state of vegetation, we are going to separate it into different sub images. And we're going to take only the image which has the red um the red light values in it. And then we're going to take only the image that has a near red, a near infrared radiation values in it. And from there on, we just plug them into a simple formula and we have a vegetation index which is what is telling us how healthy is the vegetation in this specific spot on the earth. And if we're speaking about spots, we usually speak about um 30 by 30 m per pixel or 10 by 10 m per pixel. Or if we are speaking about commercial companies, then we are saying 30 by 30 centimeters per pixel. But basically what we can do is we can separate every single thing on earth into pixels with different resolution. And then for each of these pixels, we can know how healthy and dense is the vegetation at this current moment there.

And these satellite images are taken every one day or every five days. So we can have a constantly updating picture of the state of vegetation on earth. So it is important to know this so that we know how we can apply this knowledge to solving an environmental problem. Um I have been living in Portugal for the last six, yeah, six years of my life. And in 2017, when I had this meeting at the British Interplanetary Society, we also had the worst wildfire season that Portugal had seen ever. At some point during the summer, we were bombarded with the news articles. Speaking about the devastation that wildfires were being done and how we related this to the science that I just explained to you is what fires burn where there is vegetation. So if there are some roads, they just jump over the roads where there uh water bodies, fires don't burn, water doesn't burn where there are some houses. They, the, the the fuel that is inside of the houses is burned, but then the house is left as a skeleton. But when there is vegetation, almost everything burns and vegetation is the fuel for wildfires. So in our case, we wanted to see, hey, really want to solve these environmental problem. Wildfires are a problem because they are killing a lot of trees. And the trees currently are the best mechanism that earth has to absorb CO2 and release the pressure for climate change that we are facing and we're going to face in the next 10 years.

So we started thinking of smart ways to use satellite images to help um the problem of wildfires that we have nowadays. And as you know, last year Australia burned and, and Brasilia burned and now Siberia is again on fire. So it's been, it's a recurring issue every single year that we have. We also have uh recurring satellite images being taken of earth. So what we can do is we can combine this intelligence of what is the state of vegetation and help mitigate wildfires. What do I mean by mitigate wildfires? This is to so the best way to fight a wildfire is actually to be prepared to prepare everything before the wildfire even starts. This is called mitigation and not wildfire combat wildfire combat is when the fire has started and the firefighters are already already fighting it and, and uh um um helping people run away from it. So in wildfire mitigation, what we do is we help prepare the natural world so that when fires occur and this is their natural disaster. So it's normal that we have such occurrences. But when this happens, they are much faster, uh much easier to control and firefighters can extinguish them faster. So this is the goal of wildfire mitigation.

And we discovered that satellite images help us a lot in this because we can help the government and the businesses who are already trying to uh minimize risk and mitigate wildfires do this in a more efficient way. Why we choose satellite images because they are bit um earth constantly. There is zero cost for us. So we don't need to invent anything. We don't need to learn how to build rockets. We don't need to learn how to launch them. We don't need to learn how to operate, create satellites and make sure they don't collide with any space um litter that is uh orbiting earth as well. What we need to learn is how to analyze satellite images so that we can use them for our business. And in this case is for us to use them how to help governments and businesses mitigate wildfire risk. So this is what we are doing. However, satellite images can give you a lot more information. So here we just wait and went into the specific case study, which is how do we know what's the state of vegetation? But we can use wildfires to analyze our oceans. We can analyze plastics both uh floating in the oceans and in rivers and being uh collected in land, mine, land, land, honeycode, hm.

Wherever plastic um is collected on on earth, we can also analyze um glaciers and I cap and ice caps, which helps us understand how fast global warming is happening and we can, we can do a lot of things. And the good thing is that a lot of people are still launching satellites. So they, the instruments that we get on satellites are with higher resolution and have more capabilities so that we can use this for our uh purposes. So this talk was just landfill. Thank you very much. So this talk was for me to tell you very quickly on how we have built our business on technology, which is data science and in particular, analyzing satellite images, which also makes us space tech and how we use this kind of technology to help um a pressing problem that we have globally, which is climate change and mitigating the the the the risks of it and hopefully trying to avoid it in the near future.

Um I still uh we still have five minutes but I think it's easier if we do um a little bit of question answering here instead of me continuing to talk because there is still a lot that I can, I can speak about. But I would like you to guide me in which direction should we go? So let me see. Um OK, so we have two questions for now that I can read and thank you very much for this. So Pra Prachi or Prachi, I'm sorry if I pronounce it incorrectly, what other applications do you focus on? At the current moment? We are focusing only on uh wildfire mitigation. Why be why? Because this is a massive problem. And it's also, it's, it's interesting both from a technical point of view and from a business point of view. So from a technical point of view, it's interesting because we have um a lot of challenges that we need to face. For example, if we have um a lot of clouds and sometimes we do like uh this year was uh the spring was a little bit rainy. We, we don't rely on uh cloud masks to remove the clouds because we don't believe that this um technology is very good at the moment. So what happens is if we have some cloudy days, we discard the images we need to handle, we need to deal with this.

We also need to do a lot of data engineering so that we can pipeline everything um to run automatically and to run correctly from a business point of view. It's interesting because even though everyone's still speaking about um how not still, they just started speaking about how businesses should um start uh focusing on climate change and how in general governments and businesses must work towards creating something which is both a financial value and impact value.

It's still more of a uh talk than it's it been happening. And at the moment, we are developing a new type of business model which is B to B two G or so, we are selling to businesses but our customer or our final user or consumer are governments, which uh so we do this because we believe that if we try to mitigate wildfires and work only with businesses.

We are excluding half of the lands which are uh shepherded by um local governments and governments. If we concentrate only working with the government, then we exclude the other part of the equation. And we cannot, we cannot do this alone. We need to work as a whole. OK. So let me go into the next questions. Question, which is does the smoke from the fires get in the way of imaging? This is a really good question. Uh Yeah. Yes, it does. Uh Sometimes it's not that big of a problem because as soon as we already have a smoke, we this means that the fire has started and it's very difficult to do any preventive measures at this point. What we do is we work before or our um analysis is used for wildfire mitigation preparations that businesses and governments need to do every spring before the wildfire season happens. Uh And in this case, the wildfires have still have still not uh started that if they have not started, this means that we do not have any smoke to um clouds. The images. The problem is again, uh clouds and sometimes um rain and a lot of humidity. But there are ways in which you can cancel out the effects of humidity on both the ground and in the atmosphere so that you can get correct reading of what's happening on the ground in terms of vegetation. Thanks Jennifer for this question. Um ok.

So Carmin, um another question from you, do you anticipate where wildfires will occur based on vegetation then? Ok. So two questions. This was the first one and then the second one is, what are your next steps in combating climate change? Wow, thank you very much. Beautiful questions.

So, do you anticipate where waters will occur based on vegetation? Yes, we did this, but we don't trust this very much as governments don't trust it. The problem with um predicting what where wildfires would happen is the following. There are a lot of factors that we cannot control that play role in where the fire starts. For example, 80%. Hm. I think it's 90 90% of the fires are human made. Sometimes people are just smoking um in the field near their camp or van and they don't put out their cigarette butt and the fires start. Sometimes they, they are barbecuing. Sometimes people burn things on uh purpose be based on different reasons. So we cannot predict the human element of starting fires in terms of the last 10% when fires happen naturally. It's this is very dependent on the um winds and on the humidity, both of the soil and the atmosphere. So, so humidity of atmosphere, we can get from weather predictions and weather forecasts. Humidity of the soil is another index that can be calculated by analyzing satellite images. But winds are very difficult. Even if you predict winds, they they, they just have a life of their own.

It's difficult to build um, systems which, which uh model where the winds would go next. It depends a lot on the geography. For example, in Colombia, in Colombia, you have stable winds. So it's expected where the wind would blow from. So, you know, in which direction the fire I go, if it starts in Portugal winds change direction very often, sometimes 60 times a day. And in this case, it's just um it's, it's, it's crazy to be on the ground and it's crazy to try to play God and, and try even to predict where the next direction of the wind is going to be. I hope this answered the, your first question and I'm going to, so I don't know when this is going to be cut off. We have had only 20 minutes and now we, we're talking for 23 minutes, but I'll keep on reading your um questions and we'll see when uh whether we can get all the answers done. So the um Carmen, I'm going to answer Teodora because she has a, a question and then I'll come back to your second question. So uh Theodora, from which platforms do you get your satellite images? Are they easy to procure? Yes. Very good, very good um question.

So, a couple of years ago, 10 years ago, you would have one satellite which, which is called LSAT, which is uh a NASA satellite that provided the whole world with free images. Um The resolution of it is 30 m by 30 m per pixel and it's still operational. So you can uh go on nasa's web page or on Lard's web page and get the data from there. We usually rely a lot on data from Sentinel, which is a constellation of uh satellites by the European Space Agency or ESA their resolution varies depending on what you try to measure. So it goes from 10 to 10 m a pixel to 60 to 60 m a pixel which is amazing because sometimes for air pollution, for example, before people worked with one kilometer by one kilometer, which is very, very um high level, you also have commercial players which can deliver very, very high resolution images to you something between the lines of 1 m per 1 m or 30 centimeters by 30 centimeters.

But in this case, you need to pay for this. And if you do uh satellite image analysis at scale, it can get a little bit expensive, expensive even for big businesses and governments. So the before this can happen at large scale, everything of the application and the service needs to be um perfected before a lot of money can be poured into this automatic processing of uh commercially available images. OK. Um The last question of Carmin is, what are your next uh what are your next steps for combating climate change? Um It is a good question. Uh We, when we started this company, we wanted to address the problem with plastic um in, in the ocean and in general on earth, you can still manage this and monitor this plastic. Um let's say pits from outer space. Oh, sorry, not from outer space, from out of our orbit. But um there are a lot of developments that that are happening. We've been working on wildfire mitigation for the last two years. And during these two years, a lot of new players have entered trying to solve the problem with plastic.

They're both using drones, they're using their own machines that they have built, they're, they're using satellites or some people are building their own satellites, launching them into space so that they can get the data that they need. Um So in terms of what is the next thing on climate change? We don't know, things move very quickly and I'm really happy about this because we really need to address this as people and businesses and governments. So, ladies and gentlemen, I think it's time for me to stop talking because our time is um finished. But thank you very much for taking the time to um to listen to me today. I hope that I managed to inspire you. So just to summarize to be, to do something amazing that you're proud of. You both need to have to be a good place in, at a good place for your profession. And you also need to be motivated on a personal level. So if you, if you want to go and explore something as a person, go do it, there is no stopping you from learning new things and applying them in practice. And from a point of view of the business, if you want to move into changing the world, somehow, the way that you feel that the world needs to be, go for it.

Again, if you don't start your own business, go find the businesses which are doing this and join them. I'm sure that they will be happy to have you. Thank you very much. I will give you my um linkedin profile. Let me just um share it with you. Actually, you're still looking at my um at my beautiful screen. So you're going to see my linkedin very, very briefly. So this is me and I wanted to share one more link with you, which is this one which is the uh platform from the European Space Agency where you can view and download satellite images from. So you need to uh register and log in as a registered user. And from there on you can explore satellite images by using their search um telling them which between which date you want satellite images uh from and what kind of satellite images you are interested in. And so again, thanks a lot. Go be awesome and change the world and let's go and see what the other speakers and participants in the conference are doing. Thank you.