Shravan Kumar: 0:03

Policymakers need to understand AI can actually help them rather than be detrimental in their pursuit of climate change. So that I think is the biggest factor, the intent and the strategy from the policy makers is the biggest by far factor in deciding whether these kind of programs are successful or not.

Tom Raftery: 0:23

Good morning, good afternoon, or good evening, wherever you are in the world. This is the Climate Confident podcast, the number one podcast showcasing best practices in climate emission reductions and removals. And I'm your host, Tom Raftery. Don't forget to click follow on this podcast in your podcast app of choice to be sure you don't miss any episodes. Hi, everyone. Welcome to episode 184 of the climate confident podcast. My name is Tom Raftery. And before we kick off today's show, I want to take a moment to express my gratitude to all of this podcast's amazing supporters. Your support has been instrumental in keeping this podcast going, and I'm really grateful for each and every one of you. If you're not already supporter, I'd like to encourage you to consider joining our community of like-minded individuals who are passionate about climate. Supporting the podcast is easy and affordable, with option starting as low as just three Euros or dollars. That's less than the cost of a cup of coffee and your support would make a huge difference in keeping this show going strong. To become a supporter you simply click on the support link in the show notes of this. Or any episode? Or visit Tiny url.com/climate pod. In today's episode. I'm going to be talking to Shravan Kumar from Gramener. And we're going to be talking about using AI to identify urban heat islands and make recommendations for their mitigation. And in the coming weeks, I've some really cool episodes coming up. Next week, for example, I'll be talking to Dr. Jay Hakes. He is a political and climate historian based in the US and we're going to be talking about next November's US presidential election. And the outcomes, the potential outcomes of that for climate. So that's a fascinating episode to watch out for. The week after I'll be talking to Dev and Xavi from Mittilabs, and we're going to be talking about rice and how rice has a huge environmental impact and how to mitigate that. The week after I'll be talking to Carina Gormley, and we are talking about local government, government and policy and how they impact climate. And the week after that, I'll be talking to Michael Barnard and we're talking CCUS carbon capture utilization and storage and the benefits or otherwise of doing that. So that's the next four weeks. Back to today's, as I said, my guest on the show today is Shravan Kumar from Gramener, Shravan. Welcome to the podcast. Would you like to introduce yourself?

Shravan Kumar: 2:57

Thank you, Tom. First of all, thanks for having me on the podcast. I am Shravan Kumar. I am director of client success at a company called Gramener. We are a data and AI company. I've spent the last 15 years of my career. Just after my graduation, helping clients solve the most pressing environmental, societal, and economic problems with the data, and now with AI. So I've been in the center of the data and AI revolution, if you were to call it that all my career. And I still help clients with the best solutions, that they can get from the either their own data or the data that is out there. This includes we also solve a lot of problems on climate change. And I'm also proud to announce that we have, you know, in some form touched 12 out of the 17 Sustainable Development Goals set for by the UN for governments and enterprises worldwide. So we've through our data and AI solutions, we're proud to say that you've had a huge impact on the AI for good in general, that area, and AI for earth as well. And I'm so happy to be here. And sharing my thoughts on how these technologies can help solve the most pressing problem that we as a species face. Again, thank you, Tom, for having me.

Tom Raftery: 4:29

So, Shravan, can you give me an example of how you're using AI to fight climate change?

Shravan Kumar: 4:36

Yes, certainly. So one of the things that we have done is for any AI, we start with the knowledge that for any AI to work, it needs to feed on a lot of data. Ultimately, AI is smart way of doing things with some kind of historical knowledge, and the historical knowledge is you know, typically comes in the form of a lot of data about something.

Tom Raftery: 5:04

Sure.

Shravan Kumar: 5:04

when it comes to climate change you need data about our own earth, our own climate, our own surroundings, environment, et cetera. The way how we, you know, collect data about earth. It's through Earth observation. I mean, many of us are aware of what Earth observation is, but still for the benefit of the listeners, let me state here that Earth observation is nothing but a bunch of satellite images or drone images that are collected through some means with the intent of collecting the characteristics and data about how Earth behaves in general. It can be at a city level or it can be at a planet level.

Tom Raftery: 5:52

Mm hmm.

Shravan Kumar: 5:53

So, to make this AI work, AI for climate change or AI for Earth work, we have to be able to collect a lot of historical data of how Earth has been behaving. So, that's the first step we've taken, to collect a lot of data and gain that overall knowledge. And apart from Earth observation, you have a lot of other data from the weather forecast, the wind speeds, the even water, river speeds, etc. Right? River streams and how our water bodies are behaving, so on and so forth. So you end up collecting all of this data for the AI to behave. You train the AI with the historical, historical behavior. Now you ask the AI to predict what could happen, not only in the future, but today, if a flood would happen or a water body would overflow, what would then happen to the landscape in that particular part of the Earth?. So, so we use, we are in the business of collecting all of this data, processing it, making it smart enough to understand the patterns that could emerge for different problems like urban heat islands, or floods, or earthquakes, so on and so forth. Why AI? You may ask. There's a lot of, you know, body of work, you know, organizations world over are trying to come up with risks, risk metrics, or how AI can help, right, in general, how this technology can help solve problems. We have taken the route of, you know, understanding whatever has happened in history, we are in the practice of understanding from weather forecasts or meteorological departments, whether it is national, federal, whatever it is, you have these inputs and they say the entire city is going to see some precipitation, some rain, some flooding

Tom Raftery: 7:53

mm.

Shravan Kumar: 7:53

but that blanket statement does not help in actually routing the resources to the most vulnerable parts of the city. or even the state or, or, you know, a nation. With AI, we know that, you know, climate change is happening. With AI, what we are doing is trying to route the right resources at the right time at the specific places. We are answering the where question of it, the why, what, how we more or less know from our history. But with AI, we are able to tell where should the rescue efforts go, where should the government intervene, where should the citizens actually, you know, act in planting more greenery or having a tree canopy, so on and so forth. So, AI has been used in our context. We are able to answer the where question of climate change. By means of that, we are telling, where are the people affected, where should you put a, you know, an intervention, a cool roof, or a green roof, so on and so forth, when it comes to urban heat islands. So, to answer your question, AI has been used to answer the pertinent question of where is, you know, something happening and what can we do about it? Those are the kind of questions. At a micro level, instead of a huge province or a city level, we are telling it at a micro climate or a block or a neighborhood level. We're able to say where it is happening and what can we do about it.

Tom Raftery: 9:32

Okay. For people who might not be aware, can you just tell us what urban heat islands are and why they pose such a significant risk to urban areas, particularly now, as we're in the middle of summer, it's, you know, middle of July here. It's, it's scheduled to be 41 degrees here in Seville today. So, yeah, I, I have a feeling I'm going to be experiencing an urban heat island later on today.

Shravan Kumar: 11:12

Absolutely, yes so, to define urban heat islands let’s first you know acknowledge the fact that heat waves in general are a manifestation of climate change. Glaciers are melting. We are warming the climate, you know the planet by use of fossil fuels and all the urbanisation that is, you know, taking place. So the last 50, 60 years, we've been warming the climate more than you know, any time in the history. But to the common person, how does a heat wave manifest with, with almost 60 to 70 percent of population to be in urban places by 2030 of the entire global population, right, to be in urban places, the way they will experience heat waves is through urban heat islands in a very, you know, simple way, but what is urban heat island? It is actually the phenomenon where highly built up areas like the downtown in a city or, you know, places where there's a lot of industrial establishments or commercial high streets, et cetera, experience significantly higher amount of heat because of a lot of various reasons, we'll come to the various reasons than areas that are at the periphery of the city. So, if you go to a residential place or people you know, places where the population is slightly lower, there you will experience you know, temperatures that are significantly lower than the downtown, for example. The difference between the thermal comfort that is experienced by people is actually the urban heat island. Where you have heat islands that are formed in specific places because of a lot of reasons, whether it's built up or industry, emissions, etc. You end up experiencing a lot of heat because of a lot of heat being trapped in the environment than the periphery of a city. That is how urban heat islands are defined.

Tom Raftery: 13:21

I, I got to think a lot of it has to do with just the proliferation of concrete and lack of vegetation. Would that be a fair comment?

Shravan Kumar: 13:30

Absolutely. This is the why part of it. Why is, you know, so while you know, doing anything for AI for climate change on AI for climate change or AI for urban heat islands we worked on a project called AI for resilient cities. And this is this knowledge is from there. So whatever you said is actually, you know, bang on. The, highly built up areas naturally have a lot of tendency to trap a lot of heat that comes from the sun. Even the road and the tarmac they have high absorption power than the greenery you know, elsewhere. So essentially it is the lack of vegetation and in the place of vegetation, we have built up a lot of you know, concrete or tarmac or even, you know, anything that does not happen to you know, balance the heat that is otherwise trapped in the environment or by the ground, right? Anything that doesn't you know, reflect the light or the heat that comes from the sun is likely to cause urban heat islands. And that is the result why we, when we come from our homes, by the way, I live in India I come from my home. I experienced a significant amount of heat, higher heat near the areas that are near to my office. And when I go to the, you know, place where I live, I have significant, vegetation, trees, and, you know, a lot of water bodies as well, that cool off the microclimate there. And I feel much more thermally comfortable. So that's you know, a classic urban heat island for you.

Tom Raftery: 15:15

Sure, and just tell me a little bit about the idea of using AI and geospatial data, for example. Is is the idea that you're using the AI and the geospatial data to identify where there's a danger of an urban heat island actually happening or what is it we're doing with the data? Is it just identification or is it some way to alleviate the issue as well?

Shravan Kumar: 15:42

As part of you know, our endeavor to use AI to mitigate and adapt against urban heat islands, we are specialized in answering three or four questions. And the questions are as following. One is, like I said, the where, where is the urban heat island? Where are the islands of heat present in a city, in an urban environment? That's the first question. Then we answer the what. What is causing the urban heat island effect? Right? Like you said, you were bang on saying that lack of vegetation and hence replacement of the vegetation with impervious surfaces like buildings, tarmac, etc. Can cause the urban heat island. So it's really the causality of what is causing the urban heat island. The third question we try to answer with AI is who is getting impacted due to urban heat island.

Tom Raftery: 16:40

Right.

Shravan Kumar: 16:40

So, stats say that in the Northern America, that part of the world, a lot of people that are vulnerable to heat waves, which is by the way, the you know, worst number one climate killer compared to floods or any other manifestations of climate change. Heatwave kills more people than any other climate struggle stress. But coming back to the point in North America, we've seen a lot of indigenous folks and also people of color and you know, a few ethnicities like, you know, black and a few other ethnicities are highly vulnerable to heat waves. So we try to answer through AI, through again, some geospatial data on collecting intelligence about which parts of the city have high concentration of a particular ethnicity and thereby telling who can be impacted. We also try to understand what is the concentration of vulnerable population by age infants and older people older than 65 are particularly vulnerable. And this status from right from New York City, there are 1100 hospitalizations due to heat stress. And a hundred people die every year. And that's a stat from last year. On record that was the hottest year in history. So, so we, we try to understand these patterns from different, you know, data sources, whether it's government sponsored or privately acquired, we understand who is impacted. And the final question we answered through AI is what if, what if I can put up a tree canopy in a place where vulnerable population is likely to be located. I put the AI through some data points saying I increase the vegetation by 30 percent or 50%. What would the urban heat island look like? Or what would the temperatures in that micro you know, climate at that particular block or neighborhood look like? So I can assign targets to the people that are responsible to the mi for the microclimate there to be able to take that target and say, plant 1 million trees in the vicinity in the next one year or so. Those are the kind of interventions with AI we are able to make. And those are the questions we are able to answer to AI

Tom Raftery: 19:16

And who are you doing this for? As in it's all very well if I contact you and say, Shravan, what can I do about the urban heat island around here? But I don't have the power to plant a million trees around here. So are you in contact with city chief heat officers, for example? Or city administration? Or who is it that is taking this data and these insights and acting on them?

Shravan Kumar: 19:47

Well the case in point that I was talking about is the municipal government, provincial governments and the federal governments, right? So we work indirectly with them through a few non profits who are in the business of you know, building better places and climate resilient places for the future. So we work with a non profit with a bunch of governments who then have a huge network of a bunch of municipal, federal, provincial governments. The end user of these AI tools is the municipal governments, but there is a specific persona that uses this. Like you said, it is the Chief Climate Officer who is, or their office, who are responsible for making sure the thermal comfort is intact. So, for example, there was a municipal government that was mandated to build thermally comfortable playgrounds for children. That was one of the, initiatives the municipality took. When they wanted to use this tool, they came across all the urban heat islands across the city. They said, oh, this is a revolution. We need to do something about it. And hence, they took up not only that initiative, they expanded it to saying we need to ensure thermal comfort even in commercial streets. The downtown. We need to put up a policy of white roofs or green roofing, across the city. So these are the kind of personas that we build these AI tools for, but having said that these AI tools needn't be just restricted to the governments. Like I initially mentioned we had also developed a citizen tool to be able to interact with the municipal government by You know, making a portal, a mobile app where they could log in and you know, click a picture or take the temperature, record it from a particular position from a particular place register that lat long, which would then geofence their location and say, here, here, the temperature is 46. We have to do something about it there. Right? So, it was a 2 way communication tool that we created where the citizen could register their grievances or the situation to create that situational awareness for the government, but also in turn, he or she would get the awareness of the programs the municipality was running. So, for example, the thermally comfortable playgrounds or how, by way of policymaking, like a green roofing policy, has brought down the temperature in the near, medium, and long term. They could then consume that information and be appraised of all the initiatives that are being taken by the government. So AI, like someone wise said, AI will be like electricity. So we wanted to make it so democratized that, we not only provide decision making tools and policy making tools for the municipality, but also consumption tools at the citizen end and to be able to create, take that data and make it their own. So it works both ways. So I would say primary, the primary persona is the municipality. Secondary persona is definitely the citizen.

Tom Raftery: 23:24

And is that just in one particular municipality, or is that global? And when I'm, when I'm asking that, I'm talking about the access for, for citizens and a secondary question to go along with that, if it is available to citizens globally, or even if it's just one municipality, how can you verify the accuracy of the data that the citizens are putting up? And if I'm asking that, it's because I know when I'm driving around here in Seville, for example, and I see some of those those kind of clocks that are at the side of the road telling you the time of day and the temperature and because they're out in the heat of the sun all day the temperature they're declaring that it is is usually about five degrees above the actual temperature because they've just been stuck in the sun all day so they're telling you it's 46 where it's probably 41. So, you know, how, how does all that work?

Shravan Kumar: 24:26

Right. That's a great question to actually in our world, in the AI world, we typically call it the ground truthing. So someone has to actually make sure the crowdsource information for you know, for one is accurate. And the second is how do you actually ground truth? Go there and see if you know, the reported numbers are actually, even near the reality. It could be this one thing for it to be off by five degrees centigrade, and it's completely another thing to actually report 46 degrees centigrade in the middle of a winter season, right? So, how do you kind of, look at the anomalies? Again to an extent AI comes to the fore here you have these huge, you know, set of crowdsource information or data which comes in. And you know what the usual trends are, right? If something is really, really off, like I said, 46 in the middle of December is not possible. So the AI or the algorithms are automatically looking for such anomalies and removing them automatically. That's one level of filtering at you know, the data level, but how can we do this? How can we contain the accuracy at the source? For every crowd sourced information that is entered by a citizen, we have a check, we have, you know, a checklist for them to go through. One of them is uploading an image of the context. So if they can upload an image of the thermometer reading or however they took the temperature that could be acting as first level of check, right? If that does not exist, we automatically, the algorithm automatically deprioritizes that reading.

Tom Raftery: 26:20

Okay.

Shravan Kumar: 26:21

And it also increases, increases people. We also have a leaderboard, some kind of a fun thing where you, for all the accurate readings that you do, your credentials are built up in the in the, in the leaderboard. And for example, I could say Tom has contributed to almost 20 readings in the last 10 days, and all of them are accurate. So he comes to the top of the leaderboard and hence he's incentivized to do that more often. Instead of someone who, you know, that he goes to 10 places, we have had citizens who visited the vicinity of a mall or a building and have reported 20 readings. That is an anomalous behavior, right? And we, of course, the next reading that comes from that citizen is deprioritized to an extent. So it works in, in almost similar ways. As a Google map entry by a user, right? So we follow almost the same principles, but the whole point of this is not to rely hugely on the accuracy because we already have some means, by the way, through earth observation, we are able to get temperatures up to the 30 meter level or even in some case 10 meter level, which is almost a building a residential building a small, you know house, right? So at that level if we have the temperature we can always correlate it to the nearby readings and have that as a you know, tool to filter. So that is, those are a few things that you know, we deploy when we take a crowdsource readings, but overall it is about interaction with the citizens in a very healthy way, rather than it being a source of data in itself. Right. And like I said, it's a community building interaction exercise.

Tom Raftery: 28:14

And is it, as I asked as well, is it global, or is it just one or two municipalities, or how does that work?

Shravan Kumar: 28:21

So the AI for Resilient Cities is right now scaled to almost half a dozen cities in the North American geography. And these are cities that are almost 2,500 square kilometer in, in expanse, right? Having said that, when it comes to Earth observation, or geospatial AI in general, this has the capability of scaling to province, an entire nation, or even the entire world if you keep the, you know, lowest common denominator across all of these as common for example, you can say, I want to collect 100 meter resolution data for the entire Earth. That's already available. That's already there. You know, NASA collect has collected for the last eight years. And there are federal, you know, satellite space programs that do this collection every now and then every 15 days. So it's democratized already. So with that kind of data and this kind of AI, you can put up a planet scale adaptation and mitigation process. But if you're really looking at precise locations, then you need to look at specific data sources, but the capability is always there

Tom Raftery: 29:42

Okay, but the, the, so for, for, Right now, the AI for resilient cities is half a dozen cities in North America. What, what's stopping it from going beyond those to, let's say, every municipality of more than a million people globally?

Shravan Kumar: 30:01

Right? To your point, there's practically nothing. If you want to look at it from a feasibility standpoint. There's absolutely nothing that stops it from happening. In fact, there is a program called 100 Resilient Cities. You will have heard about it. Most of us in the urban planning or smart cities world have heard about it. It is a program funded by Rockefeller Foundation and also backed by a lot of other non profits across the world. They've shortlisted 100 cities across the world and identified programs for each of the cities with the local context in picture. So. If you were to apply this to all the cities in the world, it's absolutely possible, but practically what happens on the ground is all the municipalities, all the provinces, all the nations should have an intent to actually solve it using AI. Sometimes it, it is lost in the, in translation, in discussions. A lot of cities have the intent, have the means they go for it, like the North American context I spoke. Some countries do not have the means to actually you know, solve these problems. They have bigger problems to solve. For example, Africa has a hunger problem, right? Oh, you know, some countries in Africa do have a hunger problem, which is much more important to be solved. Again, using AI or not using AI is a different thing. They cannot afford to put resources behind such problems. That is one thing from a policymaking standpoint that stops. Some countries and you know, states or provinces, they do have the intent, but they do don't have the you know, way or knowledge of, you know, AI can solve these problems. The knowledge that AI could solve these problems does not exist. So it takes a lot of education also. And building trust because for all the advent of AI, we've also faced concerns on privacy, data sovereignty, and where does the data reside, etc. So there's always that inertia of working with a third party that probably belongs to another country, working with them, whether the data from that particular country goes out of it, out of the country, so on and so forth. So there are a lot of data concerns and AI itself being target of a lot of flack. There is that inertia, so we need to get to a point where people need to be educated. Policymakers need to understand AI can actually help them rather than be detrimental in their pursuit of climate change. So that I think is the biggest factor, the intent and the strategy from the policy makers is the biggest by far factor in deciding whether these kind of programs are successful or not.

Tom Raftery: 32:52

Okay, great. And you mentioned initially as well that this can be used not just for urban heat islands, if I, if I heard you correctly, but also the likes of floods and droughts and I suppose, glacial melt and things like that. Is that the case as well?

Shravan Kumar: 33:10

Yes, absolutely. Again, I'll go back to the basics that we started off with. We are ultimately training the AI to understand in this context, the temperatures that are observed from the Earth observation data. The Earth observation data also tells you the change that has occurred in the last 10 years, 20 years, and so on and so forth. This applies to flood, water bodies being, you know, flooded, water bodies increasing there. You know, for example, there's a river it's, it's normal course versus it's course when it is flooded, right? Those two can be seen when a flood happens and you can actually mitigate, use the same techniques like how much altitude is there near the river to say which of the areas at the micro level, are going to be submerged should a flood happen. Again, the why, how, what, what if can be answered using the same set of data. It just needs the AI to understand it in a different context. It can be applied to landslides. If the land has moved from, you know, a landmass has moved from one place to another place, even at a centimeter level, that can be detected. Landslides can be detected. Earthquakes can be detected by studying the tremors you know, at the ocean bed. Wind speeds can be studied when, you know, wind speed picks up and concentrates in one particular part of the ocean. You can, that's how typically they predict Hurricanes, cyclones, etc. Using the same techniques, you can embed that with the people data, where they are, and answer where, what, and what if questions using the same basics. So, again, going back to the basics, I just wanted to let you know that it is just the, the basics remains the same. It is just the techniques that change in order to inform decision making for different climate shocks and stresses.

Tom Raftery: 35:44

Interesting. Interesting. And what's metrics or KPIs do you focus on to measure the success of the initiative and how can you ensure continuous improvement there?

Shravan Kumar: 36:00

That's a great and pertinent question. We get asked about it all the time. And this also speaks to why AI, you know, in a way, right? The first of all it's not, you know, weather forecasting in general or climate change forecast is not new to us as humankind. We get weather forecasts from meteorological departments and you know, government agencies all the time. But what has changed with AI is again, going back to something I mentioned is we are able to do this at a micro level at a neighborhood level at a building level, possibly. That has changed. Although climate change is a global problem, it needs to we all need to acknowledge that the solutions are at a local level. It manifests in a different ways in different parts of the world. So that is where AI comes in. Right. So going back to your question all of these can be you know, captured the value capture of such AI tools can be captured in the impact they bring in. And it depends on which angle you're looking at. If you're working with the governments you are tracking KPIs, like how many initiatives have been taken using this tool? How many plants were you know, planted, for example? How many green roofs were built? How many policies were made? Those are the kind of impact metrics using such tools you can report. And by the way, this is another thing that people have a lot of nausea around. When I say people, the municipalities. Municipalities, by the way, particularly in North America and, you know, all parts of the world are mandated to answer for every dollar they spent from the, taxpayers money. So if they're spending it on an AI tool, they need to actually be answerable to, for every penny they spend. So they're very conscious about saying we are using AI. So some of these programs happen to be operationalized through non profits. Non profits have their funds, and they go fund the initiative that the government is running. It's not always the government funds that are put in and if there is a government fund, it's it's routed through different strategic programs, smart innovation, smart city programs and such, right? So the impact metrics they should report is how many such initiatives have been taken and all that. More importantly, though, going back to the who, how many people were impacted? How many less hospitalizations due to, for example, urban heat or heat stress? How many less deaths? Those are the impact metrics the non profits in this case report. And that is what we also target. Of course, as an AI solution implementer, we are only restricted to creating the solution and handing it over to the authorities. The end users training them and making sure they adopt. We report adoption metrics, how many users use it, how many teams use it, et cetera, but whether they use it or not is ultimately their choice. But we have seen these kinds of metrics being reported in different reports. So using our AI tool, not directly attributing it to the AI tool, but we can correlate, you know, and connect the dots when, you know, metrics are reported. And we are proud to say, like I said, not only climate change, some of our AI solutions have also touched on eliminating hunger, eliminating poverty clean water and sanitation, life below land. We've used AI to detect species, fish species below water, to control rampant fishing, illegal fishing, et cetera. Anyways, coming back to this, that's the sense of metrics that are reported in this context.

Tom Raftery: 40:05

And do you have any success stories you can point to?

Shravan Kumar: 40:09

Oh, yes, there are many. So, I, I'll, I'll stick to the AI for Earth or AI for Resilient City program.

Tom Raftery: 40:17

Okay.

Shravan Kumar: 40:17

So we have of course, like I said, worked with at least six to eight cities in either having implemented a solution a proof of concept or a proof of value, or further. So at least, you know, six to 10 cities have taken these solutions and experimented in some way with it. Out of that, six cities are using these tools to actually make decisions. And one of the, you know, success stories is the thermally comfortable playgrounds and tree canopies implemented in a commercial high street in a city. Unfortunately, I can't name the city, but they have planted almost 80, 000 trees along the commercial street, which significantly allowed even the businesses to expand their footprint into the street and have seatings, right? Which could comfortably sit people etc. Another example from the UAE is they did something quite different. They used palm tree based shades at different, you know, places to create those pockets, cool pockets in the city. Where communities come and you know, interact, whether it's bus stops, whether it's other public places where thermal comfort, you know, UAE is, you know, 50 plus degrees centigrade at any given point in time, indoors, meaning air conditions

Tom Raftery: 41:50

Yeah.

Shravan Kumar: 41:50

Using these tools, they have been able to implement, you know, shades that are made of You know, some kind of an indigenous material with zero, you know, waste to create these cool pockets where people can come, interact and, you know, wait for a bus, so on and so forth. Right. So these are some success stories and to quantify this, it has impacted thus far around 100,000 people collectively at different. in different walks of life and at different sections of vulnerable population. So that's the people metrics, people impact. The other part of this is a lot of companies, a lot of enterprises are mandated to declare under the TCFD, what's called the Task Force for Climate Related Disclosures. They are mandated to declare what is the kind of risk they face, or their infrastructure or operations face come, let's say, 2050, 2 degrees scenario or 1.5 degrees scenario, right? While these mandates are in place, they can use these tools to quantify the risk they are under. A classic example again is energy trading. If a company has solar farms. They can use these tools to report metrics like, okay, the sun is bright and, you know, high up, you're going to get more proportion of solar energy this week, this day, right? And hence, they can adjust their rates, energy rates. At the user level. Similarly, when the wind speed is high or low, they can adjust the energy rates that come out of a wind farm and hence the proportion of renewable energy. Right? So in energy trading, this is this is becoming a really powerful tool to be able to adjust your energy trading systems and also declare risk at an enterprise level to climate shocks and stresses of tomorrow, right? So I would like to conclude by saying apart from the people metric there's a lot of business metrics that are related to these kind of tools AI for climate change.

Tom Raftery: 44:10

Sure. And what's next? I mean, where do you see this evolving to? What are the kind of next steps in the, in the roadmap? I'm sure you guys have for the product.

Shravan Kumar: 44:21

Oh, yes, so we first of all see this growing organically. First of all, the biggest metric of success of such tools is how many places this has been implemented in or operationalized in. So we want to take it in a similar form or slightly different form or customized form to as many continents as possible. That's the first, you know, and we're looking at, you know, partnerships that can enable us to do that. Like I said, it is not always straightforward. Going to a government selling it to them. We're looking for partners that are working in turn with governments to operationalize this. And the funds are not going to come from, you know, the resources, funds or people are not going to come from directly the government. So we are looking at enablers to take this to as many cities as possible in the world. That's one. The second is the accuracy of such systems right now stands at anywhere between 70 to 75%. And we are looking at microclimate, if you remember. So we want to be able to report these metrics or inform these decision making, you know, tools or methodologies in a very accurate way. We want to become as accurate as let's say higher nineties. So we hardly go wrong in our advisory or in our you know, recommendations.

Tom Raftery: 45:54

Sure.

Shravan Kumar: 45:55

So we are, we are working in that direction to make these systems as accurate as possible. Now the problem with this is the problem that the larger earth observation or geospatial industry has in general at large. The problem with that is a lot of commercial players are still grappling with profitability. In the presence of a lot of government space programs, privately, you know, held companies are facing a lot of problems in achieving, forget about thinking about profitability, even achieving scale, right? And to be able to become more accurate, we have to democratize data that is coming from private players, not government players. Because those private players give you high accuracy, high precision data. And for your model's AI to be more accurate, you need to leverage that data. And when they, the upstream is still grappling with profitability, we are going to have, we are downstream, right? We are creating analytics on top of the data. We are going to suffer from scalability as well. We are not going to take this to as many people as we want to, right? That is where the, I think, the inertia, the question between commercial and economic economic and societal viability needs to be concerned. There should be multiple players and geospatial industry is going to go through a lot of consolidation. This is my hard take and it is out there. If that happens, you know, downstream players like us are going to be folded into the upstream players. The upstream players are going to be able to create value only when they have the analytics downstream. So it calls for larger partnership with the players that can scale these program and that is what is next for us how to scale this. We know what to do but we are grappling with how to do this. A lot of commercial and you know scalability questions come on the way. So that's how we're looking at progressing this. Finding solutions and partnerships that can help us scale this.

Tom Raftery: 48:03

Okay. Left field question. If you could have any person from past or present, fictional or otherwise, to be the face of the AI for Resilient Cities, to be your kind of spokesperson, who would you choose?

Shravan Kumar: 48:24

Let me think about it. That's a very good question, by the way. No one has ever asked me that. So I think I would select someone elder, wise who has seen it all, right? And who has had the pinnacle of you know, whatever it is, it needs to whatever we need to achieve as a human, right? So let me come from that. I think probably someone like Steve jobs or or Morgan Freeman is, is the likely, you know, candidate for, at least as far as I'm concerned, someone wise, someone who has seen the history of this planet, seen our own evolution from radio transmitters to mobile phones, right? They've seen it all they are So

Tom Raftery: 49:11

a nelson Mandela or a Mahatma Gandhi or someone like that?

Shravan Kumar: 49:16

Mahatma gandhi not so much maybe but Nelson Mandela is is is you know, right bang on or even a Che Guevara because yeah had he lived to this day. I think he would have seen everything right from the hardships to the the opportunities that lie in front of us. So those are the people who can say how this is going to progress Into the future, right? So I think the Che Guevara would be the perfect ambassador for me as far as i'm concerned

Tom Raftery: 49:49

Fantastic. Fantastic. Great. Shravan, we're coming towards the end of the podcast now. Is there any question I haven't asked that you wish I had, or any aspect of this we haven't touched on that you think it's important for people to think about?

Shravan Kumar: 50:02

I think a lot of agencies, right a lot of work that has been done on the private sector or even you know, the acknowledging the role of the private sector, I think has to be covered in, in my views. People have to know what private sector folks are doing to, to mitigate climate change in general and urban heat as well. So, a big part of that is built area. We're talking about built up area of urban, you know, spaces. How do we make our urban spaces more climate resilient, more carbon zero or net carbon negative. How do we make our spaces that we live in and thrive in more climate resilient, right? So one of the things that is happening is it's a very simple technique, but yet a powerful one. A lot of architects are looking at putting up green walls. So I don't know if you've heard about green walls. Green walls is a technique of putting up a facade in front of a building with a vertical garden. What that does is decreases the temperature by a couple of degrees centigrade. And when I say decreases the temperature, the internal, you know, temperatures. What that does is reduces the dependency on ACs, or air conditions, or, you know, other factors. Which then, again, is a vicious circle, as you know, the CFCs, the, you know, gas emissions, et cetera. It reduces all of that. So even if all the buildings in India, you know, follow the principle of a green wall, we can reduce our overall temperature by almost five degrees centigrade. And, and, and today's architect firms, building designers, and even the people that provide the certifications to, you know, the LEED certifications, you would have heard about LEED silver, platinum, gold, the certification bodies, the governing bodies, green building councils, for example, have a huge role in bringing down urban heat island effect by putting in place policies by architects to be more aware of how to put you know, a greener and a more thermally comfortable building, not going for the over the top material for example, fabricating their raw material on site, what that would do is reduce the operational carbon and even embodied carbon of a building drastically down. That will bring that down drastically. And hence the private sector that is affiliated to building our spaces has a huge, huge role in decreasing the effect of urban heat islands. And this is this is imperative. This needs to happen now. So our future generations have a much more thermally comfortable planet to live in. And hence decreasing the overall urban heat island effect. And living in harmony, I should say. So I think that is a point of acknowledgement I would like to end with.

Tom Raftery: 53:16

Nice. Nice. Great. Yep. Shravan, if people would like to know more about yourself or any of the things we discussed in the podcast today, where would you have me direct them?

Shravan Kumar: 53:26

The best place is to come on to our website, Gramener. com. And come to the solutions page where we talk about all our geospatial analytics body of work and our ESG page. So Gramener. com is your, you know, point of okay entry. And the second is I could give my, you know, coordinates for people to, you know, get in touch with.

Tom Raftery: 53:57

Sure. Shoot me across anything that you want to have in the show notes whether it's your LinkedIn or whatever, and I'll put them in the show notes and that way everyone will have access to them.

Shravan Kumar: 54:08

Absolutely.

Tom Raftery: 54:09

Super. Great Shravan. Fantastic. That's been really interesting. Thanks a million for coming on the podcast today.

Shravan Kumar: 54:16

Right. Thank you for the opportunity. And I hope I didn't do too bad. Thank you so much.

Tom Raftery: 54:25

Okay, we've come to the end of the show. Thanks everyone for listening. If you'd like to know more about the Climate Confident podcast, feel free to drop me an email to tomraftery at outlook. com or message me on LinkedIn or Twitter. If you like the show, please don't forget to click follow on it in your podcast application of choice to get new episodes as soon as they're published. Also, please don't forget to rate and review the podcast. It really does help new people to find the show. Thanks. Catch you all next time.