Vladan Babovic
Description: Vladan Babovic is a professor in the Department of Civil and Environmental Engineering at the National University of Singapore. His work focuses on hydroinformatics, the development of water management through information technology and tools like AI and machine learning. In this episode we discuss the unique challenges that Singapore faces in regard to water risk and how the small nation state is a pioneering model for water issues globally. Babovic explains Singapore's flooding management systems, the concept of blue-green infrastructure, and the impacts of climate change on water concerns, stressing the urgency for inventive solutions amidst evolving environmental conditions.
Websites:
Publications:
Resources:
Show Notes:
[0:00:00] Introduction and Background
[0:02:37] Evolution of AI Paradigms
[0:06:47] Global Water Usage and Crisis
[0:11:42] Efforts for Water Autonomy in Singapore
[0:20:35] Managing Flooding Systems
[0:22:37] Urbanization and Water Management
[0:25:49] Replenishing Groundwater in Cities
[0:29:28] Rainfall in Singapore
[0:30:01] Innovative Solutions for Water Management
[0:33:04] Hydroinformatics as a Topic
[0:33:22] Mycelium and Drainage Innovation
[0:34:26] Evolution of Hydroinformatics Field
[0:34:58] Rainfall Monitoring Using CCTV Cameras
[0:37:01] Implementation of Rainfall Monitoring Algorithm
[0:37:40] Rainfall Monitoring in Jakarta
[0:40:39] Collaborative Innovation with National Water Agency
[0:41:36] Corporate Social Responsibility in Technology
[0:42:42] Addressing Water Management through Hydroinformatics
[0:45:27] Potential of Technology in Water Management
[0:46:59] Integration of AI in Data Processing
[0:48:08] Utilizing Rainfall Data for Various Purposes
[0:49:36] Players in Weather Forecasting Space
[0:53:28] National Water Quality Monitoring System
[0:55:41] Future of Water Management Technologies
[0:57:34] Societal Impact on Water Conservation
[0:58:36] Addressing Water Scarcity in Arid Regions
[1:00:59] Importance of Artificial Recharge in Water Management
Unedited AI Generated Transcript:
Introduction and Background
Brent:
[0:00] Welcome, Professor Vladan Babovic. Thank you for joining us today.
Vladan:
[0:05] Well, it's great to have you here. Welcome to rainy Singapore from sunny California.
It's just the start of a monsoon, so it will be wet until the end of your stay at NUS.
Brent:
[0:19] Yeah, looking forward to it.
Keller:
[0:22] We'd love to start off by hearing a little bit more about your story.
How'd you get into hydrology and how'd you end up at the National University of Singapore?
Vladan:
[0:29] Were well hydrology or water in general it has been passion for the better part of my life i, um i i thought i wanted to be an engineer but i do not really like an engineering youknow concrete and this sort of stuff so uh water was one of the topic which was always very interesting So it became a very natural choice of my studies, I think.
And then as the time went on, I had a chance to visit and study various parts of the world.
I actually have a degree from University in Delft in the Netherlands.
And yeah, it just grew and grew and grew.
And while in Delft, I met a certain professor, Michael Abbott was his name, who established a field called hydroinformatics, which was very new.
This was sort of late 80s or so, mid 80s.
And with his encouragement, I developed secondary interest in artificial intelligence.
So this is it it's when you start talking about this what happened in the past, it's it seems like an easy choice but it was for me very easy choice I just followed what I wanted to dowhen.
Brent:
[1:56] Did you start getting interested in AI.
Vladan:
[1:58] At that time the mid 80s um, My father, who was a medical doctor, always said, artificial intelligence?
Are you ever going to land a good job, my son?
And so I ended up doing my PhD in AI together with a visionary professor who was very supportive of all this.
And in a certain sense, maybe it was ahead of time, I don't know.
But it was, so I work in AI since, yeah, a good 1989, I think.
Evolution of AI Paradigms
Brent:
[2:38] What was the big difference between AI then and now? Because I think everyone just starts getting familiar with the term.
Vladan:
[2:44] Well, at that time, the big paradigm was something called expert systems, where everything we know at that time, the belief system was because everything we know can beexpressed in form of if-then-else rules, which is to some extent correct, but when we express ourselves in a symbolic way, it's only a very superficial part of our intelligence. you knowwhen you when you you don't drive a car saying if i press the accelerator then it reaches i don't know three thousand revolutions then i press a clutch then i shift again you can explainyourself in a form of if then else rules but you don't run those rules when you drive a car it at some stage it becomes what we call sub-symbolic below levels of symbols so that time,most of AI was very much symbolic.
We thought that whole knowledge of the world could be embodied in terms of if-then-else rules.
[4:05] Only later, maybe in mid 90s, new paradigms came like neural networks, which are still very much in use today.
Evolutionary computing, genetic algorithms, all sorts of other techniques took over and they are dominant today.
Today, what is special about AI today is It's very easy to use if you know a little bit of Python and you have these great libraries, Keras, PyTorch.
You can just write a simple script and suddenly you have this very powerful AI modeling environments at your disposal.
So this is very different. Hence, it's accessible.
Many people can use it and play with it.
Keller:
[4:58] When did you start to see other professionals and other professors start to adapt AI into their… Very recently.
Vladan:
[5:05] Very recently. Maybe last three, five years.
And I really think this has to do with the fact that it is a very accessible paradigm these days.
This was certainly not the the case.
But what we see, and this is one of the things that I really love, well, I'm very passionate about combining what we already know with AI.
You see, there was somebody by the name of Sir Isaac Newton, and he came up with Newton's laws, right? We still believe these laws are valid.
So we don't need AI to find Newton's laws for us. We know that.
And there are so many centuries of scientific advances that form a foundation to where we are today.
[6:04] There are so many things we don't understand, but we should not reinvent the wheel.
So I'm a deep believer of the fact that we have to find a way of combining what we know and then combine this with machine learning, with AI, to push the boundaries even further.
Ignoring centuries centuries of development of human knowledge is just unwise, you know, and we should just not do it.
But there are some people that fall in a trap, how is this easy to use and let me use AI and, you know, anyway, it's a long discussion and we can talk about this for hours.
Global Water Usage and Crisis
Brent:
[6:47] Yeah. Yeah. So then kind of jumping in more towards what your research is, could you maybe give us some background on the global water risk and like what the current state ofwater usage is in the world yeah.
Vladan:
[7:00] So let me let me ask you a question okay how many liters of water did you use today.
Brent:
[7:08] No idea what the exact number is but what do you think, i haven't showered yet so five liters five.
Vladan:
[7:18] Liters what about yourself i.
Keller:
[7:19] Seem like 20.
Vladan:
[7:20] 20 Well, the average person uses anything between 150 and 200 liters of water per day.
Brent:
[7:32] Directly?
Vladan:
[7:33] Well, you flush your toilet, you drink your water, you prepare your meal, you wash your clothes.
So every day it's around 200 liters. In some parts of the world, this can be very, very bad. and people water their loans.
It can go up to 600 liters per person per day. Wow. It's a lot of water, right?
And it's so convenient.
It's in everybody's home, and you've pressed a button and water runs.
[8:05] But how do we get this water, right? How does it come to you, right?
So there are two closely interlinked phenomena. First, you need a water that should be clean, unpolluted.
[8:18] Then you need a lot of energy to transport the water to your home.
You also need a lot of energy to purify, to make it clean, make it readily available.
But once you use the water, once you consume it, it goes somewhere.
And then before you return to the environment, it's very often treated so that it's not polluting the world as a bad place.
So there is this very tight connection between energy and water.
We call it water-energy nexus, in which water requires a lot of energy.
But even if you want to produce energy in any form, you need water to create steam, essentially, in it.
[9:04] So, water is much more than a natural resource.
It's much more than a natural resource. It is a matter of survival.
And so, the global water crisis, there is water, there is too much and too little.
If you remember in September, October was it? In a short space of 11 days, we had eight massive global flooding events from Hong Kong to Libya to New York City to Italy toBulgaria.
Too much water.
Then you come to the other parts of the world and there's too little water.
It's unevenly spread over the time and it's polluted.
[9:59] We can talk a lot about pollution, how pollution, you have eutrophication, all these phenomena that cause deterioration of water quality.
[10:16] But let's first talk about quantity. And nowhere else, one of the questions is, how come I stayed here? I came to Singapore for three months and I'm 20 years later still here.
I always thought Singapore will have expiry date for me.
It never came because this is one of, for water especially, this is one of the most exciting places in the world.
You know, Singapore is not water self-sufficient.
We import 40% of water from Malaysia, from province of Johor.
Hence the sense is that this water is a matter of sovereignty you know if you look at the second world war when the British army was in Singapore, beginning of second world war,they were expecting Japanese army to come from a sea but no they didn't come from a sea they come over Malayan peninsula and the only thing they did was just to close a tap, andtwo weeks later her british army did not have water to drink so they just surrendered, uh singapore today imports were still 40 percent of water from malaysia based on internationalcontracts uh that are stable and but this international contracts will go until 2060.
So just another, 35, 37 years.
Efforts for Water Autonomy in Singapore
Keller:
[11:42] Is there efforts in Singapore to try to build greater water autonomy?
Vladan:
[11:46] Yes, there are certainly efforts. So what we do in Singapore, we always say we have a diversified system of water supply.
Imported water is one very large dominant factor. factor in our water supply portfolio.
[12:12] But we have a system of reservoirs. So this is men, all lakes, all reservoirs in Singapore are man-made. There are 17 of them.
I came here to help develop reservoir number 15, which is Marina Reservoir.
And then with my my team, we develop lakes number 16 and 17, Punggol and Sanggol.
So this is the second source. The third source is what I call new water.
New water is used water.
So it's collected sewers that are treated to ultra clean level.
That's source number three. And source number four is desalination.
We are island, we are surrounded by sea, we have lots of seawater.
[13:04] But the problem really then comes to the energy. To desalinate, you need a lot of energy.
So that's that water energy in excess, right? So to produce clean water, you need a lot of energy.
It's very expensive water.
It is approximately five or six times more expensive than new water.
And new water is obviously much more expensive than rainfall that we collect and store in in reservoirs and then of course you have imported water so the, four sources gives youprovides you with diversified portfolio so the water situation is stable.
There are efforts to extend agreements with province of Johor in Malaysia, this could go on beyond 2060.
[14:06] But if you see the water demand, by 2060, our water demand will likely double, because of industry, because of population growth, because of everything else.
So obviously, these are the forces that are not necessarily aligned, makes things a little bit more complicated.
Right so that this is a problem of having potentially too little water so you have to really very consciously deal with it and in singapore this is pronounced the water has always beenprobably at the top of the national agendas uh lee kuan yew who was the founding the Prime Minister of Singapore, always said in front of the water policy, all of the other policieshave to stay on their knees because this is such an important topic for Singapore.
And the water has always played a very prominent role.
So for Water Guy, there's no better place on planet Earth to be here.
But just deal with a problem which is challenging, which is genuine, and when the people are very serious about preserving water in terms of quality and quantity.
Brent:
[15:30] Do people use Singapore as a template for solving other water issues elsewhere in the world?
Vladan:
[15:36] Singapore is very, very highly regarded as a...
As a center of water knowledge, yes, indeed.
[15:50] In addition to Orange County, California, this is one of few places in the world that relies on use of reused water.
This is a small but very subtle and very important message. Everywhere else in the world, people talk about waste water.
Here we talk about used water. The water we use, and then you can reuse it. It's not waste.
You collect it, you treat it, and you return it. Now, we don't, the new water is not supplied to general population.
[16:36] It's provided for industry. As I said, it's ultra clean. It's clean to the extent to which it has no minerals in it, just molecules of H2O.
So it has very little, it doesn't have a taste.
And it's provided to industry. You know, wafers and industry buys this water because it's super clean.
Fantastic, right? only 2% of used of new water is mixed with the natural lakes only 2% and this is only in periods where, where the lakes are not full like they are now.
[17:23] Because we just supplement water supply if you will so but you know it's a bit of in human psyche I'm not going to drink that water yeah um, But hence, it's not direct. It's notprovided for what we call direct portable use.
Brent:
[17:44] Yeah.
Vladan:
[17:45] It's only indirect portable use and only in very small percentage.
Brent:
[17:48] Yeah. So Orange County is one of the only other places in the world that does this?
Vladan:
[17:52] Orange County has a highly impressive system.
The difference between Orange County and Singapore is that they have much more space.
And their also policies are different. For example, Orange County could buy a land around the river.
Brent:
[18:13] Yeah.
Vladan:
[18:13] And they use a river, so they treat the water, and then they recharge the rivers, and then rivers slowly recharge groundwater.
Orange County is a great example of use of water.
Brent:
[18:30] Water yeah because we grew up in orange county and the school i went to as a kid i think did field trips to the plant that cleans the water yeah.
Vladan:
[18:39] Yeah it's it's i mean of course i had to go there as well so.
Brent:
[18:43] Yeah i.
Vladan:
[18:44] Didn't see you too.
Brent:
[18:46] Small to be noticed back then yeah.
Vladan:
[18:50] Yeah, go ahead.
Keller:
[18:51] How much of the water use in Singapore is industrial versus residential?
Vladan:
[18:57] It's around 65-35, industrial 65-35 residential.
So there is a lot of, because wafers, our petrochemical industry on Jurong Island, there's lots of use and needs for fresh water.
Brent:
[19:14] Do you think that split kind of holds in a lot of places in the world?
Vladan:
[19:18] Not necessarily because Singapore is small and compact and industry such that it requires water treatment.
But we don't have agricultural production and you will have a lot of places in the world that would need agriculture, but they will probably rely on a natural precipitation and othermeans of water supply rather than using a treated water for that.
But this is one part. It's a fear of not having enough water.
And at the same time, we have too much water sometimes.
Sometimes it floods, right? And that is universal because as the cities grow, we change cities in a way that very little water goes into the groundwater because cities are covered withasphalt, with concrete.
So when it rains, this water quickly goes to the drainage system and the drainage system evacuates water in order to prevent flooding.
Managing Flooding Systems
Brent:
[20:35] Because some of your work has been on managing flooding, correct?
Vladan:
[20:39] Yes.
Brent:
[20:40] So could you describe what some of those systems are?
Vladan:
[20:44] Okay. So in early days after independence, so second half of 60s, Singapore flooded regularly.
It was very normal to walk along Bukit Timah Road, Danyan Road, and after the rain, you just walk into knee-deep water. Very normal.
So Singapore then went to develop a very extensive system of drainage.
You have seen the drains, the big, large drains. There are lots of underground drains that you don't see.
Brent:
[21:21] Some of the craziest I've seen. Yeah.
Vladan:
[21:24] But the purpose of these drains was to alleviate the flooding problem, right? To make it smaller.
So what did these drains do? They essentially send the water, they evacuate the water very fast.
So it's like super highway for water. It just goes quickly to the sea.
Today, we have around 8,000 kilometers of drains in Singapore, 8,000 kilometers of drains.
And they're all built and designed to minimize flooding risk.
This also, in the past, was related to health because the stagnant water means mosquito, mosquito breeding means dengue.
And Singapore has solved its flooding problem beautifully.
But the problem is big. You know, the city, again, put yourself, it's maybe interesting, at a time of independence, this country had 2 million, population of 2 million people, and now it's5.5, getting close to 6 million people.
Urbanization and Water Management
[22:37] And these extra 3.5, 4 million people that now lives here, they all need housing, they need roads, They need, and this is, our natural environment has been altered, it's now whatonce was pristine forests where it rains and the water goes down into the groundwater.
Now it goes into the drainage system almost instantaneously.
So this is not unique to Singapore this is everywhere else in the world you go to San Francisco or San Diego or Tokyo you have these problem of a fast urban runoff, that, eventually,has to lead to flooding of a sort we don't have a space, to increase from 8,000 to 12,000 kilometers of drains or make them wider because this is still a densely populated place, right?
So we have to find different ways of dealing with too much water while at the same time dealing with too little water.
So another question, a question for you. Okay, I'm sorry, I'm a teacher, I have to ask a question.
Brent:
[24:03] Go for it.
Vladan:
[24:06] Um...
It rains a lot in Singapore, right?
Brent:
[24:13] Yeah.
Vladan:
[24:16] So how come that we are so water-stressed? The fact is that Singapore is the 10th most water-stressed country on planet Earth.
There are only nine countries that have a more difficult water situation than Singapore. It's a fun fact. It's a fact.
Brent:
[24:31] Wow.
Vladan:
[24:33] Okay. How come? It's, I mean, it rains.
It rains frequently. We have more than 200, 220 rainy days in a year.
So it rains two-thirds of the time.
Brent:
[24:48] I assume most of it's not collected.
Vladan:
[24:52] Yeah, but it's still there, right? So why are we not collecting it?
Keller:
[24:56] Don't have room to store it.
Vladan:
[24:58] Fantastic. We have no space to store the groundwater.
Because again, small country, 720 square kilometers, five and a half million people.
If you then say that there's a central catchment area where nobody lives around MacRitchie and Upper Pierce Reservoir, there are some military training grounds where nobody lives.
So if you were to remove all of this, you will get very close to average density of population of 10,000 people per square kilometer.
You can imagine one kilometer, it's not a very long distance, right? and in other direction, other kilometers.
And then you put 10,000 people. It's a lot of people.
And they all need, there's this conflicting interest in land use.
Replenishing Groundwater in Cities
[25:50] Everybody wants a house to live in. Everybody wants to drive a car.
Everybody wants, we don't have a space for water because, you know, reservoirs, nobody lives in a reservoir.
Brent:
[26:02] Right?
Vladan:
[26:03] So finding solution to these things is quite an interesting problem, I should say, right?
Keller:
[26:12] So this might be a silly question, but within cities where it's mainly concrete and the water doesn't get to replenish the actual groundwater in the land, is there a benefit in placeslike Singapore or New York that are mainly concrete, is there a benefit to replenishing the groundwater itself beneath the city?
Vladan:
[26:30] There's a great... So in...
Prevailing interest these days is in what we call blue-green infrastructure.
Unlike the drains, the drains are really designed to get rid of water, to minimize flooding.
Once the water is in that drain, it has only one way to go. It goes to the sea.
You need to slow down the water and blue-green infrastructure, and you may have heard the words sponge city.
These are cities that are designed differently that when it rains, it collects water, it slowly percolates and goes into the groundwater.
So there is indeed a lot of interest in collecting water and storing it underground in the form of groundwater.
Water but also a technical term is managed artificial recharge MAR where you collect water and then you, help it to recharge faster to go underground much faster so this is a big bigbig thing of course climate change is going to make all of this stuff much more challenging because, yeah it's.
[27:57] In Singapore we are likely going to see more intense rains when you think about it right, you've seen some intensive rain but they're likely to go even more intense and if youcouple this with urbanization meaning more concrete, more asphalt, it's a fantastic ingredient for localized urban flooding, you know, because one degree centigrade.
[28:34] Increase in temperature enables atmosphere to store 7.5% more moisture vapor.
So if the temperature of Singapore, in Singapore, were to increase by four degrees, It means that atmosphere will be able to store 30% more water.
But if you have 30% more water in the atmosphere, that water has to come down in the form of rain, which runs a risk that our precipitation, our rainfall intensity will increase by 20%,30%, which is tremendous, tremendous impact.
Because, you know, what, again, 8,000 kilometers of drains? Are we going to make them bigger?
We can't. So we really need to start looking for very innovative ways of dealing with this.
Rainfall in Singapore
Brent:
[29:29] How much rain does Singapore get?
Vladan:
[29:32] 2,400 liters of water per square meter per year.
It's around 200 liters per square meter per month.
I'll give you London. If I mention London, it rains in London all the time.
They get lots of rain, right? well london gets 650 liters per square meter per year so our three months three and a half months is what london gets in a year.
Brent:
[29:59] Yeah that's crazy.
Innovative Solutions for Water Management
Keller:
[30:01] Yeah and what are some of the leading i guess solutions to this problem like are people looking at different materials for cement or what are some of these options.
Vladan:
[30:12] So so a big option now and we are all getting very excited with.
It's something which people call building with nature.
You see, engineering solutions of 60s and 70s, when we designed those big drainage systems, right, this was a good solution for that time.
But now we are reverting back to the basics to say, hey, you know what? There is.
[30:48] The drains need not be concrete covered drains maybe they could still look like nature inspired solutions where we have plants and vegetation and what have you, so it looksnice it performs reasonably well and also souls, Vegetation cleans the water. You see, all the stuff that we don't like to have in the water, like nutrients, like phosphorus and nitrogen, etcetera, et cetera, et cetera, actually plants need that to grow.
So you can, by reintroducing vegetation in our systems, you are also cleaning, improving water quality in low-energy regimes.
So it's not sent to a treatment plant to improve its quality, but actually every meter of the flow in the drain helps improve water quality.
So this is a big, big area of work now, right? Yeah.
Keller:
[32:03] We talked to someone when we first started the podcast. I was trying to use mycelium, like mushroom roots, and they're drainage from agricultural areas.
Vladan:
[32:11] Yes.
Keller:
[32:11] And it seemed really promising. Yes.
Vladan:
[32:13] Two things. First, root systems of plants, they of course make this more porous, so it restores percolation.
But then you have biofilms, microbial organisms.
I'm told, I'm an engineer, I don't know, but I'm told there is more biomass on planet Earth in microbial life than in microbial life, in all plants and us taken together.
And if you just think about metabolic rates of activity in microbial life, as you realize that this is really a factory of how you can improve.
So the films that get formed on plants as they grow, they're actually very powerful ways of improving water quality.
Hydroinformatics as a Topic
Keller:
[33:04] And then I guess going into your specific research a little bit more, could you explain hydroinformatics as a topic? Yeah.
Vladan:
[33:14] Yeah, well, it's an obvious, easy question and yet at the same time somewhat complicated topic.
Mycelium and Drainage Innovation
[33:23] As I mentioned, I had a visionary professor as my mentor who developed this field of expertise.
[33:37] And it was really meant about information technology for aquatic environment in the broadest possible sense from computer modeling of the flows to databases and geographicinformation systems and AI and these sort of topics.
So it was really none of us in all honesty.
[34:08] Is better computer scientists than fully trained computer scientists.
But we hydro-informaticians have learned, to some extent, adopted these techniques and adopted them for our need in water management.
Evolution of Hydroinformatics Field
[34:27] Maybe it's easier for us to learn a little bit about computer science than it is for computer scientists to learn about water.
So, this is how the field evolved, now it's a very, it has its own journal that exists, I don't know 20 plus years, it has its conference series and it always looks for a new ways of dealingwith water management.
Rainfall Monitoring Using CCTV Cameras
[34:58] So i i have developed interest for example in monitoring rain rainfall using cctv cameras sounds weird right could.
Brent:
[35:12] You define what those are because in the u.s we don't really talk about it like that.
Vladan:
[35:16] Okay so cctv camera people install cctv cameras all over the place I mean, if you come to, next time you enter in metro station here in Singapore, just lift your hand and you willsee there are probably 20 or 30 cameras just on that station.
The intended purpose is different, you know, to monitor activities.
What this particular technology that I got interested in is it subtracts the background, so it anonymizes video images and it just looks at the rainfall streaks and, obviously the wider therainfall streak it is the bigger the raindrop it is right and so using some mathematics and pre-processing you can calculate how many raindrops do you have what is their size and youadd them together with your rainfall intensity.
The idea is that.
[36:26] Every CCTV camera, for which I don't have to pay anything, they're here.
They're installed. Somebody else sank the cost to install these cameras.
All I have to do is just get the video feed, which I anonymize.
I don't care about faces or whatever.
I just like to see my fields of falling raindrops.
And to measure this, so if you just think, there are probably 30, 40, 50,000 cameras in Singapore alone.
These are all my free rainfall monitoring stations.
Implementation of Rainfall Monitoring Algorithm
Brent:
[37:01] Do you have access to the...
Vladan:
[37:03] At the moment, this algorithm is used on half of CCTV cameras operated by our national water agency in Singapore.
Brent:
[37:11] Was it hard to get access to it?
Vladan:
[37:14] No.
We started... So, our water agency, National Water Agency, POV, is really forward-looking agency.
They're very open for innovation.
So we did some small in-house research project, showed a promise and they said, okay, maybe we want to do something together.
Let's do some tests with the three or four cameras, which we did.
Rainfall Monitoring in Jakarta
[37:41] And they said, oh, this looks great. Why not put it on 150 cameras, which we did. and then next thing you know they run it on half of their cameras which is great which isfantastic, But what I want to argue is that, so Singapore has, again, another question.
How many traditional rainfall gauges are there in Singapore?
Brent:
[38:06] 20.
Keller:
[38:08] 300.
Vladan:
[38:10] 75.
Keller:
[38:11] 75.
Vladan:
[38:14] How many of them are in Jakarta?
Brent:
[38:21] 74.
Vladan:
[38:22] Five. Only five. And these are not real-time operated.
Somebody will have to sit on a little moped and drive and see how much rain it fell on that particular spot. There are only five.
And it's not that people in Jakarta don't care about rainwater.
They do care a lot about rainwater and floods.
But society has different priorities and they have to be addressed.
How many CCTV cameras are in Jakarta?
Tens of thousands. So suddenly you have this possibility of accessing the data, I call this opportunistic sensing.
Somebody has placed an instrument for different purpose, CCTV camera, and I'm repurposing that computer vision information into rainfall intensity, And I'm too informed for flying.
We currently work with a second largest telco in Singapore, StarHub, on using the data.
[39:40] On signal strength between telecommunication base stations and antennas.
They have 800 antennas that are backbone of our telecommunication.
And that signal, when it rains, that signal gets weaker. And we found a way of relating the strength of that signal with rainfall. Again, I don't have to pay for cameras.
They're there. They're there for us to talk over the telephone, to exchange data.
Brent:
[40:14] Did you come up with these ideas of using the cameras or signal straights?
Vladan:
[40:21] Yes. But look, you can't say, oh, this is all my thing. But there are.
Brent:
[40:27] Was your lab one of the first to do it?
Vladan:
[40:29] Yes. Yes. And all this stuff with telco as well. Well, there were people that were using this, but nobody managed to bring this to this scale.
Collaborative Innovation with National Water Agency
[40:39] We are now using it. So it's a confluence of visionary partners, like National Water Agency here, that is willing to test, to work with you, to test new techniques.
And, of course, you have to show that this brings value.
And then if you do it once, twice, three, five times, and it becomes really a relationship where this becomes a living lab.
And this is what Singapore says. They want to be a living lab to test some of these things.
Brent:
[41:13] So then what does the telecom company get out of working with you guys? Or just help it out?
Vladan:
[41:19] Two things.
Of course, we're not getting this for free. We are paying somebody for this data, but it's a small fee. Don't tell them that.
Brent:
[41:30] It's really expensive.
Corporate Social Responsibility in Technology
Vladan:
[41:36] But it's also part of corporate social responsibility.
The second largest telco in the country. Shouldn't you be doing something about our climate change?
Shouldn't you be doing something about enhancing resilience against flooding?
And they are really and I have to give them a full credit because they're really good people with hearts in the right place that, have decided perhaps to risk to some extent to work withus to actually develop this technology and make it available and again let's go back to Jakarta are there mobile phones in Jakarta?
Brent:
[42:21] Plenty Plenty.
Vladan:
[42:22] You go to really rural India, they may not have a hydrometallurgical network, but they will have mobile phones.
So, this is what makes me very excited about bringing the latest technology to advanced water management.
Addressing Water Management through Hydroinformatics
[42:42] So, this is perhaps a long answer to your question, what is hydroinformatics?
But it is about opening avenues of processing and information and data in a way that enhances our water management.
Brent:
[43:00] And then what are your goals for how this water like rainfall intensity like that data how is it going to be used and how can that help the people.
Vladan:
[43:08] Well uh you are probably familiar with the fact that there are parts of california where people can't buy flood insurance anymore, because the house the housing is being floodedand they live in the flood plains, why did they build a house in floodplains in the first place how come nobody told them don't build a house there it floods so you have to understandrainfall to understand a flood to advise people the damages due to floods loss of life due to just remember what happened in Do you remember the floods two months ago? Was it?
Keller:
[43:51] No.
Vladan:
[43:52] Hong Kong, very well-developed economy.
New York City, what was the damage?
Economic damages, loss of life, even greater, because we can't express this in dollar terms.
So providing this information helps people plan.
Uh for the future better right how.
Keller:
[44:20] Is the accuracy with like the telephone signal modeling or the cctv compared to the standard traditional.
Vladan:
[44:28] I mean not not not comparable i would i would argue this is this is uh light years ahead of more traditional you see the traditional ways of monitoring rain rainfall is aninstrument called Tipping Bucket, it's sort of a 19th century invention, that gives you very precise information about rainfall, on that spot where the instrument is placed.
So these Tipping Buckets, rain gauges, they give you 75 point observations.
But if you have a telco base stations, well, didn't I just say there was 800 of them.
Plus you have a lines between them. So it's not points, it's a line.
So the whole country is sort of triangulated, crisscrossed by these line.
Potential of Technology in Water Management
[45:27] So you have a higher spatial resolution, you have a data that collected, if you will, even at a range of milliseconds if you care, you can collect the data in milliseconds.
This is a completely new paradigm. And again, scaling this up in places where people can't afford, it's too distant, it's too expensive, it's too, If you go to Manila, you go to Davao, butyou go to places, Papua New Guinea, a range of hydro-metallurgical network is not the first thing that is on the top priority list of some places in the world because they have othermore pressing issues.
Wouldn't it be fantastic that this technology helps instantaneously address all of these issues?
Brent:
[46:16] Have you started to use it elsewhere in the world?
Vladan:
[46:23] Yes, we are starting to use it in Indonesia, Vietnam.
We are trying to start using it in Bangkok, but it's an effort.
Technology works. You deal with more socioeconomic rather than technological challenges because you have You have to find a good telco, a good partner to work with them, and soforth and so forth.
So you're dealing with other non-technology issues, right, to make sure this technology is applied.
Integration of AI in Data Processing
Brent:
[46:59] And then how are some of the newer technologies, like the more advanced AI machine learning, being implemented in this data processing?
Vladan:
[47:08] Say so the data so what you are receiving is a raw signal uh or attenuation or how much power is lost between two points, um you can try to derive laws that uh that relate tosignal strength attenuation with with rainfall, for example.
But these are not necessarily well-understood things to do.
So we're trying to describe it using equations, but we also more and more are able, better and better able to use AI to find this approximation for us.
And that can, every new observation gives you more data, more data can be learned on the fly.
So we can create our own chat rain or whatever you want to call it, GPT.
Utilizing Rainfall Data for Various Purposes
Keller:
[48:09] Yeah. And so if the prediction models are getting stronger with AI and we know more accurately how much rain is falling, what is that used for mainly?
Is it used to indicate policy? What happens?
Vladan:
[48:21] Depends what you want to do. So if you're a farmer, obviously you want to manage your produce, right? what is a good time for you to apply pesticides?
You don't want to apply pesticides before the rain because the rain, it's all gone.
If you are a traffic agency, obviously rainfall affects the traffic a big way, and you may want to re-divert traffic from areas affected by it.
Policy is also very important, but this really falls more towards government agencies that need to decide what infrastructure needs to be built.
Drainage system, capacity, blue, green, All of this stuff gets to form one coherent picture.
So this is where everything comes together.
So policies are typically what government institutions should be doing, defining.
Whereas guys like me, academia or companies are trying to implement this and help them do this. right.
Players in Weather Forecasting Space
Keller:
[49:37] Who are like the major players within the weather force casting space like is that government organizations.
Vladan:
[49:45] Private tradition well it depends where you find yourself there is in the United States there is something called NOAA National Oceanographic and Atmospheric Administrationbut you also have IBM weather company, then you have dark sky you know your Apple weather, there's lots of interest in this as as a business um, Every country will have its ownhydrometeorological organization, but it's becoming more and more widely used outside governments.
Many businesses are affected by the rain. How much is your Uber surcharge if you book it during the heavy rain?
Four times? Eight times?
Keller:
[50:43] No. Yeah.
Vladan:
[50:45] So, would you like to know when to take a taxi?
Brent:
[50:50] Yeah.
Vladan:
[50:50] So, you know, it's a... Logistics, first my last, my delivery, food delivery.
So many human activities are affected by the environmental conditions.
Brent:
[51:03] Yeah, and I think this type of conversation is great because it brings some of those back to the forefront of our mind because especially nowadays most people are so detachedfrom what goes into all the things they interact with on a daily basis from hence.
Vladan:
[51:19] I come to the last question water is not an inherently sexy topic i i'm insulted am i changing your mind a little.
Brent:
[51:31] We didn't need to visit but i think just There's so many people with, everyone's like, oh, water.
Since it's so readily available, when anything's that present in your life, it kind of goes to the back of your mind.
Vladan:
[51:48] Until the moment it stops being there.
Brent:
[51:51] Yeah, exactly. And then do you see the world having more and more water issues coming up as global warming intensifies? Yes.
Vladan:
[52:04] I really think so, yes. I really think so.
I think the issue of having too much, too little, too much water or too little water is going to come.
Sooner rather than later.
I see many people are getting serious concern about it. Climate is accelerating.
There's definitely climate change accelerating, all of that.
Water quality is going to be massive massive massive issue you know i mean we apply so many fertilizers uh uh but if you the uptake of fertilizers by produce is actually small it's It'slike 15%, 85% just goes, and it ends up in our water bodies.
And this is why we have lots of eutrophic water bodies with lots of smelly pea soup looking like lakes.
This is, yeah, the water quality and too much, too little is going to be a massive thing.
Brent:
[53:25] Are you doing research on like water quality?
National Water Quality Monitoring System
Vladan:
[53:28] Yes.
Brent:
[53:30] And is it primarily just focused on Singapore or are you like extrapolating that out?
Vladan:
[53:35] Well, you know, I'm a hydroinformatics guy, so I'm not a water quality guy per se, but I know something about water quality.
So for Singapore, we have, together with my team, we have developed a national water quality monitoring and modeling system of the entire coastal waters of Singapore.
We have eight buoys, very large buoys. You can think of them as a wet chemistry labs placed in the sea.
They float. They continuously take water samples and analyze them.
And then by means of telemetry, they sent the data to our computer systems where we then take this data, combine it with meteorological forecast to continuously produce 72-hourforecast, of water quality anywhere around Singapore. 88 water quality parameters.
Brent:
[54:22] And then you're producing a forecast and then does your system check how accurate it was?
Vladan:
[54:28] Oh, yes, of course.
Brent:
[54:29] And it's very accurate?
Vladan:
[54:31] Very accurate it is. I mean, uncertainties, of course, this is Singapore national system. It's called system Neptune.
But we don't know, for example, what happens at every instance in Malaysia or Indonesia.
So there are some uncertainties of what comes from further places, places that are further away.
But within the national confines, yes, it's very, very good.
This is used by National Environment Agency here in Singapore?
To, as I said, produce any spot of the coastal waters in Singapore, to produce, water quality of 88 water quality parameters it's amazing from e coli to trivial stuff such as temperature toyou.
Brent:
[55:20] Know are you able to measure microplastics.
Vladan:
[55:22] Uh no not really because for microplastics you would have to have grab samples grab samples you have to collect them yeah but we have as i said everything else that ismeasurable from from hydrocarbons to nitrogen, phosphorus.
Future of Water Management Technologies
Keller:
[55:42] And for water management, like looking towards the future, do you think there's more opportunity in building more efficient systems that we already have regardingmanagement or new technologies, like different types like desalination plants?
Do you think there's more of a space for new tech that's more efficient, less energy use, or is it more so individual energy consumption, industrial energy consumption, changing thosepatterns of behavior?
Vladan:
[56:08] It's a great question. The technology that we use can and should improve.
I happen to know a company in Southern California called Energy Recovery, and all they do is recover the waste or suboptimally used energy in water treatment plants.
[56:34] But I think the greater potential is societal.
So we just talked about how much liters of water you use today.
Now imagine if everyone were to make an effort to use, let's say, 10 liters of water per day less.
This is 10 flushes of toilet, 10.
You know, one small flush is 3.3 liters. You know, it's on gallons, so have, you know, 3.3 liters.
Three flashes, 10 liters.
[57:12] Five and a half million people times 10, that's 55 million liters of water today, times 365 days.
It's a lot of water. So it's actually, this is what we often call here, this is our next national water tap. In addition to the four, this is a fifth.
Societal Impact on Water Conservation
[57:35] And this is just being conscious. If I brush my teeth, do I leave the water running or do I... These small things, society, save it energy.
You know, shall I turn all these lights out or not?
Can I be a little more conscious? So I think this is the big, big, big thing that actually comes from generation like you.
You know, this is where, you know, if I, being a lecturer here, I think there are two things of which your generation is concerned and excited about, is sustainability, and the, And AI,data science.
Brent:
[58:22] Yeah.
Vladan:
[58:24] Hence, hydroinformatics, the sexiest water subject alive. What do you mean it's not sexy topic?
Leave my office.
Addressing Water Scarcity in Arid Regions
Brent:
[58:37] What do you think is going to really help the places that are maybe more arid and really struggling for rainfall and just water resources as a whole?
Vladan:
[58:46] Groundwater. When it rains, it rains, capture the water, put it on the ground.
Because if it stays on the surface, it will evaporate.
Brent:
[58:54] Yeah. But what about places like the deserts where it's really not raining much?
What do you think is going to be most important? Is it transporting water to them?
Because I know California does a huge aqueduct system.
Is it going to be desalination? How?
Vladan:
[59:09] No, well, California really struggles with the snowpack.
You know, that's gone. on and then there's a massive almond in the trees.
Yeah, I mean, you know, it's a, again, it's a policy, the biggest water consumer in California is these plantations.
And the water comes from snowmelt. And if snow, so the climate change, you get less snow.
What is going to be the end result? I mean, is the audience plantations really going to survive?
So, sustainable, take as much as you need, you know.
You can't really overproduce something at the cost of something else.
Sooner or later, it will come and haunt you.
So, these are arid places indeed. And in some places, it's just unfortunate enough not to have it.
I don't know what can I advise to, you know, Saharan Africa about. but it rains.
A friend of mine was in Dubai the other day, sent me a video of flash floods in Dubai.
Brent:
[1:00:25] Isn't their soil too dry to really uptake it well?
Vladan:
[1:00:30] No, this is really...
Brent:
[1:00:31] Or does it soak at all?
Vladan:
[1:00:34] No, no. So this was really in a city of Dubai, which is, again, like here, covered by asphalt and concrete and everything else, so the water, you know.
But if it rains in a you can have a mechanism of capturing the water and then recharge it, manage artificial recharge, storing it underground and then it becomes a part of yourunderground aquifer that you can extract, three months later, four months later when you need it.
Importance of Artificial Recharge in Water Management
Keller:
[1:00:59] I think as we wrap up here do you have any other advice for students?
Vladan:
[1:01:04] Enroll in water management course.
Keller:
[1:01:06] Is there any particular one at NUS you recommend?
Vladan:
[1:01:09] Hydroinformatics because that's what I teach it starts in a short few weeks no no no it's a look a climate, and climate and water and this interface with digital is very excitingarea very very and it gives you very meaningful career and then we'll be in hot demand so i would hope also leading to to a good financial recognition. Good job.
Keller:
[1:01:45] Wonderful. Thank you so much.
Vladan:
[1:01:46] Thank you, guys. Thank you very much for stopping.