Listen on Apple Podcasts
Listen on Spotify
Listen on Google Podcasts

What Lurks Beneath: How Robots Can Save City Plumbing with Vanessa Speight

In this episode, we’re going underground, undersea and into your water and sewer pipelines with science fiction’s favorite problem-solvers…robots! Jay sits down with Vanessa Speight, a professor of Integrated Water Systems at the University of Sheffield, to learn how new, spider-like robots have the potential to locate and fix leaks in aging water pipes.

Jay and Vanessa discuss when we might actually see these pipe-traveling bots in action and what they can realistically do for developing nations, where drinking water loss can be as much as 70 per cent due to aging and unmaintained systems.

In our Last Word, professor Lucian Busoniu tells us about SeaClear, a project funded by the European Union, building the first fleet of autonomous robots to collect litter from the ocean floor.

Guest Bios

Vanessa SpeightVanessa Speight

Vanessa Speight is an internationally recognised expert in drinking water quality with a focus on potable water distribution systems, applications of models and data analytics for decision support, public health risk, and regulatory/policy development. Her primary interests are understanding and modeling the reactions between hydraulics, microbiological constituents, and chemical contaminants within urban water systems to design the sustainable water systems of the future. Speight’s approaches include laboratory experiments, field investigations, and data analytics. A new but rapidly expanding field of her research relates to Intermittent Water Distribution Systems – the dominant form of water supply for billions around the world – to better understand the impact of such operations on asset deterioration, water quality and public health.

Ongoing research is investigating how pipes fill and drain under different operating conditions, and what impact those operations have on water quality including biofilm growth, supported by field work in South Africa and Nepal. Speight was the Managing Director of TWENTY65, the EPSRC Grand Challenge Centre for Water working with 6 universities and 100+ industrial collaborative partners across the water cycle to develop flexible and synergistic solutions to meet future challenges for water. She is a Fellow of the Royal Society for Public Health, a Professional Engineer (Virginia, Maryland USA), and Deputy Editor for the journal AWWA Water Science.

Lucian BusoniuLucian Busoniu

Lucian Busoniu received his Ph.D. degree cum laude from the Delft University of Technology, the Netherlands, in 2009. He is a full professor with the Department of Automation at the Technical University of Cluj-Napoca, where he leads the group on Robotics and Nonlinear Control. He has previously held research positions in the Netherlands and in France. His research interests include nonlinear optimal control using artificial intelligence and reinforcement learning techniques, robotics, and multiagent systems. His publications include among others a book on reinforcement learning and several influential review articles. He serves on the editorial board of the Elsevier journal Engineering Applications of Artificial Intelligence, and was the recipient of the 2009 Andrew P. Sage Award for the Best Paper in the IEEE Transactions on Systems, Man, and Cybernetics. He has competitively obtained research funding in excess of 1.5 million EUR and was invited to present research on prime-time national TV and in other media.

Dive Deeper

  • Some companies already use tethered robots to investigate pipes that are inaccessible. But most of the network is currently inaccessible without digging. This is where much smaller, artificially intelligent machines come in. A new generation of underground robotic pipe patrollers is being tested at the Integrated Civil and Infrastructure Research Centre (ICAIR) at the University of Sheffield. (BBC)
  • Intermittent Water Supply (IWS) refers to piped water supply service that is available to consumers less than 24 hours per day and affects at least 1.3 billion people around the world. (International Water Association)
  • Today, more than 2 billion people lack access to safely managed water services. Around 4.5 billion children, women and men live without safely managed sanitation services. (UN SDG 6 Update)
  • The replacement value of UK buried water and wastewater pipes, a network with a length of approximately 1 million km, is between £300B and £600B. (Pipebots)
  • In developing countries, roughly 45 million cubic meters of water are lost daily with an economic value of over US$3 billion per year. (World Bank)

Photo Credit

  • Vanessa Speight – Submitted
  • Lucian Busoniu – Submitted

Full Transcript

Vanessa Speight: We’re hoping to have a world where there’s no more digging up the pipe in the middle of the road. There’s no more of that, you know, do not enter sign because of construction. And that all of that can be done without disturbing the surface and hopefully without disturbing people’s service. The hope is to get to a sort of two or three inch kind of diameter and a small swimmable kind of robot.

Jay: According to one estimate, a full 1/3 of the world’s clean and treated drinking water is lost before it even reaches our taps. That’s about one in every three glasses of drinking water gone. And in developing nations, that water loss accounts for as much as 70%. I’m Jay Famiglietti. On this episode of What About Water, we’re going underground to look at how tiny pipe-traveling robots could be used in city and rural plumbing systems to locate leaks and catch contamination.

The joke goes that one day, we’ll all be replaced by robots. But really, when it comes to some jobs, robots are just better than us. Especially when it comes to getting into places that we can’t. Vanessa Speight is a Professor of Integrated Water Systems at the University of Sheffield. She researches how micro-robots can help detect issues in pipes before they lead to disaster.

Welcome to What About Water, Vanessa.

Vanessa Speight: Thank you. Nice to be here.

Jay: Well, it’s great to have you. And this is a really, really important problem that we know is happening all over the place. Water loss in leaky pipes is a humongous problem everywhere. Why is that?

Vanessa Speight: Well, we have a lot of infrastructure that’s 50 to 100 to even older, years old, and we really haven’t necessarily kept up with the replacement of that pipe infrastructure.

As our cities age, we actually need to keep repairing the pipes and ultimately replacing the pipes to keep the leakage numbers down. So what we’ve ended up with is a situation where you really have a lot of water lost. You had good statistics there at the beginning. It varies by country, by city, but we’re spending a lot of energy and chemicals to treat that water.

That water is increasingly scarce, and then we’re giving it away to the environment, and there’s just no reason to be doing that.

Jay: So I have two leaky pipe stories that I want to tell. One, I lived in a small town called Sierra Madre, which is just outside of Pasadena, and you could tell where the city water mains were from the patches on the street, right along the street.

So rather than replace the water mains, they just go fix them. So you knew exactly where the water mains were from all the patchwork. Second one was a huge water main burst near UCLA. The day after my wife and I bought our daughter a new car and she parked it in the garage at UCLA and she just missed a colossal lower level flood by about half an hour.

And so it’s a big deal everywhere. So even a regular consumer could have a leaky pipe and not know it and be paying for water that they’re not even getting, right?

Vanessa Speight: Yeah, absolutely. And in fact, toilets are one of the huge culprits. And you often don’t even realize because that water is not showing up anywhere.

It’s just going down the drain. So you can be losing 300 liters a day of water through your toilet without even realizing that that’s happening.

Jay: Is that why Donald Trump had to flush his toilet Like 15 times?

Vanessa Speight: Well, no, that’s another issue of water-saving toilets, which are a good idea, but only if they’re not leaking.

Jay: I also wanted to say that I had an irrigation leak. So, you know, in California, water can get pretty expensive. And we were renting a house and had an irrigation leak, and we did not know, just like we talked about, we didn’t know. I found out when I got a hang tag on the door that it was like a $350 fine or something.

Vanessa Speight: Yeah, no, it’s really–

Jay: We think you have a leak and you owe us $350.

Vanessa Speight: It’s really a challenge to find the leaks in your house, almost as much as it’s a challenge to find them out in the road.

Jay: Yeah. So, I understand that this problem is worse in developing nations. Why is that?

Vanessa Speight: Well, developing countries are struggling with trying to build infrastructure quickly.

Many of the countries don’t have the standards and oversight of how the pipes are built, So there’s not the quality of construction to keep them in good shape. When they’re first put in and then they’re subject often to illegal connections, very big problem in many developing countries. People are trying to get water any way they can.

So they’re creating their own water connection to the main pipes and you’ll see all sorts of interesting homemade infrastructure kinds of projects when you walk around and look at some of these pipes. So there’s a real problem with keeping the infrastructure in good shape as well as with building it in the first place.

Jay: Well, so when you think about the complexity of urban water delivery systems, I mean, you know, half the time we don’t even know where the pipes are because as you said, they’ve been around for a hundred or more years. And then combined with what you just told us, it must be a real nightmare to find those leaks, to locate the leaks and then to fix them.

Vanessa Speight: Yeah, we’re getting a lot better at locating the leaks. I would say there’s a lot of techniques that have been under development in the last 20 years. Acoustic techniques where you listen for the sound of the gurgling water or there sort of makes a whistle actually a lot of leaks, much like you can sort of hear the water running in your own piping in your house.

So you can listen and find the leaks. You can do some clever math to add up what’s going in and out of a specific area and try to figure out what’s missing. But fixing the leaks is actually, I still think the harder part. We can find a lot. We’ve got a very good record in a lot of countries of where they are.

Then you’ve got to go dig up the road. You’ve got to make those patches, as you said, which often are the point of the next pothole starting and cause problems in the roadway that no one likes it and people complain or that the roadworks are there in the middle of where I want to go and they’re blocking me. So it’s a really big problem to fix it, and it’s getting increasingly expensive to repair or replace pipes. So a lot of water utilities are just sort of pushing on with trying to get the most life they can out of the existing pipes, but isn’t necessarily solving the problem very efficiently.

Jay: So let’s talk about robots.

Can you tell me about these robots that you’ve been working on that can go into our pipes? How big are they? What do they look like? and are they scary?

Vanessa Speight: Well, if they’re not scary, there’s a very large team.

Jay: Are you sure they’re not scary? ‘Cause I’m scared.

Vanessa Speight: Yes, and they will not jump out of your toilet to bite you.

Jay: Okay, I’m glad we got to that right away ’cause I did not want a spider biting at my butt or climbing out of the toilet and walking into my living room.

Vanessa Speight: No, so there’s a very large team led by one of my colleagues, Kirill Horoshenkov at the University of Sheffield, working on a number of aspects to do with these robots.

This is not a simple kind of technology. So mainly it’s a source of power that’s an issue because they need to swim around in the water. There are some, I would say it’s more advanced in sewers which are full of water and they can kind of walk along the bottom during dry periods, whereas a drinking water pipe is full and under pressure.

So there’s different kinds of concerns there for how you move around, how you would collect data, so different sensors making them small, and then how you would report that data out, because of course you’re buried underneath several feet of street and possibly buildings and other pipes in the way that are metal.

So it’s not a straightforward thing to navigate for these robots. They can’t see anything. It’s dark in the pipe, obviously, so you can’t use any of the visual navigation like we would think about for the autonomous cars, that doesn’t work in a water pipe. So there’s a number of technical challenges. They could work together in swarms, which is a whole theory of robotics, the swarm robotics, where they would collectively do tasks, maybe come together when needed to do a task and then go back alone.

Jay: Okay, I want to share with you the visual that is building up in my brain, okay? And that is one of a subterranean, you know, basically river network, like an underground river network that is now being populated by swimming and crawling creatures in swarms.

Vanessa Speight: Yes, but they’re helping us.

Jay: Okay, I believe you, I believe you.

I really need a visual here on the robots. What do they look like? How big are they? Do they look like spiders, crawly things, caterpillars?

Vanessa Speight: So they do look a little like some of the toy robots that you see there based on some of those similar platforms. They are a bit spider-like actually with little legs on the bottom and sort of a disc on the top.

Jay: So, okay, so we have these spiders. So like how many inches are they? How big are the legs?

Vanessa Speight: That’s changing all the time. So let me see. The legs really are, right now it’s probably two or three inches in diameter and the legs are an inch or so long. So you’ve got that sort of small, small toy robot with little legs that can walk along the bottom of the pipe.

Jay: And how, what would the diameter of a pipe like that be?

Vanessa Speight: The pipes that run in the street to your house are probably four to six inches out on the main street. And then those pipes connecting to your house are three quarters of an inch or half an inch. So that’s the first place where that will block the robots from getting in, is that the connection to your house is quite small compared to the big pipes out in the street.

Jay: I’m gonna send out an army of robots from my house to fight your robots. So what are you hoping these robots would do differently?

Vanessa Speight: Initially, the first thing you actually really need them to do is just map the location of the pipes. As you said, we don’t necessarily know and record keeping is not great.

And there are a lot of places where we might know two endpoints of where the pipe starts and ends, but we don’t know where it goes in between. So first of all, find the location of the pipes and map out the pipe network, map out the diameter, and map out the connections. That’s actually a huge accomplishment just in itself.

The next level you would go to would be sensors, either monitoring pressure or flow or potentially water quality. Now we’ve got quite good sensor capabilities to measure things in water. And probably the ultimate and most distant goal is to do repairs, do small repairs, be able to find a small leak and fix it before it becomes a big leak.

But that requires a lot of energy and how to deliver that energy to a very small robot is still very much a big question that the research is working on.

Jay: So how did you get into this?

Vanessa Speight: Well, as I said, there’s several of us colleagues who have come together. We had a center that we ran for five years called 2065.

It was the EPSRC’s Grand Challenge Center for Water. And we were tasked with thinking 50 years ahead. So we began in 2015, thus the name, and really trying to think, forget some of today’s issues. Let’s think what would a sustainable water system of that kind of future look like. And the first thing you have to think about is there’s no more digging holes in the road.

We don’t have these mysteries like where are the pipes that we have to solve that, you know, the water systems of the future hopefully can do all these things. And the robots then become a way to make that happen.

Jay: So a quick question. So are you doing any visuals down there? Can you do any video? Because it’s a leading question.

I want to know if there is a little man in a rowboat, like in the Tidy Bowl commercials, like down there. Or alligators.

Vanessa Speight: In the water pipes? No.

Jay: OK.

Vanessa Speight: The sewers, whole other story.

Jay: OK. OK. So you’re talking about pipes then and not sewers, too. So let’s be clear about that.

Vanessa Speight: Yeah, so drinking water pipes that are full of water, yes.

Jay: Yeah, yeah, yeah. OK, because there’s great opportunities also on the sewage side as well.

Vanessa Speight: Yes, and actually, the technology is probably a bit more advanced, as I said, because it’s a little more accessible. There’s manholes everywhere. You can get into the sewers and they’re not always full. So there’s just a few of the problems have gone away in sewers that you have with drinking water pipes.

Jay: It’s a whole other subterranean world. There’s a science fiction– I mean, there have been. But I like this one, the drinking water stuff and the swarm. And when she said swarm, she got a script there, I think.

Vanessa Speight: But they’re good and helpful. They’re not dangerous. And they’re not–

Jay: You’re tapping into my dark side. I’m sorry. You’re tapping into my dark– – No I understand just how valuable it is. And listen, I mean, when I was in, when I lived in the LA region, there was a professor there, Stephanie Pincetl, who was trying to map the LA water pipe, water delivery system, just trying to map it. And it’s virtually impossible.

So what you’re talking about is incredibly valuable. So yeah, I’m really looking forward to seeing that. So our listeners understand, in addition to drinking water, basically the gravel beds that many of the pipes reside in underground are actually incredibly transmissive for groundwater. People get confused with groundwater and think there’s underground rivers, but there’s this sort of like a human-made underground river system, a very complicated network.

So super, super awesome. I’m really looking forward to seeing some of those results. So you are trying to tie this to the sustainable development goals, and especially number six, which is about wash and sanitation and hygiene for all. Are these robots actually going to help developing nations?

Vanessa Speight: Yes, I think so.

Because if you think of traditional picture of a developing country water problem, you are probably thinking of some of the charities’ ads where there’s a desert and there’s the lone tree that’s kind of dying. That’s not the reality of life in developing countries now. More than 50% of people already live in cities, and the urbanization rate is through the roof.

The predictions are probably 70% or so in the next 20 years. So really, the problem in developing countries is going to become, if it isn’t already, delivering water through pipes again. So not that the rural water problem doesn’t exist, But the city and urban landscape is changing so quickly and building the pipes.

And as I said earlier, the construction problems, the illegal connections problems, and just the general leakage rates of pipes in developing countries mean that it’s a really critical problem to get water in cities right. And it’s more critical in developing countries really than it is in the developed world.

Jay: So it also sounds to me like this is an opportunity for in the developing world where those pipes are just being laid for the first time. This is an opportunity to get it right, right from the beginning.

Vanessa Speight: Yes, absolutely and can you put sensors in them and can you put the kind of data collection that would help us use advanced tools to analyze where are the leaks happening, where might the illegal water connections be taking place?

So there is an opportunity, but people need to be aware of these technologies and sort of build that in right now, which is not really where the state of the art is in developing countries.

Jay: Yeah. So what do you think the awareness is on this? So if you went to, I don’t know, like, is there a country that you think is really in need of this or that you’re working, you know, trying to work with?

Vanessa Speight: Oh, I mean, every country.

Jay: Yeah, all of them. Everybody.

Vanessa Speight: But, you know, there are a few countries where I think they’re trying quite hard to get on top of their water systems. And an interesting country I’ve worked with quite a lot is South Africa, where the government there has guaranteed a minimum amount of water for free to its households.

And then beyond that, they pay for the water, you know, much like Western water bills are. So in that case, they’ve got some metering capabilities that are needed there and they’re collecting data at the water meters. And that could pair very nicely with data that the robots are collecting. And, you know, another line of work that we’ve been doing lots of research on is and machine learning analysis of a lot of those data sets to learn more about how the system’s performing.

Jay: So I’m curious then, as you roll this out, are you working strictly out of university lab? Are you partnering with industry? Do you have like your own business? How is this rolling out?

Vanessa Speight: The current funding is a university consortium that’s been funded. There’s another year or so to run on that. A lot of industry partners very interested And one of the key goals of this was to make the robotic industry aware of water as a place, as a market for them really, you know, as a place where their technologies could be applied and to make some of those matchmaking connections because I don’t think they’re necessarily aware of the problems and the potential.

Beyond this, then it will need to roll out more to prototype sort of, you know, early technology stages.

Jay: So Vanessa, I’m curious about who pays for this. So what sort of costs are we talking about?

Vanessa Speight: We do not have a great estimate of costs right now just because the manufacturing side is moving very quickly and the miniaturization is moving quickly.

I think what we need to think about and what we’ve been trying to get our stakeholder partners to think about is the trade off. So you can have a very simple, very cheap device. That probably can’t fix the pipes. It can probably run around and make a map and send you the data, but that’s all. Maybe that is a good solution, and you have a lot of them, and if a few break or a few go missing, you’re not too worried about that because they were cheap.

So that’s sort of one end of the spectrum. The other end would obviously be to have this super, very expensive, it-can-do-everything robot, but you better make sure you don’t lose that then, and you better make sure you don’t break that. And I think the reality is somewhere in the middle, but it’s really going to become an issue of what functionality the industry wants and how much to pay for it.

If you think of how expensive it is now to manually go listen for leaks, to manually go and dig three or four holes before you find that leak, you actually can afford a fairly expensive robot if you eliminate doing all of those other wasteful things.

Jay: Vanessa, at the pace of development that you’re going at right now, what do you see the timeline being?

When can we expect to hear about robots crawling around in our pipes?

Vanessa Speight: It’s probably five years out at least for drinking water, but possibly a bit sooner for sewers. Some of the other issues in drinking water in particular are regulatory, because anything that’s in the water has to be not contaminating the water, so we don’t want the robots to poison us.

So ensuring that’s the case and building, especially for the drinking water side, building the infrastructure to sort of launch them and collect them and manage them. On the sewer side, I think that’s possibly a bit closer, again, because you’ve got easy access through manholes in the middle of the street.

You lift the lid and you’re immediately into the sewers. Drinking water pipes are a little harder to reach. So…

Jay: You know, it just sort of struck me like people probably don’t care about any any kind of sensors and robots in the sewers because that stuff is outbound, right? It’s the drinking water that everyone, that’s coming into your home and into your body.

And I’m sure there’s gonna be people that are thinking like, you’re putting little robots in my body, aren’t you?

Vanessa Speight: Well, you’re not actually in the water. It’s staying in the water, but making sure that that’s the case, yes. The sewer though is a bigger problem than people think because sewers overflow, sewers also leak. They leak the sewage into the environment and we don’t want any of those things happening. So in some countries I would say the sewers are the priority. The UK in particular has had a recent focus on sewer overflows and finding blockages is a huge business and if we can do that more quickly and efficiently then there’s potentially a big application there quite soon.

Whereas the drinking water problem, it’s going to be a few more years because it’s just a more challenging environment to work in.

Jay: This has really been enlightening in the sense that you’ve really sort of illuminated this dark underworld of pipes and not scary, but very intelligent and very helpful robots.

And I think they’re going to make a huge contribution. So thank you so much for joining us today, Vanessa. We really appreciate it.

Vanessa Speight: Thank you.

Jay: Vanessa Spite is a professor of integrated water systems at the University of Sheffield. Pipes aren’t the only places where robots could be useful. According to SeaClear, 94% of ocean litter ends up at the bottom of the sea, and the SeaClear project is working to change that using autonomous robots. It’s received funding from the European Union’s Horizon 2020 Research and Innovation Program.

Lucian Brasaniu is a Professor at Cluj Romania, working on this project.

Lucian Busoniu: The SeaClear project aims to build the first team of autonomous robots which cleans up the ocean floor. Most of the litter in the oceans is on the seafloor. At the very least, it’s on the order of tens of millions of tons of waste.

There are four robots in the team. Let’s say the brain of the operation and the robot that supplies power to all the others is an unmanned surface vehicle or USV, which is about six meters long. So you can imagine something maybe the size of a big car. And there’s two underwater robots and the small one is called the mini-tortuga and we use that one to look for the litter.

So it has cameras and it has a sonar and we travel with it pretty much close to the bottom of the sea in coastal areas and we find the litter. Then the second robot, which is bigger and has a gripper, which is used to pick up the litter. So something like a mesh hand made out of two halves, which close and grab the litter and then the liter is deposited. And then the fourth robot is another mobile robot. It’s let’s say an eye in the sky drone. And the role of that one is to identify any liter on the surface and perhaps establish correlations with the liter on the sea floor and also to position sometimes the robots that are underwater so that we know where they are.

In addition to that, it might also watch for obstacles so that our autonomous system doesn’t run for instance into other ships that are roaming the area and so on. So to differentiate between litter and other things in the environment such as fishes or plants that we should not harm we use artificial intelligence specifically deep learning in order to basically look at an image and classify regions of the image.

Is this litter? Is this not litter? If it’s litter what kind of litter is it? And yeah that’s how we are able to tell what to pick up and what not to pick up.

Our system is still in development, which means that we still have to test all these automated components, some of it is still manual and so on, but in our final demonstration, which is planned for the autumn of this year, we are hoping to cover an area of, let’s say, some hundreds of meters on each side, and to be able to pick up the litter from such an area.

in a final commercial version of the system, we should be able to do that more cost effectively than if we were using human divers and of course, more safely.

Jay: That was Lucian Briceanu. He’s a Professor in Cluj, Romania, working on SeaClear in collaboration with partners across Europe. You can learn more and watch some neat videos at And we’ll put a link to that in our show notes. We record this podcast at Arizona State University, which sits on the Homeland of the Akimel O’odham and Pee Posh Tribal Nations.

And we produce this podcast in Saskatchewan, on Treaty 6 territory, the Homeland of First Nations and Métis people. What About Water is a collaboration between The Walrus Lab and the Global Institute for Water Security at the University of Saskatchewan. This podcast is a production of Cascade Communications.

Our producer is Erin Stephens. Our fact checker is Taisha Garby. We’d like to thank our studio crew here at Central Sound at Arizona PBS. Our crew at GIWS is Mark Ferguson, Shawn Ahmed, Fred Reibin, Andrea Rowe, and Jesse Witow. I’m Jay Famiglietti. Thanks for listening.