with ποΈ Brian Moloney β Founder & Managing Director @ StormHarvester
π§ StormHarvester is the leading company in smart drainage systems (and more).
This episode is part of my series on Water Digitization β check it out! π
What we covered:
π How a simple, sturdy yet brilliant concept could have saved Queensland 3 billion pound worth of infrastructure damage
π How hyperlocal weather forecasts and existing network sensors make for a perfect combination if you leverage them cleverly
π How and Why machine learning outperforms modeling
π How it can save utilities tremendous money by preventing sewer blockages long before they happen
π How better rainwater management is an opportunity for greywater reuse and landscaping
π How actuation of networks might be the next frontier
π On-Premise Vs. Cloud Solutions
π Brianβs Vision, Stormharvesterβs Mission & the role that Venture Capital plays in the Water Industry
π₯ β¦ and of course, we concluded with the π§ππ₯ππ πππ§π π¦πͺππ¨π©ππ€π£π¨ π₯Β
Teaser:
Resources:
β‘οΈ Send your warm regards to Brian on LinkedIn
β‘οΈ Visit StormHarvesterβs Website
is on Linkedin β‘οΈ
Full Transcript:
These are computer generated, so expect some typos π
Antoine Walter:
So, hi, Brian, welcome to the show. Hi, how are you doing well? Iβm very good. Thank you. Let me start with a postcard. Actually. Itβs the first time I have the pleasure to welcome someone from Ireland. So youβre in Belfast right now. What can you tell me? That would be crispy about Belfast.
Brian Moloney:
Itβs actually quite a nice day here today. Itβs 12 to 15 degrees, which is unseasonable for us. So the window is open here a bit and I can actually see the sun coming in for the first time this year. So itβs a beautiful environment to be doing the podcast from.
Antoine Walter:
So I hope you have no hard feelings, you know, Iβm French youβre in Ireland. I think for the first time, since a decades, friends won against Ireland in rugby. So I hope you donβt mind if I brag a bit. No.
Brian Moloney:
So weβll give you this one. Weβll be back again next year.
Antoine Walter:
Itβs letβs say very probable. So letβs go to the matter. Actually. I was looking at your path and I was noticing that you started up in letβs say hardcore engineering, civil engineering, hydraulic environmental engineering, and now youβre more on the clouds, artificial intelligence. So thatβs kind of a transition from hard to soft from far, but can you guide us through your steps?
Brian Moloney:
Yeah, so I think it all goes back, I suppose, at the start to my university days. So I went to Trinity college in Dublin and actually I did a course that allowed us to study different disciplines of engineering. So I studied both civil engineering and also some software engineering for the first number of years, and then specialized into the area of civil engineering and drainage and hydraulic modeling and everything associated with that. But it gave me a good, broad understanding for the initial years of the software and the building of software and how that might be applied to stormwater and wastewater drainage networks. So I think it all started maybe back then, and then it developed them to my time in Australia was really where it came to the forefront. Then
Antoine Walter:
Actually your times in Australia, according to what I read on your website, itβs the, the time and place where you had your aha moment, which led you to creating actually your companyβs Stromharvester. Can you tell us something about this aha moment first that we understand that one? And then can you explain me how you have these aha moments down under and you create your company in the UK?
Brian Moloney:
Yeah, absolutely. I suppose I was graduated in Ireland as a civil engineer and started to work initially in drainage, but very quickly we fell into 2006 and 2007, 2008 and the global recession. So in that instance, there was quite a lot of work happening in Australia around engineering and civil engineering. So I decided to go for my career and to move to Australia. I actually started living outside of Brisbane in Queensland, and it was just 2009 when I started moved over here and, and started to get established only a couple of months when Iβd been living there, there was a, a huge catastrophic event called the Queensland floods, where there was 3 billion pounds worth of damage done to the infrastructure of Queensland. And I was living in Queensland at this time outside of Brisbane. So one of the things that struck me when I was there, the flood wasnβt just a one day event.
It happened over the course of about a month. It was continued high rainfall, which would come intermittently every couple of days. But the thing that really confused me is that we knew the rainfall was coming a day or two in advance because we could see from the accurate local forecast information that they had in Australia at the time, but we couldnβt do anything about it as engineers. At the time I was working as a drainage engineer for a local council in Australia. And what we used to do is essentially we would go around to all the housing developments and we would ask all the individuals in the housing developments, could they please empty any of their storm tanks or their rainwater harvesting tanks like a day before the big rainfall would come. And I just taught as, as an engineer, this was crazy, such a reactive way to try and solve a problem of trying to mitigate the first impacts of the heavy rainfall.
So it was actually around the same time in Australia when I got my first iPhone. And I couldnβt understand how there could be so much technology in this little device and then our infrastructure, which is the thing that protects us against these real environmental catastrophes is so reactive, literally everything in the stormwater and the drainage networks in Australia and every country in the world was the same at this time was reactive. So something would happen and then the drainage network would react to me. It just seems so silly that you would have infrastructure there when you have technology available in the world that you could have this really clever, predictive drainage network. And it would be able to see the rainfall that was coming in the coming days. It could understand what that would mean for a pumping station in a sewer network or for a flood tank or a reservoir, and we could drop levels.
And it just seemed that there was such a lack of use of technology in civil engineering compared to how other industries had developed like oil and gas and other elder industries had raised ahead embracing the internet of things and different technologies, subsequently machine learning. So that was really where it came from. It was just so the story was when we were working as part of the council, we were going out and we were actually manually knocking on a door and asking somebody, can you empty your flood tank tonight? Are your rain water harvesting tank tonight? Because thereβs a big rainfall event coming tomorrow. So it started to work in my garage and actually built a little device. That was a little really basic internet of things, connected device that would open and close a small valve based on first of all, an automatic command that I could give it a remote command, link that up to a web portal where I could then open and close the valve from the web portal.
And I suggested to the council, instead of us actually going around and manually opening the valves, why donβt we issue these valves to everybody? And then we can do that remotely from the control center. And then once we got that idea embedded, I started to explore the idea of, okay, why donβt we automatically link these valves to a forecast platform, which would allow us to actually drain down the tank remotely. If we knew that was one of these big storm events coming and they happen regularly, you know, that one in 2011 was a particularly serious rainfall event, but in Australia itβs such a seasonal climate in parts of Queensland. They get this heavy rainfall every year or something similar. So, so this was a repeated problem that they had, but they had no technology solution for us. So really the first, the first, it wasnβt even a company at that stage. It was just developing a solution to solve a problem that I saw in the market. And it was just to help out the council that I was working with. It wasnβt necessarily to try and go and really push a commercialization or anything like that. But that was really the first instance. It was seeing all this other technology that could affect and that could help us and we werenβt using it. So that was the Genesis of the idea.
Download my Latest Book - for Free!
Antoine Walter:
So you were scratching your own itches, which is probably the best way to come out with something, but that was still in Queensland. And you brought that back with you to Dublin or to dub, to, to Belfast. Where did you come back and why did you come back
Brian Moloney:
To Belfast! Yeah, so I was working there and we were getting a lot of success with that project, but ultimately once the recession had sort of started to ease and construction had started back in Ireland, there was a need for engineers back in Ireland, again. So from Australia, I contacted the university, Queens university in Belfast, and we started with doing a remote project initially. And then I moved over to do a closer project with the university as they were very interested in this type of technology and how it could be applied. And, and we started to get some students doing finally a projects. And then we ultimately developed out into, as a student doing a PhD, ultimately on this type of solution. So really there was a strong interest from Queens university to help me develop the concept. And, and they saw the potential. There was for these types of solutions and then, you know, just a desire to, to personal desire to come closer to home.
Antoine Walter:
So that means that your original product or the original service you were bringing is your Stormwater solution that you have today still in portfolio. So that was the origin of the company. Is it still the heart of what you do today because you also have the other part, which is linked to the networks. So what is, do you have a preferred child? Thatβs a difficult question.
Brian Moloney:
No, I donβt. I donβt think we have a preferred child. So the product is itβs been quite a successful product with us now, and weβve got lots of installations of that product around the UK. Weβve got a really strong partnership with the company Wavin, and weβre using Wavin flood to install these smart valves around different projects around Europe. And some of those projects are fascinating, fascinating applications of this new smart technology, really, really optimizing what we can do with stormwater networks. What we started to look at then was that, you know, weβre doing this at localized flood tanks for helping municipalities to, to manage their flood tanks or their storm tanks. And we started to look at the utilities and how technologies like this assist water utilities. So we started to look at drainage networks that might have some infiltration, or might have some storm water coming into them.
We started to look at wastewater networks and the influence of stormwater on wastewater networks. And what we found is that certain situations, the wastewater network, the flow could increase 10 to 20 times during periods of high rainfall on certain areas of combined source. So we said this predictive system that we had built to link infrastructure to the weather forecast, it had to be useful also to utilities because they have a lot of the same problems with their infrastructure being very reactive. So we started to look at Don and we found when we went to talk to utilities that they already had sensors installed all over their pipe networks, that pumping stations and CSO points, combined sewer overflow locations. And they already had a network of sensors, but everything was still reacting. So we started to bring that predictive and smart technology into the utilities and applying that in various different ways, anywhere where we see an opportunity to improve the performance of the drainage network using our predictive technology is something weβre going to be interested in helping a utility to do
Antoine Walter:
Letβs deconstruct that thereβs two elements here, which I find very interesting the first and weβll conclude with the first two, itβs a do it reverse. But the first is you mentioned that there are already a lot of sensors in place on the networks. And I was wondering if thatβs the case on all the networks or if thatβs really the advanced utility that has all these sensors. And the second is this element of the forecast, because as an outsider, letβs say for me, really, for the stupid, I have no clue about how you might program all of that. And I might think that to start with, I wouldnβt start with hyper-local weather forecast. I would do a sturdy, simple thing that might eventually work at 5% off of its potential, but just to do a proof of concept, but apparently you had it embedded from the very beginning. So it was part of your thatβs used the big names of your minimum viable products. To have these hyper-local weather forecast, how easy is it? Does it sound much more complicated than what it really is? Because to me that sounds really like, like complex.
Brian Moloney:
It is, and it can be very complex, but you can simplify it as well in the description and how you explain it. So the hyperlocal forecast information is absolutely critical to drainage networks. So it has to be a key component of what we do. So weβve, up-skilled ourselves and weβve upskilled all of our professionals that are working with us in understanding the different forecasts that are out there, what the abilities of different forecast providers are, how we can get this forecast information into the correct format, how we can use us and then how we critically, how we can link it to different points around the drainage network. So if you take, for example, a sewer network, so take the city that youβre in. If you look at the sewer network of the city that youβre in, there is no point in us predicting that there is going to be rainfall across the entire city for a couple of hours.
So say weβre going to get 10 mils of rainfall tonight, and then it will be gone tomorrow across the whole city to have benefit. We have to know exactly whatβs coming at exactly what part of the city at what time. So then we can start to do something really smart and move the water around or the stormwater or slow it down. I really good analogy. I like to put out there for is could you imagine driving into your city tonight? If there was no traffic lights anywhere, it would be chaos. Everybody wouldnβt know at what junction to move and how to travel around. Thatβs what the drainage networks are below our city. At the moment, there is no smarts. Thereβs no predictive in there. So itβs critical to put this predictive and the smart technology and you can get so much benefit from doing that. Itβs basically like running a city with traffic loads or without traffic lights.
Antoine Walter:
So it is mandatory or itβs, it brings a lot of value that I get it pretty easily. But whereβd you get your data from you have a partnership with a local weather company, or do you do that yourself? Is it data that is open source freely available to everyone? How can I imagine that your weather data, whereβd you get that from?
Brian Moloney:
So weβre not weather forecasters, or weβre not a meteorological agency and ourselves. So what we do is weβll partner with various different meteorological areas in different countries, and sometimes multiple providers in countries. Some providers are good at giving really short term, maybe up to six hours, rainfall for customer go to the meat, more longer-term stuff. Some are good at hyper-local. So itβs by combining various different weather for cost providers all into the mix, all into one system, and then being able to pull out the correct information for the correct asset at the correct time in real time, and be able to do that quickly and make calculations on that quickly. So weβre not experts in predicting the weather, but we are experts in putting various different prediction sources together, putting our expertise at top of them and taking that information and making it relevant to storm water and drainage networks.
Antoine Walter:
Okay. So thatβs very clear, which brings me to my first part of the question, which is the sensors on the network, because you have this data of the weather and the other part of the data is what you get from the network. How common is it to have sensors on the networks? Itβs really everywhere in the UK. You would see those sensors everywhere in Europe, everywhere in the world, or is it really the highest end utility thatβs have their networks fully equipped with sensors?
Brian Moloney
I suppose it all depends. It comes down to how many sensors do you need to make a system like this work? And the bottom line is not as much as what people think they need. Thereβs a real drive over the last number of years to put more and more and more sensoring into storm water and wastewater networks. Itβs not always the right thing to do. You need to, first of all, understand the data that your existing sensors are giving you. So what existing sensors do you have in a network? So a wastewater network is made up of several different components, so that thereβll be pipes, thereβll be pumping stations, and thereβll be usually outfall pints or combined sewer overflows or sewer overflow points where you discharge from. Usually the pumping stations will be sensored. So thereβll be some sensoring and the pumping stations. Cause, cause do you tell him he needs to know how the pumps are operating?
If theyβre deteriorating, if theyβre running well, not running well. So this is the first point you will have monitored. So this is, this is a basic, and this is in pretty much all utilities. The next step then is to CSO as the combined out the outflow points across the network. And usually utilities will have some sensoring at their overflow points. In most instances, if not, they will have a plan over the next number of years to put sensors in there. Because if theyβre going to have some discharge to the environment, they need to know the regulator or the environmental agency of the country will usually want to understand how much theyβre going to discharge, where theyβre going to discharge if theyβre discharging so that they understand, if there will be an, a consequence to the environment of the storm discharge. And then the really sophisticated utilities might supplement this with some sewer level sensing around different areas, particularly maybe areas where theyβve had some blockages or areas where thereβs had some issues in the network. Our system can work once weβve got the pumping station. So we can work at basically level one, censoring level, two censoring, itβs going to optimize. Youβre going to be able to do more and level three years to all singing all dancing system, where you could really, really develop out your benefits.
Download my Latest Book - for Free!
Antoine Walter:
You mentioned your system, and thatβs now the part where reading your case, studies, reading your documentation. I wasnβt able to understand exactly how that works because on that same microphone I had an interview with with Kando I donβt know if you know them can do is putting some sensors in the network and then they do modeling. And from what I understand from your technology, youβre not modeling you do something, which according to what I read is even better. So how do you do to process this data if youβre not modeling?
Brian Moloney:
Yeah, thatβs really interesting. So I have a confession to make, so my background is actually hydraulic modeling. So thatβs what I did for a lot of the time I was working as a consultant. So Iβve come from the industry of hydraulic modeling. Nobodyβs perfect. So the big difficulty, the hydraulic modeling has, is in terms of trying to apply that to real time. Itβs very, very difficult to run the complex hydraulic models that are required in real time to actually give us accurate information in real time so that we can make operational decisions. Cause an operational decision will need to be made. You know, if youβre making it in 15 minutes time, that might be too late because the, you know, in a pumping station, typically itβll turn on maybe after 10 minutes and turn off after five minutes. So you could miss two cycles of the pump.
If youβre a 15 minutes or a half an hour doing your calculation or your collaboration, as well as that, I think most people in the industry understand that hydraulic models are theyβre inherently inaccurate and itβs difficult to calibrate them continually in real time to make sure theyβre always calibrated. So what we do is we actually use machine learning as an alternative. If youβre looking to make operational decisions as to how you should run your network in real time, we believe that machine learning is a better solution than a detailed hydraulic model to do this. So the reason for this is you take the actual data from the actual site, from the sensors in the site. And we use this, we build a machine learning model around the actual data, also using the hyperlocal forecast. So a time series of the information that youβre getting from the sensor time series of the forecast information, also critically, any other influencing data that you think might be influencing the level of the sewer or the pattern of the sewer, like local borehole level for perhaps cause you might have some groundwater infiltration into a sewer network, like maybe a load the river river network.
And I can tell you some fascinating examples of where weβve added different information into the feature set up the machine learning. And we were able to really get our prediction window really, really, really accurate on the back of dash. The theory here is that if you have a sensor in the network, you should use this sensor not to collaborate a hydraulic model, but actually to predict whatβs going to happen at that sensor point, based on the data you get from that sensor and some, some form of sophisticated calculation, like an AI model or machine learning model. So really thatβs the Genesis. Iβm not saying we donβt need hydraulic models as an industry. If you want to build another hundred houses on the outskirts and you want to understand how your whole network is going to be affected, hydraulic models are really good for that. And that thatβs really what theyβre required for. If you want to change the behavior of your infrastructure operationally in real time, we feel strongly and weβve seen the results that machine learning models with sensors are a better way to do that. Thatβs our fundamental principle of the offering weβre bringing into the utility sector. So
Antoine Walter:
The data is gathered is reconciliate it with this weather forecast, data feeds into machine learning and artificial intelligence. But where does all of that happen? Is it a cloud platform? Is it an on premise?
Brian Moloney:
Yeah. So most often we can do on premise solutions, but most often weβll do it in the category recommend thereβs huge advantages to your data processing, to being able to scale up different data processing applications in real time. And then scale back down again, we would recommend strongly that you would do a cloud solution, but itβs not mandatory. We can also do on-premise solutions if required.
Antoine Walter:
So data gets collected to the cloud and the cloud sentence, then a recommendation. How much of actuation do you find on the networks today? Because you have a recommendation there, there would be two outcomes. Whether you have a network which is actuated, then you can directly move and you can benefits at the maximum from this reactivity that you have through your system. Or if you have a recommendation to an operator that whether pushes a button or maybe has to go somewhere with his car and open or close the valve. So how is it today? How, how much activation do you find on the networks as much as sensors or even less, even more?
Brian Moloney:
Yeah, Iβm certainly much less than sensors. I think the, the application of the actuator would have been my guess, but yeah, yeah, absolutely. So the application of the actuation across the network is not, there is certainly some standout examples of some utilities who have that and have it working really well. But in general, thatβs not something thatβs practiced by utilities generally around the world. I think itβs something thatβs calming. Utilities are becoming open. You know, all the factors that everybody understands, climate change, greater populations, constraints, budgets, theyβre pushing utilities to have to try new types of solutions, to get more out of the existing infrastructure that they have. Something like predictive control valves and actuators are critical to that, but you donβt have to go to that step to get the benefit. Weβve done projects like blockage prediction, real-time blockage prediction using this technology, using the forecast and the machine learning.
So where you would be developing a prediction for each asset, you be giving a trench, hold an operating threshold for each asset. And then if you go outside your operating threshold. So if thereβs a anomalous level or anomalous flow in the network, so something thatβs not expected in the network, then thatβs highlighting to you that there might be a problem there. And weβve been able to get really accurate results with a case study we did at Essex water. Weβre also rolling that out with a couple of other utilities, but West six water or a big utility in the Southwest of England, we were able to identify running for three months, across 3000 kilometers of the source. We identified 60 blockages in real time that they were able to clear a pipes out and ensure we didnβt have a pollution incident. So we were detecting the early formations of blockages by identifying anomalous flows in the network.
Antoine Walter:
Actually thatβs an interesting this book story because I have to confess, I had no clue that the problem was thus important. I read in your documentation again, thatβs solely in the UK, itβs 100 million pounds every year to remove those blockages. And from your case study, I did quite a stupid statistics. But I found out that it was roughly one blockage for three kilometer fives every year. So it sounds to me like a huge problem and in what I read as well as that you were able to detect those blockages 14 days before they really happen. So how much of the blockages can you eliminate in percentage?
Brian Moloney:
Yeah. So this is all correct. And I think that a hundred million figure is actually from a couple of years ago. So not that that has increased because the problem is continually increasing sources. This is a huge problem for utilities. The blockage problem. If you go on most utilities, actually, why, why, why is it noise increasing? So both. Yeah, so really itβs, thereβs two factors that contribute to about 90% of the sewer blockages. The first one is wet wipes where these non-degradable wipes are being used more regularly. I read a figure today that last year, I think there was 8 billion wipes sold individual wipes sold in the UK. So all of those are being used. Some of them then are being flushed down the tilers and they donβt degrade like toilet paper might degrade and they stay in the sewer when that den is combined with the Isles and the fatβs coming off of cooking and the facts that are okay to go down the drain.
But once they actually hit the sewers, they cooled down and they form a solid. So this all melds together and it forms like a rock like substance in the sewers. And thatβs what you call a fatberg and thatβs whatβs causing a lot of the issues is these thought these, these, they look grotesque when you see them, but theyβre essentially like the big, big rock formation in the sewer pipes, which obviously is a huge problem because itβs a, itβs essentially a big blockage. So what do utilities are trying to do is to get to these blockage formations Airlie before they turn into these big fatberg formations and actually get in and get them moved on and jet them out so that they get through the network, because if they start to form in the pipe, then youβve got a big problem. So understanding the correct flow, doing really accurate flow windows are level thresholds in the sewers, monitoring them, using machine learning to understand when youβre going outside of an expected threshold or are outside of an expected level. Itβs really, really valuable to allow that utilities are starting to understand the potential and the benefits around.
Antoine Walter:
So itβs good because you defined the fatberg. So I was really wondering what the fatberg was. Now have a clear definition. Itβs crazy. When you think of it, 8 billion wet wipes, only in the UK. If youβre listening to this and youβre still flushing wet wipes in your toilets. I mean, I donβt know how we as professionals shall repeat it, but donβt do that. Just that just donβt. So you find out about this blockage is 14 days in advance. You can clear them and how much of them can you detect all of them? 90% of them, 50% of them. Whatβs your accuracy there?
Brian Moloney:
Yeah. So, because weβve developed out this technology with the machine learning capabilities and because weβre using the hyper-local rainfall forecast and weβve been at this since 2011, believe it or not, that was when the Queensland floods happened. Thatβs when I started to develop out this technology in the projects that weβve run with utilities, we actually havenβt missed a single blockage. Weβve got every single blockage rice weβve got 95% accuracy on our alerts are maybe nine 90, four 95, something like this. And then youβre talking about 5% false positives. So what weβre finding is that you need to have very high levels of accuracy or else the operations crews into utilities lose faith in the solution. So every time you send a message to the operation crew, that you need to go to a certain site at a certain location, they need to find something when they get there, or else they start to say, Oh, this machine learning doesnβt work or this technology isnβt good.
So itβs really, you have to have that level of accuracy or else you wonβt get buy-in from the operational staff. And what we found is we were running proof of concepts with different utilities, and we would actually get a phone call at the end of the proof of concept from the maintenance are the operational crew. That would say, please donβt turn this off. We like this, this, this is working for us. Finally, weβve got a solution. Thatβs able to get us through the blockages. So we donβt have to deal with the big fat burgs down the line. Please keep this running and the utility, but ask us, please keep this going. And to me that proves the value of a proof of concept. When the operational crews are the ones that are asking you, you know, this is working, we shouldnβt take this out.
Antoine Walter:
Thatβs for the networks, if I go back to the attenuation tanks, if I got it right, what you do there is that. And I mean, you are on the donβt waste water podcasts. So trust me, I like the Idea. Itβs about not wasting water. You can reuse a part of that water because you know that the storm event is gone. So thereβs no new rain for the next two days. So why not use that water or to use it as grey water to flush the toilet or non-potable reuse? How does that work?
Brian Moloney:
Yeah, so thatβs a, so weβre jumping all over the place. This is a first solution that we described was putting a smart file versus smart pump, linked to a forecast platform into a storm tank or a flood tank. So let me give you an example of a project. So weβve got some really fascinating projects being installed currently and more coming as we go across this year in the Netherlands. So the net are nuns, certain cities. Weβve got a big project in Austin and a big project in us and dull in these cities. This is what our partner company woven, who do the flood tanks with this solution. Weβre often are putting and retrofitting in these new flood tanks into these cities. And the municipality has understood that thereβs an increased flood risk. So theyβre trying to minimize and make the city safer by installing these flood attenuation structures and tanks in different parts of the city.
So weβve come up with this solution that we control the level in this tank based on our smart files. So we can predict what waterβs coming when itβs coming. And what weβre doing is weβre recycling water out of this tank because now we can hold water in the tank, but if thereβs a big flood event coming, we can release it and drain it down. And then we can, the next time we get some small rainfall, we can hold the water in the tank. Again, weβre able to use this water to actually water the vegetation in the town center, all around the cities. So instead of during the summer when thereβs a potential hose pipe ban on that, all the vegetation dies and all the nice flowers that exist around the center of the town. Now weβre able to water them from the tank that has been installed.
So the tank is not only increasing, there are mitigating the flood risk. Itβs also providing the ability to recycle and keep all your greenery in the area. Green. Sometimes itβs been linked to the local public toilets in the area. So thereβs all different sorts of things. If, if we put it in on an individual development for a particular building, we can actually reuse the water for drinking or for non drinking water. So for a toilet flushing or for cleaning or for irrigating, the, the plants as Iβve described. So itβs about being smart with the water weβre using. Thereβs lots of benefits here because this stormwater otherwise would have ended up in a combined sewer, which mightβve contributed to pollution. It would have probably have ended up if not being discharged in, in a pollution event, it would have ended up in the waste water treatment works, where it would need to be treated, but instead of needed to be treated, itβs actually providing value by allowing it to recycle some water back in the network. It just makes so much sense on so many levels to actually implement this tire of technology. Now
Download my Latest Book - for Free!
Antoine Walter:
It exists. So I know that it sounds a bit like Iβm dragging you left and right, and Iβm sorry for that, but Iβm really curious. Itβs a fascinating field, but of course you donβt want to treat all of that water in your wastewater treatment plants. But I was discussing that with with Marie Launay, hi Marie, if listened to that. And she made some, some studies and some research on micropollutants, which would be given back to the environment through this storm attenuation tanks. And what she found out is that there are waves there. The first backwash of a parking lot is heavily polluted. I mean, heavily is full of those micro pollutants. Whereas the next 20 minutes of a rain event is much less sensible to this pollution. So if you were able to say, okay, letβs take this parts, letβs store it there.
We can put it directly back to the network because thereβs a storm of itβs an ongoing right now, but when the storm event is over, letβs push back this part of the water, which is polluted and letβs reuse the rest in grey water. So you could really do a separation. And when I saw in your explanation video that your, your attenuation tanks are basically split in two. I thought if itβs split in two, why not pitch it in three or in four and have these different waves covered that way? Does it sound utopic or is it like marginal gains and not worth it? Whatβs your thought process here?
Brian Moloney:
Yeah, I think itβs a very interesting suggestion. Very interesting solution. I think the beauty of this technology is that it allows us to explore and really push the boundaries here and things like this, because what we now have is we now have a dynamic control system thatβs installed in the flood tank. So really we can control the flood tank in whatever way we want. And we can retrofit that. We can change that by altering a bit of software so that we run it in a slightly different sequence. For example, we post the valves open every now and again to make sure we get some movement of water in the time to provide it to if I had water sitting there in one position for too long and, and avoid those issues. So if there was a requirement to drain off some of the first flush and then hold the water back and then later release it back in.
So it could eventually get to the treatment works. Itβs all absolutely possible. And itβs the beauty of putting a system in there that can be dynamic and remotely controlled using software is that, you know, if you put a big concrete tank in the ground and you walk away, itβs there forever and it can only perform in that way. It can change its behavior. So if climate change changes, it changes the flood pattern or the discharge pattern, the software can change with us, and then we can keep the tank constant and updated. And itβs just a technology that exists to do that now, which weβre putting into the, into these tanks.
Antoine Walter:
You mentioned this reference in the Netherlands, and you mentioned your partnership with, with Wavin. You are currently distributed in 20 countries, right? Letβs move to the business side of things. First, whatβs your business model? What are you selling right now as itβs a service is itβs a fee based on, on the installation of a software and then a service fee. How does that work?
Brian Moloney:
Merrily? Weβre a SAS company. So it will be a software as a service charge. So thatβs primarily how we work. If we need to put in a valve, a smart valve, there will be a charge for putting in some hardware. Only if itβs required. If the hardware is not required, weβre not pushing the hardware. Weβre all about the software, but sometimes if the utility or the development doesnβt have that hardware themselves, we might have to help them arm provided to them. There can be a hardware charge if they need the hardware. But primarily what weβre trying to focus on is the software side of things. We see ourselves as a software solution that can do prediction better than anybody else in the industry. And that allows us to do all these other technologies, like the automated control of the valve to improve the flow of water and whole buck water, to predict blockages and wastewater networks to do real-time control for utilities. So itβs, itβs the software that enables all this and the real smarts are in the software. The hardware is just a facilitator so that we can, we can put the control mechanism in there, but how you control that is where the brains are. And thatβs where all of our art our IP is.
Antoine Walter:
And what about this distribution in this 20 countries? Is itβs you directly? Is it through partnerships? How do you expand internationally?
Brian Moloney:
Well, we deal with some countries ourselves, and then we have a deal with woven to go around. A lot of countries around Europe woven are obviously a big presence there, a big and their parent company RBR really environmentally focused. And thatβs where the alignment really came really well because our mission is to improve drainage networks so that we can improve the environment. And that really struck a chord with the parent company of Marvin who are RBF and theyβre a Mexican company. So that distribution happens with woven sales network that exists around various different countries. Now, some countries weβll go to ourselves, some countries, weβve got people working on our say, working in ourselves. Weβre very, hands-on in the UK. We do have rainwater harvesting partners in the UK, but with the utilities, weβre very hands-on ourselves and the UK. Weβre obviously very hands-on ourselves in Ireland and then various different countries. Weβve got different models, but a lot of it is true. This wealth and partnership model, which has been a, for more point of view, a fantastic success.
Antoine Walter:
I was surprised I have to say positively surprised because of the radical transparency aspect of that, which I really love in companies, but to see from your, the pitch of your company on that, you mentioned that you are backed by venture capital. And Iβm always wondering when I see companies backed by venture capital inside the water industry and had several of them are on that microphone. Iβm always wondering if hypergrowth is something which is achievable within the water industry. So I would be interesting here to have your take. Yeah.
Brian Moloney:
We are backed by venture capitalists financing. So, I mean, itβs really exciting because it gives us the ability to scale quicker than what we would otherwise be able to do. So weβve installed systems, weβve tested systems, weβve proven the concept in a number of countries and a number of different concepts that weβve all discussed across this podcast. Now we have the ability to take those concepts that have worked across a number of countries and quickly scale them into, into a lot of different countries at the same time. So itβs a, itβs a really, really exciting time and where it aligned really well with our investors, because, you know, within the, the investment funds, the people that we deal with are, believe it or not civil engineers. Now, they may not have worked as civil engineers for a long time, but thatβs where the background was.
Thatβs where theyβve come from true education. Thatβs where they started their own careers. And a lot of those, these people are entrepreneurs themselves that have had successful companies, you know, in this fear. So they really get what weβre doing from day one, theyβve bought into our vision for what we want to do. And, you know, seeing that there is a huge environmental problem is a huge, huge area of growth to improve sewer networks and drainage network, stormwater networks, you know, and thatβs the primary thing that drives us. Itβs the primary thing that drives them. Theyβve also got an environmental slant, which is also a very positive thing from our point of view, that they do care about the environment they are investing in companies that can improve the environment. And I think were bought on, on, on a really aligned path. And weβre both going really going after the market in terms of trying to bring the solutions out to as many countries as we can, and really push ourselves.
Antoine Walter:
You mentioned that they are aligned with your vision, but what is your vision? Where do you see yourself in, letβs say two years, five years on the long run.
Brian Moloney:
So I think really what weβre trying to do is weβre trying to improve drainage networks or wastewater and storm water in our networks. We need to, our mission is to improve the performance of them. And we do this using technologies like hyper-local weather, forecasting, machine learning. So really getting out and being able to improve as many drainage networks and wastewater networks as we possibly can. So we can save as much water as we possibly can take as much water out of the combined store network, prevent as many overflows and pollution incidents as we possibly can in as many countries as we can nearly make an environmental difference so that we can, we can really say, and look back in five years time that we did something, we made a change, we took some technology, we applied it into drainage and we were able to improve the performance of drainage networks and improve the,
Download my Latest Book - for Free!
Antoine Walter:
So itβs not only a vision. Itβs also a mission somehow. How many people are you right now within the company
Brian Moloney:
At the moment we are between 15 and 20. So weβre growing very quickly. So maybe last year we were, we were 10, seven the year before that we were four or five. So weβre growing really, really, really quickly. And with the clients that weβre signing on every week, weβre growing quicker and quicker, the challenge that weβre having is getting, getting people as quick as we possibly can. Thatβs our challenge. The clients seem to want our services. Theyβre interested in what we do, but we need to be able to deliver for the big utility clients that weβre taking on. So weβre constantly trying to push ourselves forward.
Antoine Walter:
You mentioned that once you have something which works, do you deploy it in other countries? So basically once you have your, your product market fit around your solutions, then you can deploy it to other countries. Is it all going to happen through waving or do you also plan to have in a close to mid-term future? I donβt know where a us office or an Australian office would make about sense given your, your path.
Brian Moloney:
Yeah, thatβs a, something very much on my horizon as the Australian opportunity and, and tackling that because of my history and the company coming from Australia, initially, it wonβt all be done through love and love and have great coverage across Europe for the storm tanks in the flood tanks. But the strategy will be to look for other partners that can help us, certainly our partners across America that can help us. And weβre talking to a number of those at the moment. And similarly in Australia, weβre really keen to get in and get active in the Australian market. We think weβve a solution. Theyβve got all the same problems across Australia and New Zealand. So we think that thereβs a really good fit there for what we do. So part of it will be true additional partnerships to supplement the wildland partnership. And part of it will be true direct outreach. We will do a certain amount directly in these countries as well. I think stay close to a market. Itβs really important to have the representation from your own company in those, those markets youβre interested in targeting.
Antoine Walter:
And now, you know, the, the golden last question, I had very honest answer to that question. Sometimes some other times people just told me they donβt know. And because sometimes they didnβt know when sometimes they didnβt want to tell. And some other times they told me, no, thatβs a stupid question. So I, up to you really, all the answers are possible. Do you want to build the next big thing in this water industry? Or do you like building, and at some point you want to exit and start building something else?
Brian Moloney:
I think I like building, I want to take this company forward, but we need to make sure that what we do gives the company the best chance to, to use this technology. Itβs not, itβs not the company, itβs the ability to use this technology and to take this technology to the market. So whether we have to do that through a merger or two acquisitions or how we do that, but itβs what Iβm really passionate about is taking this technology. I can see the benefits of it. I can see what it can do now. And itβs bringing that into this industry. This industry is, you know, true education. Itβs been, my whole career is working in drainage and storm water and wastewater networks. And I really see that thereβs a big difference can be made from using this technology. So however that happens quickest and at the biggest scale, thatβs what Iβm interested in. And if thatβs also taking more funding and really pushing forward further, if itβs also merging with other companies or itβs also acquiring companies, or if itβs also going in under the umbrella of another company that can really promise us that growth and get, and make the biggest difference, weβre open to all of those suggestions, but Iβm just so excited about moving this technology and getting it out there.
Antoine Walter:
So youβre following the mission while you believe in what youβre doing. So I think thatβs a good combo. Makes about sense. Brian, would I propose you is to switch to a rapid fire questions? Thatβs okay for you. Yep. Perfect.
Download my Latest Book - for Free!
Rapid fire questions.
Antoine Walter:
So in that section, I try to keep the question short and Iβd be happy if you keep the answers short, but Iβm not cutting the microphone. So donβt worry. My first question would be what is the most exciting project youβve been working on and why? Yeah, so
Brian Moloney:
The, I think one of the really, really exciting projects weβve worked on is the project weβve completed with West six water. The one weβve completed with waste six water to me would be a really exciting one where weβve been able to predict blockages. We havenβt missed a single blockage in the network that weβve been working on and itβs, itβs transformational for their maintenance crews. Theyβre moving to a condition-based maintenance approach, which is fundamentally changing and improving how theyβre, how theyβre able to respond to these blockage issues. I think that thatβs been a really practical, exciting project.
Antoine Walter:
Okay. Let me cheat here right away. I just have one number, which I didnβt understand the, in this case study, when I read it, itβs minus 96 persons control room alert. What does that mean?
Brian Moloney:
So how the utility is working is at the moment theyβve got older sensors out in the network. So theyβve got their pumping station sensors. Theyβve got some CSO outflow sensors and during periods of heavy rainfall, basically what happens is their control room lights up like a Christmas tree because all of the alerts go during rainfall and the entire control room. You know, it becomes unmanageable to a point where a lot of utilities have to silence for a day and just put all their alerts, turn all their alerts off for a day because they canβt understand where the maintenance crews are meant to go and where theyβre not meant to go because they canβt understand what a genuine alert or whatβs just normal high level in the pipes because of rainfall. And what weβre doing is weβre using our machine learning to determine when is a high level, okay?
Because itβs just rainfall and itβs within the operating parameters are when is a high level not okay because itβs showing something out of sequence. So whatβs, the pipe is usually during this density and intensity of rainfall, duration of rainfall fills to 50%, but our machine learning has arranged then of 55 to 45% that we should be within for that sewer. If the sewer goes to 70% or 80%, then thereβs a problem because something is happening in the sewer that shouldnβt be, and thatβs where the ops crew should go. Not to the other 50 sewers that are behaving in the normal practical way that they just fill up an empty. So itβs, we were able to turn off or silence 95% of their alarms and leave on the 5% that are critical, where they need to get out, or thereβs going to be pollution that shouldnβt go out. That could be avoided.
Antoine Walter:
Okay. All clear. So sorry, Iβm the one that sidetracked to you, but that was intriguing to me. And Iβm glad you made it much more clearer. Whatβs your favorite part of your current job?
Brian Moloney:
I think itβs seen the benefit, seen the rewards that we get when the team comes together. So we had a project today where we got really, really good feedback from the utility client, that they were really happy with something that we were doing a piece of work that we had provided to them. And there was four people involved in that project, all doing a small bit different in terms of, of their own piece, of what they were doing. All the four people came together to provide this one solution that had our clients basically saying, this is absolutely what we need to do. Itβs helped us. And it saved us money. Itβs the reward that we get when the team is working well together to achieve the mission.
Antoine Walter:
What is the trend to watch out for in the water industry?
Brian Moloney:
Oh, well, I, I think I canβt I canβt preach for the last 40 minutes about smart network, smart drainage forecast, prediction control. We say more than real-time control. We say forecast predictive control because real-time control is still, itβs still behind time. If you can do predictive control, thatβs what weβre preaching. And I think thatβs the area where thereβs huge benefits coming down the line.
Antoine Walter:
What is the thing you care the most when youβre working on a new project and what is the one you care the least?
Brian Moloney:
Oh, ah, thatβs, thatβs a really difficult question. What we care the most about a new project is delivering a positive outcome for the client so that we so that we can help them to reach their goal, be that financial or be that environmental. So if weβre trying to, to provide a system in for a client that can actually reduce pollution and, or save them money at the same time. So that would be the, the most important thing, the least important thing. I donβt know, thatβs too difficult for me to answer.
Download my Latest Book - for Free!
Antoine Walter:
Surprisingly, thatβs the most common answer I get to this question. Itβs hard to say what you care, the least I get it. Do you have sources to recommend, to keep up with the water and wastewater market trends?
Brian Moloney:
I think nodding in particular, I think the conferences, the industry conferences are fantastic. So I went to WEFTEC in Chicago last year. Well, not last year, the year before, when it was possible to travel. And that was just such a fantastic experience to see all the different companies, just under one roof, with all their different solutions, all challenging, different problems coming with different mentalities. I highly highly recommend WEFTEC also Aquatech and Amsterdam, a fantastic show, a fantastic presentation, and really, really, really good and seek out people like Swan forum and people like imagine H2O, who are the, who are the groups that bring these innovative technologies together and showcase them to the industry. Theyβd be really the places that Iβd be looking out it. I know thereβs not, you can always, always check out storm harvesters, LinkedIn page, cause weβre very active and weβre really trying to show what weβre doing to everybody.
Antoine Walter:
Well-Placed would you have someone to recommend that we should definitely invite as soon as possible on that same microphone?
Brian Moloney:
Oh very interesting. So the European product manager for, for so motor with woven is Amman called Herschel van from the Netherlands who was a real advocate and Lavin have done some really interesting stuff yes. With what their partnership with storm harvester, but also in enabling more vegetation and more trees using their crates and their planter boxes. And theyβre able to leave the vegetation to breed and get the moisture around the vegetation without causing root damage. Theyβve got a number of solutions like that that are really innovative and real positive. I think heβs a really interesting person too, that has a really good aspect and real good vision on this. I know not, you know, some of the utility managers and leaders that we work with are really, really innovative. And theyβve actually had the chord face of using this and lots of other technology.
Thereβs a utility called West six water. Iβve mentioned before where weβve done a lot of really good work and Jody Knight manages their marketplace or heβs involved with some of their marketplace work. And he is fantastic advocate for solutions that will improve the performance. Even if they are very innovative, that will provide, will improve. The utilities are performance. So Jody is a, a real strong candidate I would feel. And maybe her chill as well. Thanks for the Dublin recommendation. If people want to follow up with you just after this interview, you mentioned your, your company page on LinkedIn, Iβll put the link to your LinkedIn, to your, your webpage. I was also impressed with see, on your webpage number of case studies you had. I mean, itβs, itβs interesting to, to get a better grasp on, on what youβre doing. And I saw that all of you on your, about page, all of the people that are listed are directly listed with the LinkedIn.
So I guess that must be a, a medium of choice. Yeah, very much. Any other place that you would recommend to, to link to? I think our LinkedIn is our strongest the strongest one at this moment in time, we also have a Facebook page, which weβre very active on. Some people like using that platform other than not. I mean, just, I think just LinkedIn, keep an eye on our blog, on our website as well. Weβre quite active in that. And then weβre always trying to add case studies onto our website. So keeping an eye on our case studies and our website also are very interested. And then my personal LinkedIn, Iβm always very active on there. Thatβs I, I find that a really good tool to engage with other industry professionals and share tots on, on different things. So my personal LinkedIn, Brian Maloney on LinkedIn is also a good one. Awesome hook puts all the, all the things, of course, in the, in the show notes. Like every time Brian has been a pleasure, Iβm slightly overtime in know Iβm French. I have to be not really accurate with with the times I give. But I hope itβs still fun for you. It was a pleasure. And I hope to have the chance in a couple of years to check on him.
Other Episodes:
