In this episode, we interview Ricardo Baibich, CTO of Kando. We discuss the use of IoT data from water utilities to provide visibility into energy efficiency, infrastructure status, and population health risks and trends. We also explored the unique challenges faced by utilities due to the complexity of their networks and their inability to control the volume and quality of water flows into their systems.  

Kando is a data intelligence platform that brings data from waste water to help improve and support public health and the environment. Kando works with public authorities, municipalities, and water utilities to create operational visibility by leveraging IoT hardware and machine learning. 

 Key Questions: 

What are the biggest challenges that utilities face in managing waste water infrastructure? 

How can utilities use waste water data and machine learning to understand population health trends?  

How can legacy software and hardware be integrated into a modern connected system?  

How can data analytics make a circular water cycle possible and preserve water in arid regions? 

Erik: Ricardo, thank you for joining us today. 

Ricardo: Hi, Eric. Thank you for having me in the podcast. 

Erik: Great. Ricardo, I got to ask. First of all, you've got a bit of an accent. I know you're sitting in Israel, but you have a nickname that's a bit more Latino. Where are you from? 

Ricardo: I'm Ricardo, but a lot of people call me Caco. I'm from Brazil. I am 22, 23 years living in Israel now. But yes, I'm a Brazilian. I'm living in Israel doing wastewater. 

Erik: It's funny, because I looked at your CV. Whenever somebody's from Israel, I always look at the bottom and I look what's their military experience. Because sometimes they have this interesting background of, they were working with a special division. I didn't see that on yours. So then, I was thinking okay, I'm curious. He's probably not as an Israeli. 

Ricardo: Yes. 

Erik: But you do have a very interesting background. I'm curious how this led you then to join Kando. Because Kando is a very specific field that you're in, this wastewater field. When did you first touch the topic of wastewater? How did that come on your radar? 

Ricardo: I think it's more of the water, energy, food, and now Climate Nexus, which I am doing for the last 23 years. It comes with connecting, digitalizing this world, bringing IoT data collection and acquisition analysis, geographical information systems to try to understand, tackle these problems. I come from the technology side. I'm a software engineer, but I played with Legos from really early and basic programming, logo programming, and then C. So, I think that coming to Israel--which here, this is a topic which is really of national importance, I think--quite connected me with this. Well, I can do this and that. This is one. I'm the CTO at Kando for the last five, almost five years. But prior to that, I was also working in precisely agriculture and sensing, and distributed sensing systems. Yes, I think that water is something that we don't even understand, we as people, how important it is to us. We don't pay enough attention to it. It's a mission now. 

Erik: Fantastic. Well, I'm glad to hear that your journey started with Legos. I've got a two-year-old who's just getting into them now. I hope, eventually, he ends up as a CTO somewhere down the line and can reflect back on where it all started. But it's really interesting that you mentioned that this is of national importance. So, I'm sitting here in Shanghai. We're just at the end of the Yangtze River, and it rains a lot. Here, water, it's definitely a big topic because there's a lot of of pollution. It's a huge city so, obviously, it's important from that perspective. But I guess for Israel, in countries that are very arid, it has an entirely different perspective, which is really of a national security almost. So, let's talk about why is this such a critical topic. What are the different perspectives that make water management an important topic? 

Ricardo: I think that we're now focused on the wastewater part of it. Of course, the droughts in the world we see. Having potable water is getting more and more expensive, energy-wise and the sources and et cetera. But we, as humans, are generating tons of waste per day in cities that we are dumping as waste and also as wastewater. Now, this wastewater, it has to be collected. It has to be treated. It has to be ideally reclaimed and reused as water, as resources because a lot of stuff is in there. What happens today is that this water is sometimes just returned to the river with minimal treatment, if at all, and so on. It is a water cycle. So, in the end, this water we are dumping, we are drinking it back if by actively repurposing the water or if by dumping back into the water cycle. It will come back to us--the micro plastics in the oceans and the pollution that we create. It's a matter of survival, I think, for humans to be able to manage this expensive resource, this imperative resource for life that is water. Wastewater collection and treatment is a huge part of it, that we don't hear about it every day in the news. It's like underground. But it's impressive when you go and see these people managing these huge networks of thousands of kilometers of pipings under our feet. Every day, they're doing this work. There's where we are trying to help these authorities, municipal, state-wise, federal-wise authorities that have to collect all this wastewater and treat it and try to keep the public health and protecting the environment while also doing that in a timely and financially possible manner. Because it's expensive to do, so it's all thought of. 

Erik: When we talk about the value of data here, I guess we could look at this from diverse perspective. Part of it is the quality of the water at different points. I guess there's also, as you just mentioned, a lot of very expensive infrastructure here. There's potentially an infrastructure maintenance perspective. What are the points where you see the biggest challenge is today in terms of managing this wastewater infrastructure? 

Ricardo: What we are focusing in when talking about the infrastructure itself, because we tell you in a bit that in the end, we understand that what we do is wastewater intelligence. There is more into the wastewater than only the water and the things you can recover. There is a lot of information about what's happening overground. Now we really truly understand that and can give these information and insights as actionable insights to a number of things, not only the management of the wastewater network itself but focusing on the wastewater network or the collection and treatment of the wastewater. The treatment, there is this treatment station. Generally speaking, there is a treatment station that collects the wastewater from the city. There is a big factory that has raw material wastewater and has to put out clean water in different levels depending on the kind of treatment plant and the purpose of the water in the end. Some cities today in the US are already talking about drinking potable water from wastewater, doing the full cycle. Some are actually doing that already in pilot phases. They're really doing that. In Israel, most of the water goes to irrigation. So, it's also the fruits we eat. The watermelons now in the summer is 70% to 80% wastewater in the end. The thing today is that the collection network until now was a big question mark. People knew that they are getting this raw material in their factory, the treatment plant. They have to do something about it, but they are just now finding out what it is. So, it's a hard thing to do. They have to over engineer. They have to have backups over backups. It's a hard work. So, we understood that by being able to see what's happening in the network, we can help these people know what's coming, prepare for it, optimize their work. There are biological processes in the way, so it's important to know how to improve. At the same time, when looking into the network, we are able to see the sources of these anomalies, of these pollution events. So, we are able to pinpoint them and help the authorities go and discuss and talk, even legally do something about it. There are regulations, but our focus and our understanding is that it is much better to talk and discuss and fix the problem. A lot of times it's not bad people. It's something that went off in a process and nobody knew. So, we are seeing that it can be fixed and it's for the greater good, even cheaper for everybody in the end. 

Erik: That's fascinating, because I am coming in a bit more from a manufacturing background. I've never really thought about this before in detail. But I suppose a wastewater facility is very much a process plant. It's a process plant that has, I guess, that you just mentioned, a unique set of challenges that most don't have. So, if I'm a chemical plant or if I'm a beverage manufacturer, I basically know what my inputs are. I very tightly control them. If I have a new supplier, I know who that supplier is. So, you can really optimize that. I guess with this, it's doing a lot of similar things in terms of processing liquids and filtering, and so forth. But you don't have so great control about the inputs, as you just said. You might have a factory release something that they haven't released before, or you might have a flood event, or you might have a pandemic with a bunch of people sick all of a sudden. These are all, I suppose, new inputs into the water that you didn't have before. I've never really thought about that. That's a very interesting perspective. 

Ricardo: I come also from the industrial IoT before. It's called IoT, right? Looking like that, the city itself is a big, big, big factory plant with all the sources. They all come to this central place. It is a big challenge. Because in the end, you don't want to have or you would like to maybe, but it's tens of thousands of kilometers of piping. Really complex network. So, what we tried to do here is to bring AI and Edge and cloud to have this ability of put something that is manageable. You don't want thousands of units if you can have hundreds and maybe tens depending on the kind of IoT unit and sensing that you need at the same place. A lot of machine learning on the side of understanding the data and understanding the geospatial data, so it is a big challenge. I think that, yes, few have seen it as that before. Of course, the treatment plant has a process. It's something that people worked for for decades. But the collection system itself was a bit neglected, I think, on that side of the technologies and the services, the offerings for this new kind of technologies can help. From the infrastructure itself, some factories when discharging some things, creating an environment that allow bacteria to really eat the concrete. Then you have this piping that disappear with time, to the problems that of the amount of organic matters and heavy metals that you may have in the sewage that get to the treatment plant and really make the process much more expensive, if not impossible, sometimes. You hear about those extreme events that crash the treatment plant. Really, they have to restart the whole thing. It may take weeks to restart the whole thing. Imagine that during those weeks, they are overflowing and treated almost. So, these are extreme events, but they happen. We are, in a sense, helping them have a bit of visibility on those. Because as you said, it's like a big factory. They don't know the prime raw material they are getting. The sources, now they have a bit of the visibility. We use a number of dashboards and data analysis to create these actionable insights. In the end, it's the idea of reducing the energy, the C02, the greenhouse gases generated by using this energy and so on. 

Erik: We have this very complex system. Say, I flush the toilet in my apartment and that water goes through a lot of different pipes and infrastructure before it ends up wherever it ends up. Where along this path are you collecting data? How close to the, let's say, supply to the household and the factory? Are you trying to collect it at entry points and do sampling there? Are there specific points in the process where it's aggregated, and you do sampling? Are you more sampling the end product and trying to do a quality assessment there? Where do you do the sampling along this path? 

Ricardo: This simply happens in older network, not in households. I think you could say about the pandemic, which is another thing we started doing but then, it's more on community level. When talking about industrial discharges, sometimes, yes, there are units in front of a specific factory. But these are different kinds of customers as well for the water utilities that are managing. These have permits of how much every kind of material they can discharge. Then they are being monitored and worked with to make sure everybody's on permit. Now, for example, it was something we started in the pandemic. It's that this understanding that you can see the COVID virus in wastewater. Then subsequently, more than that, that you can see, detect future pandemics. You can understand levels of population diet and population, diabetes and all kinds of endemics. So, it's the same system but working on a community level. It is an aggregate level to be able to understand exactly this aggregate level. It can be used from the source. It makes sense to use it for the specific use cases in different parts. When talking industrial sewage, I think it's always a good idea to have this big visibility of everything. So, we try to monitor the bigger industrial areas as areas. But then, we have this ability of high definition in an area, going in and zooming in with more units and more capabilities to pinpoint exactly and understand exactly what's going on. The system is always suggesting where to write all the system by learning what's happening. It's now asking for more details in specific areas. 

Erik: You mentioned a couple of interesting things here. I'm trying to just get a sense of where we are in the market. You mentioned, for example, being able to track things like the obesity across maybe a community or across the city, where there is a higher prevalence or lower. Then I'm imagining fairly precise and unique sensors to do that sampling. Whereas, if you just asked me yesterday what kind of sensors do they have in water facility, I would assume it's temperature and it's acidity, and maybe a couple other things that are blunt instruments for doing general testing. But now you're talking about quite specific things. Can you just share what is the typical status quo sensing equipment in an average wastewater facility in Israel or in the U.S.? Then where do you see there are being potential? I guess it sounds like you're working with some some facilities that are now really taking a few steps forward in terms of going into a more granular detail. What is the status quo in the average facility? Then we can start getting into the more interesting areas to extend that. 

Ricardo: To make that clear here, I think the thing is that not everything we are measuring is online sensors on water or on the medium. There are different data velocities, if you will. So, we have sensors that are touching the water, measuring all the time. We have sensors that we call remote but they are in the pipe, not touching the medium. We have sensors that are outside in some research stations, which the water is being pumped out, checked, and sensed, and then put back into the same place. We have centers. Even in the center family, there is a grab sampling and spot sampling. Meaning, a sensor saw something of interest, we take a sample. It's really a pump filling a bottle that goes to a lab. When I talk about public health, community health and we talk about the COVID project, for example, that is being today measured, I think in average treatment plant, at least in the US, I think the CDC has a program for this grab sampling and measuring in the lab. It's a sensor. It's not an online sensor. We got the sample to the lab. These are where the specifics when talking about the online sensors in the sewage and specifically in the collection system. Because the collection system, you don't have power. You have a hard time transmitting data. So, there is where we identify as the most challenging place that would bring the benefit. There we are measuring more of macro. So, you're right. Temperature is one. Electro conductivity, pH, there are a number of parameters that can and are being measured depending on the exact application, but not to those really parts per billion things you only find in the lab. What we do have as part of the IoT kit is a sampler. The sampler is activated or it works in different modes. It can do composite sampling. It can do a full day sampling. It can do a spot in event sampling. So, that's how the system works, by leveraging all these different data velocities on open-source data as well. But of course, a big part of the hardware system and the field system is to bring this first-hand data set. It's data that comes from the sewage. We assume we're not necessarily using only sensors in the sewage. Lab samples is also a part of it. Of course, not for the real-time alerts. The system has these fingerprints so we know for a lot of pollution events that we have them on the samples in the lab and know exactly what they look like. We have these fingerprints, these models, that can say, "Hey, this is happening now." So, we don't have to have another sample to prove that this is what's happening. Then we can use the sensors to give us visibility without having to sample all the time. 

Erik: Okay, understood. You identify specific factors that correlate with an event. Then you could just track those instead of doing the lab sampling. I think I have a pretty good understanding of your value proposition. It's almost, I guess, it's on the one hand visibility in terms of what is in the system and where there might be issues, and operational excellence in terms of how do you, as a system, reduce energy consumption, improve quality output, reduce CapEx spend by understanding where pipes are eroding and whatnot, and so forth. So, you have this, putting the data to use to accomplish these goals. Let's then talk a little bit more about what Kando is doing here. Are you a system integrator? Are you providing the whole stack from where the integration to also the visualization software and the AI, or are you more of a technology provider that is providing the software and working with integrators who do the hardware installations and so forth? What is Kando's role in this putting a system together? 

Ricardo: We are working with public authorities, with municipalities, with water utilities. We provide a service. We are a service provider. The whole service, we can provide. We leverage sometimes our hardware, and then sometimes it's data from the customer. There is this and this hybrid mode. We work with local partners, so we have these team locally maintaining the system. It's all on us. We provide the actionable insights, and we make all the things that have to happen behind the scenes happen. Regarding the hardware itself, we design things but we do not produce hardware. So, we buy sensors. We design and then buy the unit itself. We are not a hardware company for this sake. Our system is agnostic to the hardware. So, we work with data from customers, being it from sensors or historical systems. We leverage open-source data and, of course, the data we acquire for the project depending on the situation, what needs to be done, and with accordance with the customer. So, we are working with the authorities and local partners in that sense. 

Erik: Great. So, you're wrapping this up as a service. Then it sounds like you're able to provide the hardware if needed, but you would buy that from manufacturers. You provide the service. You provide the software. I'm sure there's legacy sensors. On the software side, is there already a lot of legacy software that you have to integrate with or work around? I guess, there is, to some extent. What does that legacy environment look like from the software side? 

Ricardo: I'm not sure we have a lot of touch points with legacy software as we have data interfaces. We have this cloud system. So, our customers go in the cloud, in the website, where they have this application where they see their dashboards and maps and events. They can understand what's happening. Different levels of decision-making in the end. We do have, of course, APIs, both from data from the customers, from different systems. I wouldn't necessarily call them legacy, but I think this technical part is not really the problem most of the time. Of course, to be able to also provide our insights not only in the website as dashboards but also as data for other systems to integrate and use. So, I always see that the real challenge there is the data product definitions, rather than the real technical software interfaces. I think the engineers normally find common ground easier, if you will. I wouldn't put it as a big challenge but it's something that happens, of course. I think we see it in all the industries now, more and more that this open data thing, this at least standardized data. 

Erik: If we look at that question of how do you make use of the data, who would be the typical users here? Because I guess you have quite a few functions within the facility itself that could make use of this data, and then also external stakeholders that would also want access to the data. So, what does the user profile usually look like? Maybe we can also cover the buyer profile. Where is the budget coming from, and then who are the users of the analysis? 

Ricardo: Regarding the public health project, let's call them, these are public health organizations. In Israel, their national infrastructure. So, we are providing data there and the dashboards for public authorities to really decision makers, policymakers and decision makers. This is regarding sampling for COVID now and wastewater-based epidemiology or public health through wastewater intelligence. The case for the utilities, the collection system that utility manages, they have this operational team. They do sample. They do have this pre-treatment team going around and trying to find the problems in the quality and the permits and so on. This is one thing that is gaining a lot of value from the system and working with the system. At the same time, the people treating in the wastewater treatment plant, they are also made. These dashboards are made for both teams and in different levels of real-time slash decision. They are use case specific. There is a difference between looking for a specific event that comes from a specific factory right now and then four hours getting to the treatment plant and between looking into the monthly discharge pattern of an area. Both are valid use case in the system that have the tools. So, the water utility is different depending where in the world. Water utility can be a municipal authority. It can be a municipality that owns the utility. Some places in the world, they are private companies. Some places, they are public, or half, or hybrid. But they are the ones responsible for collecting and then treating, or collecting and someone else is treating this wastewater. These are the customers. Again, for the public health, part of it is more of health departments, municipal, state-wise, federal, with health departments. These are the stakeholders and policymakers using the system. From operational, OPICS and really reducing energy and materials and labor to decisions in public health now with the public community health. So, it has become really diverse. 

Erik: Do you work at all with industry, with chemical plants or other plants that are using a lot of water, and have a lot of wastewater issues? Are you, more or less, 100% focused on municipalities? 

Ricardo: With specific industries, it's not something that we did with one factory or something like that. This is not where we are. Some big compounds, you have these industrial parks, which if you look into them, they are like mini cities with a huge industrial presence. These are places that have big enough network, I would say, or the need to really have this intelligence. Because it's not only measuring one point. For measuring one point in the process inside the factory, I'm not sure we have something special. There are a lot of sensors that are made to be on pressurized pipes specifically for process monitoring. So, we are after that. We are looking from the first municipal point in the network to the treatment plant. Really, the collection network. Where you have collection networks, we can work with. Bigger compounds, industrial compounds but not a specific industry like one factory. 

Erik: Maybe we can just look at one or two case studies. I think it'd be interesting to walk through a couple of use cases here. What were the problems, and then how did they end up addressing those? Any interesting ones come to mind? 

Ricardo: Yes, I think to showcase or to talk about different kinds of the user, two of the use cases, we have, for example, one of the interesting ones in the US, El Paso, Texas. We are deployed there, and really seeing the results. They have this big treatment plant. They have a lot of industry, a lot of petrochemicals that are heavy. They are building a beautiful project of potable water reuse, using this wastewater and water and then making drinkable water. They are using us there now. We are helping them there. Because you cannot do that without knowing the sources and what's coming. People will be drinking then. It's really, really interesting, I think. Because it showcases the importance of source control, of really understanding what's happening there to be able. It's an enabler of something very important. This is an interesting one. I think that in Italy, in general, we see that they have these operators that operate a number of treatment plants. So, there is also this interesting case of looking into comparing plants and looking into the comparison of cities. But more specifically on these places, the problem is always regarding metals, the industrial areas and metals in the sludge. Because when you treat the water, then you have this sludge. If it's not Class A for example, it can't be reused for creating biogas even. It has to be disposed as toxic waste. So, it costs them hundreds of dollars per ton to dispose these thousands of tonnes of waste being produced. The other option is to have a high-quality enough sludge, that you can produce biogas from it. You can even make them into fertilizers. In Israel, it has done with a lot of it. Then in some cases, you have not only people not paying for disposal, but getting paid for the sludge. It's a big operational cost that can even be turned into an operational gain here. It's a combination of the amount of the pollution and the kind of the pollution. In this specific project in Italia, I'm not sure now which one of them, they reported almost 30% such gain in electricity after using the system coming and talking to the polluters and lowering the level of specific factories, and then being able to lower the electricity use by almost 30%. So, it's huge. 

Erik: The electricity used in... 

Ricardo: In the treatment plant. 

Erik: Interesting. By asking the factories to better control the inputs into the plant. 

Ricardo: Yes, by making sure the factories are on permits and are working as they should. If you're producing less pollution loads and the loads are less toxic, it's easier to deal with. I'm simplifying the toxic. It's a huge area in itself of how wastewater is treated, and it's a big industrial process. So, we are trying to help here, really, give this heads up, in a way, of what's coming and how to optimize them. Having this information, you can optimize from one hand but on the other hand, at the same time, saying, "Hey, it's not only by optimizing the process and knowing what you're getting. You have a way of changing the inputs. You can talk to those factories, and you can agree on some. You can then have better inputs as well." It's feeding the cycle of the economics here for both saving the environment, at the same time better treating the water that will come back in the cycle to us. Whatever we do with that water, in the end, it's in the water cycle. It will be back rather heavy, clean, and safe. 

Erik: Absolutely. Water is becoming evermore precious. So, this is one of those issues where you know that the ROI, that the benefit is only increasing year after year. I think we've covered a good bit of territory here, Ricardo. Anything that we haven't touched on yet that would be important to share with folks? 

Ricardo: Well, I think it really comes back in the end, really in this thing of the impact we are trying to do. It's really trying to not only do and help others do, but I think it's about a change of how we think as human. 

Erik: Just a quick question on that last point as to how we think. I mean, not to insult any of your customers, but governments are not universally considered to be the most forward-thinking in terms of using technical solutions to solve problems. Maybe that's unfair, but I think that's at least a stereotype. But have you found receptiveness to having more data-driven? 

Ricardo: I think it's a great question. I think that you would be surprised about the water people. Because these people do embrace innovation and do embrace technology, because they understand that there's no other way. You see it across the board with the water people. From desalination to water treatment, drinking water treatment, wastewater treatment, they're all the time looking for the next. At the same time, you're totally right. These guys are right as authorities and public authorities to be risk averse. It's not they are not innovative, but they cannot allow themselves the risk of failure. They have this responsibility. We chose them to be there and have these responsibilities. They're not allowed to have this failure. So, this is a challenge. I think we are of a risk management tool as well. We are a risk insurance as well. We are monitoring exactly that. It hasn't been hard for us, I think, to have this understanding from the customer side because we also have to know what's happening to really manage risks. I think I agree that you think about public authorities or policymakers as conservative on the technological side. I think it's more of risk averse than not innovative. So, we work hard on it to show and to prove that what our AI systems say is true by, first, when coming to a new customer, showing samples and show validating the data. Yes, I think it's a good thing that they are the gatekeepers of that. They are right to require that rigorous systems, that they're not a new idea and just an idea but something proven, proven technology. It was a challenge, but we're in the last years. We have billions of gallons of water, and there are sensors and hundreds of thousands of lab samples. I don't know how many. After so many data, I can say with confidence, yes, I think it's risk averse more than not innovative in the tech. I think it's important from us, industrial suppliers, technology suppliers, and service suppliers to also understand that it's our responsibility to prove. That is the risk management around it. It's an important part of it. 

Erik: That's a great perspective, actually. I think it's easy for a technologist to point the finger and say, "You're too afraid. You're too slow to adopt new technology." But of course, the technologist is not the one that's on the line if something goes wrong. It's the operator. The politicians are also behind them. So, it makes sense to be in. Obviously, this is such a sensitive topic. 

Ricardo: In my experience, it's always working together and having this talk with the customers and being open in partners. We want the same good together. It does work. I think that I see the water people, at least, which I can talk about when I see them. I think it's a huge shout out. Because in the end, we are all drinking this water. We go to the bathroom, and everything works. It's under our feet. But people are working hard every day. 

Erik: It's remarkable. It's really the hidden luxury that we all live with and don't think about. Well, Ricardo, thank you so much for taking the time to share this with us today. Maybe just a last question. It would be, how can folks get in touch with you or with the team if they want to learn more about what you're doing at Kando? 

Ricardo: Eric, thank you so much for having me. Well, kando.eco. First of all, I'm there on the LinkedIns and on the website. Everybody can write, and we'll be more than happy to talk and listen. 

Erik: All right. Perfect. Well, thank you, Ricardo. 

Ricardo: Excellent. Thank you so much, Erik. Thank you.