The Digital Twin Readiness Guide – The Path To Implementation

[00:00:00] Max Herzog: Welcome everyone to this month's edition of Water Data Forum. My name is Max Herzog. I'm with the Cleveland Water Alliance, and it's my great pleasure to have you on here on behalf of the collaborating partners that put on this series,  Cleveland Water Alliance, the Water Environment Federation, and the Midwest Big Data Innovation Hub.

[00:00:33] This is, I believe, now our 3rd year running the forum. We're really interested in engaging you all in a variety of conversations around water and data, demystifying these topics and engaging experts to answer your questions and help sort of paint a picture of what's going on in different areas of work.

[00:00:55] Today, we're going to ask that you all be submitting your questions via the Q&A function at the bottom of your toolbar there, you may see a chat as well, but we ask that you please submit via Q&A, it allows our panelists to also respond, via text, during the presentation. So we may answer a few more questions that way.

[00:01:17] If you do pop your questions into the chat, we will try and get to them, but we will be answering Q&A questions first. My apologies here. I have the wrong title up. But it is my great pleasure to introduce our facilitator for today , Elkin Hernandez, he's the director of maintenance services at DC water. And a member of the West intelligent water technology committee. So, with that, I'll ask Elkin to take things away and introduce today's panelists.

[00:01:51] Elkin Hernandez: Good afternoon morning for some and evening for others. First of all, thank you for making time to join us today, to this webinar on, the twins, before anything, I like to before introducing the panelists that we have today.

[00:02:16] I like to make a reference about the term digital twins and how we understand it. So, one common understanding of the twin is that it's a dynamic digital representation of real world entities and their behaviors, using models and static and dynamic data. That will enable insights and interactions to drive actionable and improve outcomes.

[00:02:43] So, just keep in mind that, that emphasis on dynamic data, that allows us to take these digital modeling concepts to the next steps. It's important for us to be aware that the concept itself is not new, this concept has been around for decades and the underlying principles for even longer than decades, but they were not really, viable for use, in a wider scale before because we didn't have the massive processing and storage computing power that we have right now.

[00:03:25] And neither we have the magic of the internet that allows to, not only share knowledge, but share data, at very high rates. Certainly, in the last few years, a couple of decades and the last few years, we're seeing that the wider sector is embracing digital technologies more and more.

[00:03:48] And, with the idea of, obviously, gaining efficiencies in what we know as the water cycle, should allow us to improve prediction preparedness, for the different challenges we have,  in front of us, it allows process optimization. As is an operational health analysis and, ultimately,  investment optimization, which is a big concern on the industry.

[00:04:17] Today we have a very special panel, and the reason I say that is because we have a very well rounded set of experts that can help us to go from concepts and fundamentals to implementation and to the benefits of the digital twin concept in the water industry.

[00:04:46] We have Mr. Mike Karl. Mike Karl is the VP for, global Digital Leader, for water at AECOM. He's also the chairman of the Swan Americans Alliance. He's been working for several years in this field of digital twins. He has over two decades of experience. He serves the digital transformation initiatives with ACOM's water sector.

[00:05:16] His specialty lies in harnessing digital enabled solutions to boost organizational efficiency, underscored by his depth of expertise in smart water automation, digital twins, and advanced analytics. He's recognized as an industry leader, and he's been consistently at the forefront of digital innovation.

[00:05:38] His dedication is to the water industry's digital evolution. He will be our first presenter. ,after him, we're going to have Dr. Pushkar Regni. He is with Brown and Caldwell, and he is a  wastewater innovation leader. He's a board certified environmental engineer named 40, under 40 award, D by, in 2022 by the American Academy of Environmental Engineers and Scientist.

[00:06:10] Pusker is created with pioneering advanced biological nutrient removal technologies. So we have these, the wastewater treatment experts here, and he's the author of over 50 peer reviewed publications that have served as principal investigator for multiple Water Research Foundation projects.

[00:06:30] He's currently the vice chair of the Web Research and Innovation Symposium and the vice chair of the Web Research and Innovation for strengthening engagement. So certainly,  what we could call an expert on water treatment technologies. 

[00:06:47] And the final panelist that we have today will be, Dr. Agnethe Pedersen. She's Digital Twin Lead with BAN Center SID in Denmark,which is also called BCS in Denmark. This is a utility company, and she has made research about the reliability of the digital twins, especially, within the application of the simulation models, she's in the steering committee of the strong Interpretable Utility Group where utilities come together to share a direction of the digital for the digital digitalization of the water sector.

[00:07:27] So as you can see, we have the different components that would allow us to get a well rounded view of this topic with that introduction. I'll let Mike start his presentation. Thank you.

[00:07:43] Mike Karl: Thank you, Elkin and welcome everyone. Today's going to be an exciting presentation and I'm thrilled that you were able to join us today.

[00:07:54] So I just want to, start the presentation. So it's an honor to work with Puskar and Agnath, and work with the large team that came together to create a path for intuitive water system operations to guide the industry to apply Digital Twins. So let's just start off with what Digital Twin is, some of the benefits. Digital Twins means different things to each person.

[00:08:29] Elkin was talking about the dynamic nature of how it's critical, and it's critical to us as well. One way to think about it is it's really a common environment that's established to allow utilities to achieve their business goals across the life cycle. Whether that's in planning, design, construction, or even operations and maintenance phases, so this holistic perspective across physical and the virtual elements or the digital elements associated with it.

[00:09:04] Really provides a common knowledge repository for each employee to get a much greater perspective and each person supporting them, whether it's a contractor, an engineer, or members of the utility itself. And that's been lacking in the industry and bringing those information systems together and then doing it dynamically is what provides a whole bunch of value to grip, get a lot more information.

[00:09:34] So decisions. and actions are much more well informed. So let's start with the definition of a digital twin. So we worked really hard, over a number of years, to have a great amount of debate and discussion around what is the definition of a digital twin. And we partnered, not just within SWAN, who is a global organization focused on advancing smart water, for utilities in the industry.

[00:10:10] But also AWWA's Digital Clean Group to come up with this consolidated definition. And I think it's worth reading and highlighting a couple of elements here. So a digital twin is a dynamic digital representation of real world entities and their behaviors using models with static and dynamic data that enable insights and interactions to drive actionable and improved outcomes.

[00:10:42] So a couple of key things within that that I just want to discuss for a second is, the dynamic digital representation. So, that means not just static and legacy data, but that doesn't mean all real time or near real time data either, and then the models, you'll hear a lot of advancements in AI and ML.

[00:11:06] And we have an immense foundation of mechanistic oriented models that have been used for hydraulics or biological elements for many, many years. It's really the power of those models and working together that are enabling this new outcome. And then the last part the application of this technology has to be done within a way that it's driving the organization goals forward and that's done through actions that are creating outcomes.

[00:11:43] So, as we began as a group within swan, I'm defining what the consolidated view of digital twins was. We started getting a lot of critical questions that we spent a good couple of years discussing and debating, evaluating forward thinking utilities that we're implementing to answer some key questions.

[00:12:10] One of those questions was, how do I get started? We'll be able to dive into that. And there's guidance and and the readiness guide that will give a link to a little bit. So that was a key question that utilities were really struggling with, how do I get started within the pressures? Another question was, a lot of utilities were hearing about a revolutionary framework that we provided called the digital twin architecture and they were wanting specific guidance on how to apply that architecture.

[00:12:44] And so what we spent those two years doing was really taking all of the confusion And bringing clarity to that to give specific guidance to all of you and others in the industry of how to really tactically move forward. So that's what the team that brought together to really advance the industry did.

[00:13:12] And that was. Over 18 months of specific development on a digital twin readiness guide with over 49 companies participating, we had over 70 meetings and 14 workshops where we developed and matured this readiness guide. So it's with great pleasure that I'm able to give you a screenshot with a QR code that you can scan.

[00:13:39] Where you can download the copy of this readiness guide and it's freely available for you. You just have to enter some information into the website to get the guide sent to you and that guide was really, to enable each person to create actual results that helped optimize performance and better inform your decision making.

[00:14:08] I just want to introduce a couple of concepts that Puskar will go a little bit deeper in. But this concept of digital twin maturing over a journey was foundational to the entire guy and that that journey is built upon and advanced by 4 pillars becoming stronger and taller advancing the organization.

[00:14:35] And that was really the number of outcomes you're able to provide for the organization. The amount of technology and connectivity that you had between your technology and your humans. The insights that those systems are providing, in cooperation with the individuals within each organization, and then the interactions and actions that were enabled by this technology.

[00:15:08] So those 4 pillars are foundational, they may start small, as each journey matures and be built up and advanced over, the journey of growing the digital twins within organizations. The other element was really to establish the path of the journey, and after many different analyses, what became clear as the most successful path was deciding a specific and desired outcome that was shared by each organization and looking somewhat narrow for that early outcome, and then, looking for gaps in technology, skills and organizational readiness. For the need to be closed to achieve that outcome. And then you see the little Ferris wheel as the next icon, which is really the incremental development process that is agile in nature of developing a digital twin. And then, that unified symbol, of expanding the initial digital twin outcome,  to create exponential benefits for each organization.

[00:16:39] So this readiness guide was built around 10 case studies from around the globe that focused on these six areas of work. So from water network, water treatment, wastewater network, wastewater treatment. Urban drainage and pump station optimization, and I love this quote from a good friend that says, “with this, every single drop is under control”. 

[00:17:09] I think that's really exciting to be thinking about the level of maturity that we can achieve with today's available technology. So, I just want to have a brief call for your support and engagement. We're all in this journey together of advancing the way we serve our communities, and we have been charting a course together where we've started with maturing the digital twin architecture, coming up with those foundational pillars,  evolving and tuning and refining the journey that each organization takes to implementing digital twins, releasing the readiness guide, and we're really excited to continue and accelerate that growth across,  really the entire industry.

[00:18:01] And please reach out if you want to get involved, make your mark on the industry and really contribute back to your communities. Our contact info will be at the end. Please reach out. We'd love to  have you join us along this. With that, I'll turn it over to my good friend Puskar and dive into his presentation.

[00:18:27] Go ahead, Puskar.

[00:18:29] Pusker Regmi: Thank you, Michael. So, before, we get started here with my section of the presentation, I just want to start with a global perspective here. Right? And the global perspective here is, there are like 2.4 billion people without access to sanitation, and there are additional 70,000,000 people who lack clean drinking water, right?

[00:18:55] So this is the context that we have, and I don't think digital innovation is kind of an option at this point. It's a necessity if we want to address these global challenges. Right? 

[00:19:13] So I think in general, the modern utility landscape and the modern utilities, they do understand the importance of digital transformation, and they are willing to take steps in the direction of digital twins, in terms of like, it was revolutionizing utilities and how we can do more with less kinds of concepts.

[00:19:40] So, the intention is there, but as you would see, you know, the steps required to get there are still lagging. Right? So, that is the one of the utility concepts here, you know, with intention being there, not having clear direction. And when it comes to digital twins, it sounds the promise of digital twins, the benefits are there, right? But it also comes with a lot of challenges. And challenges that could be new to a utility. These are like complex systems. There are no clear guidelines, daily operation takes a lot of tools and to focus on something innovative and requires a different set of skills is always going to be a hard thing.

[00:20:26] And there is a question of, you know, how do you keep up with the changing landscape of rapidly evolving digital technologies, right? And how do you find the time as an organization to allocate? So these are the main challenges that we have, but the good news here, is that we have made some progress, in terms of getting started.

[00:20:49] And Michael briefly introduced the digital twin readiness guide, and I think it's a kind of an important resource, which outlines how you can get started for successful adoption? And how can you draw some of the insights, that has happened, some of the early wins across the world with the key studies, how can you borrow some of those concepts and implement yourself in your utility, right?

[00:21:16] So with that, the journey of creating this digital twin readiness guide. We had a very clear mission in mind. We wanted to build a common language so that we can have a conversation like this, and we also wanted to align the water industry ecosystem around these, you know, common language and common architecture, right?

[00:21:44] And also we wanted to answer some of the pressing questions around implementation, things like, Why does utility need a digital twin? Where do I start? How do I apply the digital twin architecture? and things like that. So, it had a very clear mission, and I think we were quite successful in getting there.

[00:22:05] And it was all possible because of all the people coming together. So we didn't do this in isolation, we had a broad spectrum of people who informed this readiness guide from academics to utility managers, consultants, and technology providers, they all ship these concepts that are included in the guide.

[00:22:33] I think they are very powerful in that regard. And it captures the product perspective of the industry. And when we started, like, you know, working on this readiness guide, one of the 1st things that we wanted to tackle was, how digital twins are different from conventional models.

[00:22:54] You know, everybody who has done design of big systems, they have used models, they're familiar with it. But what makes that digital twin different from some of the conventional modeling concepts that we already have? Right? And it comes down to, you know, using the real time data, and which is continuously updating.

[00:23:18] These models are generating insights, which are real time, which are automatically updating. And as the result you get, like, real time insights and actionable results, right? So that is the differentiating element of digital twins versus some of the static concepts of conventional models.

[00:23:38]  And Michael already brought up this digital twin architecture. The way that I want to kind of visualize this is that you already have a physical system, right? Like most of the utilities, they have the physical system, both on the infrastructure side and the digital side. The digital twin component comes into play when you start, gathering information, gathering data, analyzing them, visualizing them and at the end of the day, creating some sort of user experience where you gather insights, which can help you, create like informed decisions, optimize performance and actionable results.

[00:24:24] If we are not closing the loop with these actionable results, it's not a digital twin, it's maybe a digital twin in the making, you may have parts and components of a digital twin, and maybe you are in the right direction, but you have not achieved digital twin. So closing this loop with actionable results is a critical kind of a thing.

[00:24:44] And at the end of the day, once you start doing this, the benefits will propel, using these concepts more and more and you'll get more into implementing this, at a wider scale within your utility. 

[00:25:07] And Michael already showed us this journey, but the thing that I want to highlight here is, you know, every utility is in a different place when it comes to their digital twin journey. But these simple concepts of identifying what you need to achieve, have very clear objectives in mind, then moving on to assessing the key questions that you have.

[00:25:32] And also what are the key gaps you have in your technology, what you already possess in terms of skill sets and infrastructure in place, is an important couple of steps. And then you go into this implementation phase, which is kind of underscored by the iterative nature of implementation, right?

[00:25:54] And it could be a little bit different. You know, we are used to designing systems that last for like 50 years, the design needs to be perfect. Otherwise, we are gonna miss our permit or something like that, right? But here with digital twin implementation, it's an exile kind of implementation.

[00:26:14] You can make some mistakes, but you can learn from them and keep making progress, right? So that's the idea. And once you get comfortable with some of the small wins, then you expand, and add more capabilities. And you try to do this system wide, right? And as you start expanding this, the benefits and the return on investment will get more attractive.

[00:26:38] And that's going to fuel this initiative, and move forward, right? So that's the idea with the digital twin journey. Obviously, Michael went through some of the pillars that support this journey, but at the end of the day, you just have to do it to, to get the experience and get into it.

[00:27:00]  So the 10 case studies Michael highlighted, I think they're very important in the sense that these are like some of the early success stories that we have from across the world. They represent different parts of our water system, but at the same time, you know, the way that we have included these case studies, they all use digital twin architecture.

[00:27:28] So it makes it a little bit easy to follow along, but at the same time, they give you enough insights, to apply on your own, utility scale. And these capture some of the wastewater, water network, urban drainage, aspects of our water system.

[00:27:50]  You know, I've been thinking about this, for a while now, based on some of the activities that are happening in our industry, some of the IPs that we see on the digital space that is coming out. And it seems like there are four stages of the utilities that we have, and the first is: you have a process, but on the digital side, you still need to build it up.

[00:28:19] You need to develop your infrastructure. The infrastructure is not quite there yet. Right? And then there is a second stage, the intermediate stage where you have the digital infrastructure, maybe, but there are some of the technological augmentation that is still needed. Right? And if you have gone past that, if you are at an advanced stage, then you want to be aligned on how to use these technologies that you have the digital infrastructure that you have, right?

[00:28:49] And then finally, if you are strategically aligned you can be thinking of doing continuous, you know, optimization, and this is what we can call optimized states. And where you would start leveraging the concepts of digital twin and leverage like digital transformation within your organization.

[00:29:10] And the good news is, if you look at the industry right now, despite some of the struggles with, you know, hiring resources, those still persist, but it seems like there is a positive trend towards, you know, optimism of digital technologies like AI, the concern is decreasing.

[00:29:34] And with this sort of, you know, kind of a landscape, where more and more people are warming up to the ideas of digital leveraging digital technologies in their daily, day to day work life, maybe the adoption of digital twins will also benefit quite a bit. Right? So this is a positive, you know, kind of development in recent times, and that will help utilities employees to make these changes within their organizations. And finally, the next slide. Finally, I want to kind of share this example of a progressive utility, who I would call them, like, almost stays, you know, for optimizing stays a continuous improvement. 

[00:30:25] Where they believe they have, you know, invested enough in terms of digital infrastructure and digital tools. Now they're ready to start looking at digital tools to support their real time decision support, harness like part of the data, and leverage advanced automation throughout the processes.

[00:30:47] And in the process, what they want to achieve is to lay the foundation for digital twins so that they can do this at a large scale and we see this as the utilities progress in this journey,  some of the questions that they're asking is going to get more ambitious, and talking about like progressive utilities, I think we have a excellent kind of presentation from Agnethe who is going to talk about, their experiences, in doing digital twin in Europe and with that, I'd like to pass it to Agnethe.

[00:31:27] Agnethe Nedergaard: Yes. Thank you, that you want to have me speak to the audience here. I am from VCS Denmark, which is a smaller utility, but we are the third, largest city in Denmark, but we are a smaller utility compared to the US I guess, with about 230,000 inhabitants in the two main municipalities that we are servicing in our company.

[00:31:58] I originally did a PhD in digital twins and these simulation models, if we could, try to see if we can replicate the reality in the simulation models. And that's also what this Presentation will be about. So in our utility company, we have the waterworks distributions.

[00:32:24] We have the urban drainage system, and we also have water resource recovery facilities. But in this presentation, I will only talk about the urban drainage system. And in that system, we have a lot of sensors, and we also have a fairly good hydraulic model, which we are applying for a lot of design purposes and also to give updates to the municipality, what happened during the year, how many overflows did we have and so on.

[00:32:55] So we use these models for a lot of things. But one of the questions, because I've been working in this industry for decades now, and one of the questions I had, and I also asked in my PhD is, how good are these models to replicate in reality? Because we have a lot of puzzle pieces, like attributes, the asset database, the surface descriptions, and we also have possibilities in the models and observation and in these analysis and updating tools and all of these puzzles, we are trying to replicate the reality, but do we see the full picture?

[00:33:30] That was one of my questions. And that's what the purpose of this digital twin is, can we evaluate how good the model is? We actually started in 2008 doing a daily comparison of the model and with the observations. So we have been doing this for a long time, but it's the first into the last five years that we actually systematized this a lot.

[00:34:02] And I will show you in this presentation how we did it. So we can follow this architecture,  which is made by us by this group. And I will show you here how we did it because we have the asset database, which is this data collection and sources. And from here we can extract the data from the database, put it into this model building updating tool that we have in Denmark.

[00:34:36] And we can also put in a lot of analytics about the model of the service model of infiltration and water loads. We can put this into this updating tool. and from this updating tool, we can extract an operation model of our entire system, or actually, we can also extract design models and planning models.

[00:34:59] But for this presentation, I'll focus on the operation model because we extract an operation model. And all of this, we are updating this regularly, not day to day, but regularly, maybe every third month or so. And this is the static data in the terms that Michael talked about, and then we have the more dynamic data, where we have the sensing and control and the data collection sources.

[00:35:25] This is the boxes from here, where we have the interim census, the permanent census, and all of our ring gauges, and radar in this area, and all of this we integrate into our data storage,  where we can extract these data and do some analytics. The data quality control, we can make soft sensors, and we can also make this aerodynamics, and model evaluation, which I will be talking about, and everything we can put into a visualization platform.

[00:35:57] And then we can take action through this visualization and see what happens, what needs to be changed. If we see that the model is not behaving as it should, can we find a reason for that and then change the model and then it will hopefully be better the next day. One thing that I want to show. Or say, that it is, there is not only one provider of this, and there's no, not only one platform. And that is especially important for us in the utility company, because we would like to see this as a puzzle piece or like a big puzzle. And we need to change this puzzle. The one puzzle pieces that. That does not match into this picture, or if something happens, we are not relying on only one provider.

[00:36:51] We will rely on more than one. It requires a lot of ownership from us, but we are willing to do this because we think we can become more flexible and we can also put in new providers that have a good solution. Then we can change some of the puzzle pieces. So that is especially important for us. Then I would like to show you some of the results because that may be interesting.

[00:37:17] And we thought that when we do this model evaluation, it's always easy to see it graphically just with a traffic light, so if it's red it's not as good and green is better. And then there are some details about the hatch and all that stuff, but I will not go into detail with that.

[00:37:39] But also to say that I have made 2 plots and one is for everyday events when it's raining, just a minor amount and then, the situations where we have an overflow occurring. And that's the differentiation in what you are looking at is not something that we in the community have been really good at.

[00:38:03] We are just saying we have a model for everything. But what is it really good at? We are mostly using our models to design the network and that is especially in the overflow. But actually, our model actually performs better in everyday events. It's not surprising, but it also realized a lot that we need to be better at looking at these, particularly overflow events, for example, seeing if we can optimize the model.

[00:38:38] And then you can ask, why don't you just calibrate the model, then it will fit. But I have written an entire paper about this. And if you're willing to, or interested, you can read about it. But we would like to make this error diagnostic instead and say, okay, what is wrong with the model?

[00:38:56] Because there are so many places in a model that it can be wrong. So it's not only the parameters where you typically calibrate. It can also be in just the understanding of some of the model tools that you don't understand, and maybe you should look into understanding the model, or maybe it's the physical attributes that simply, or the system attributes that are wrong, or maybe it's the rain that is falling different in the area or something, but I could talk forever about this. So this is just an example of the beavers from my PhD and now I'm, I have implemented all of this into our system. And this is also the table from my PhD, where I tried to smarize how many sensors we had, and at that time, we had around 350 SCADA systems, and now we have a total of plus 400 sites.

[00:39:59] Actually, that's the 270 that you need to compare. So, we are expanding our sensors rapidly and we are putting sensors everywhere and we need to look into these sensors and try to automate some of the analytics that we do, and that's what I'm trying to do so we don't just have basic data.

[00:40:25] But we actually also are using this data. And just to show this tool that we have the visualization that I can just plot. This is the model with the dark blue and the observations with the light blue. And then I can visualize all of this data with the time series. And then I can also get, this is not quite finished yet, but I can make these analyzes to see how good is my model performance like I did before,  from these case studies, then I just tried here from the overflow and it's only based on one and a half years of data.

[00:41:03] So I don't have that many overflow events. So, it's kind of red and there's also some things that I need to check, but that's the purpose is that we get a map. And when I am going to design something new, I can look at this map and say, okay, can I trust my model before I even start doing some simulations in this area?

[00:41:30] So this was my model evaluation and what to do now? Because we have so much data in our cloud now and in our data storage, so now we are progressing doing different stuff. So we have started up doing anomaly detection, both being proactive. How can we prevent that? That we need to go do maintenance on the weekends and do the weekdays.

[00:42:01] And also if something happens, if there's a system anomaly, how do we become aware of this, quite soon instead of maybe seeing it firstly Monday or Tuesday or something? We also use this, simply as an awareness tool of where we have sensors because they're really costly and if some of our operational staff are not aware that there is a sensor in this manhole, then they can destroy it. So we are also using these tools for, and we make these overflow reports instead of having them done annually, then we can have them what happened on the weekend, how much overflow do we have? We have cloudburst alarms for faster information knowledge to our customer care and the project managers that we have.

[00:42:53] If there is a cloudburst, they can inform the customers. And we can also, easily extract these complex data, which are often hidden somewhere, in some SCADA system, and we can now easily extract this and give it out to the external people that we have working for us. And one important thing to say is that we are having a lot of collaboration internally, and we also have a lot of departments supporting this detailed transformation and the digital twin. And it is important for me that all of these departments that are contributing information to the system also gain knowledge from this system, because then they are aware of the information they give, and if there is a failure, if there's something misunderstanding, then they report and then they actually get better results back.

[00:43:56] So, I think that's especially important when you are a utility company and you need to collaborate with all your colleagues. So that was everything for me. I have one slide. If you want the papers, then you can find them here.

[00:44:14] Elkin Hernandez: Well, thank you for the representation. Now I'm going to switch. We have about 14 minutes for a Q&A session. So please feel free to send your questions using the Q&A tool on zoom. And I will myself , try to moderate these questions and we have a couple. So let me do something here.

[00:44:47] Elkin Hernandez: The first question we have is a very interesting one. It says, What are barriers for the adoption of visual twins, is the cost of software sensors and talent? So if anybody here can help us and describe what is the vision and experience of making technology affordable for utilities.

[00:45:11] Pusker Regmi: I have some thoughts on that. So in terms of the software sensors, some of the technology behind it, I think the cost curves for those are like, favorable, the cost is declining overall for the industry. But I think talent is something that is going to be a big deal, right? Like, how do you address the talent issue?

[00:45:37] Because if you think about talent, especially on the digital side of things, it's the same talent that is used by tech companies, and it's going to be hard to get those, you know, resources working for like water industry, just based on some of the pay gaps that we have. Right? So I think one of the ways to address that issue is.

[00:46:01] Training programs, collaboration between like utilities in terms of training and grooming like people, and educating them, grooming these individuals, like within utilities, as opposed to, hiring like 100 percent from the outside, there's always going to be a blend of like, what you bring from outside, but also I think grooming needs to be a big part of that. So that's my kind of high level talk. Maybe others have other ideas.

[00:46:34] Agnethe Nedergaard: I have one comment. Yeah, because that's also one of my concerns, because I'm a utility company, and it is expensive to do all of these solutions and to do this information. It is just expensive. But what we are working on with the interoperability utility group that I'm also in is sworn. Is that we are trying to figure out how we can help each other. So whatever is being developed in Australia, I can use it in Denmark and vice versa. So we are not alone in this transformation and that also smaller utility companies can join this without having the big engine running, but maybe only smaller parts, but it's still running.

[00:47:19] So, how can we make a system that is not only for big utilities with a lot of money. But also for smaller utility companies that can afford the small solution of itself.

[00:47:33] Elkin Hernandez: Thank you. And, I would think that plays along with your comments on trying to make sure, there's not just one solution for everything, but there are different boxes that can be exchanged and fit according to the needs and the means of the different utilities.

[00:47:49] So thank you for that comment. I have a couple of more questions. One, it's very interesting. States, how we foresee the environmental impact of using a digital twin, building and maintaining the infrastructure necessary for digital twins can have environmental consequences including increased energy consumption, electronic waste, and ultimately how are we thinking to control the carbon footprint due to digital twins, considering the climate change do we think is sustainable in the long term? An interesting question that anybody wants to give it a try.

[00:48:31] Mike Karl: I'll start, so I think this question is about, you know, more of the computing impact of digital twins. And, you know, the trade offs associated with that.

[00:48:49] So, there's quite an extensive opportunity within utilities to reduce the impact of energy, climate, carbon, etcetera. And some of the best ways are leveraging these types of technologies that require more computer and intensive computing resources. So you have to be confident, which I think we are, that leveraging that computing resourcing power is a net benefit to the application of it.

[00:49:29] So, when you leverage digital twins, it does require more computing resources, but the benefit of optimizing energy chemical, carbon footprint, even analyzing the maintenance worker route history and, you know, service patterns reduces the amount of time on the road to achieving maintenance input.

[00:49:55] So that's one element. Another element is data center optimization of how their place cooled energy on that. That focus is a little bit outside of the discussion of the digital plan, but I know there's quite a bit of research and advancement in those areas as well. I don't know if I got the question completely right.

[00:50:20] Elkin Hernandez: I think it was an interesting question and a bit open, but I would agree that the optimization that technology can facilitate to a utility, very likely if he's properly executed, should be in excess of whatever negative impact it brings, right? So the savings, as you mentioned, chemical, the carbon footprint from transportation, you know, when you have less chemicals, it's not only the production, the transportation, production of biosolids and the like should be enough to cover those needs

[00:51:01] Pusker Regmi: I have one follow up on that. If I may, you know, so even if we say, like, there is a net positive of a digital twin, you know, regardless of all of these environmental consequences that it may have, there's a net positive in terms of efficiencies gained, reduction in energy, chemicals and everything in between.

[00:51:23] At the end of the day, I think one of the philosophies that we have to follow is that if it goes back to identifying your objectives for the sake of doing digital twin or adding more infrastructure. We shouldn't be doing that. We have to be critical in terms of our objectives.

[00:51:39] What we are trying to solve, you know, and align everything based on that. Having more than what is needed is never a good idea. It creates more confusion in terms of maintaining them and to sustain them. Right? So. Not only is it a hassle, from a perspective of maintaining them, it's not good for the environment as well. So we have to be critical and go back to what we are trying to achieve. Thank you.

[00:52:07] Elkin Hernandez: We have a couple of other questions on the board. One says, for a mature facility that already has its own control monitoring software platform, how can the external modeling software be integrated into? Does everything have to be redeveloped in order to be compatible, or can we reuse some of that existing infrastructure?

[00:52:39] Agnethe Nedergaard: I think that we would like to see that we can set up some standards for the input and output from different tools so that's something we're trying now that we have an a data standard for all utilities that this is the standard of our asset database or our pumps, how they provide data, but it also requires that the software develops for the different. Maybe there is one company who is really good at doing data analytics or data cleaning or whatever you can.

[00:53:23] This tool can do, but it requires that we also get the output standardized from this. So, the next step or the next puzzle pieces know exactly. Okay, this is the output from this tool, whichever company does this tool. And then the next tool can do it, but it requires that we as a community talk a lot together and also agree on the different standards so that we can develop on each other instead of all doing data cleaning, because everyone needs to do that. Maybe one or two companies or standards can say, okay, this is what cleaning does. And then this is the output. Next step, and then we can progress a lot more.

[00:54:08] Max Herzog: Thank you.

[00:54:09] Pusker Regmi: Yeah. And I have one follow up on that. If I may, just to, you know, so on the data collection side of things, I think data is data, whatever modeling, external modeling software that we are using should be able to utilize the data without the need to change anything in place.

[00:54:26] Right? So that should be one of your selection criteria of the modeling software that you shouldn't be changing what you are collecting already, like, it should not result in changes in your data collection and monitoring infrastructure. So that's one of the criteria that you have to follow.

[00:54:44] Otherwise you're not only having a new platform for modeling. You're also changing everything that you know, that's not good, you know, that's not an easy place to be.

[00:54:54] Elkin Hernandez: So we're running out of time. I think there's a question that came up in the chat, so I'll see if we can go quickly over that one. Talks about the characteristics of the SCADA chimaya historian that she's using and how is she using them together with the other systems? Maybe you can try to summarize it.

[00:55:18] Agnethe Nedergaard: I cannot see this question. 

[00:55:21] Elkin Hernandez: It's on the chat

[00:55:27] Agnethe Nedergaard: Sorry. Just, I am.

[00:55:29] Elkin Hernandez: That's fine. That's fine. Okay. That's okay. Good. Thanks for giving it a try. I think we're coming to an end. We have only a couple of minutes. What I want to use this time for is to remind everybody that this series is going to be recorded and it's going to be posted and available for everybody to watch later or share with other people.

[00:55:55] This is the Water Data Forum, this is a joint initiative with the Cleveland Water Alliance, with the Midwest Big Data Hub and the Water Environment Federation. We've been working together for almost 3 years on these areas and we intend to continue.

[00:56:18] In November, we have a new session coming, tilted toward a resilient future managing the impact of a changing climate on community infrastructure. So it's a topic that is well current and upgrades greater interest for the communities and for the policymakers. So hopefully you guys can join us during that presentation. And with that said, I'll say thank you to the panelists for your time, for sharing the expertise. Their contact information, I think it has been shared, so please if you have any follow up questions, contact directly with them or with the information for the Water Data Forum thank you and have a wonderful rest of the day.