 For this global webcast, this, my name is Omar Siddiqui with EPRI. I'm Senior Program Manager in our Electrification and Customer Solutions Group, and it's my distinct pleasure to host and kick off this digital grid virtual workshop integrating customer resources. Today is day two of our series for this week, and we have a very distinguished panel here, and we're very much looking forward to an interactive session. I am speaking on behalf of my colleagues at EPRI, including Mark McGranigan, our Vice President of Innovation at EPRI, who will be making a few remarks, and our distinguished moderator, Mahir Cebo, and our distinguished panelists for today, and our colleagues at Stanford. Before we begin, I do want to make one statement and inform you all that our colleagues from Stanford University, who have been, as I said, our co-hosts for this week's digital grid virtual workshops, will not be participating in today's webinar. Like many universities, Stanford is suspending all normal business on this day, including participation in meetings in observance of the shutdown academia and shutdown STEM movement that has sprouted online over the past 24 hours as an expression of solidarity against systemic racism in society that disproportionately impacts Black people in the United States. We fully support the decision of Stanford and others to take this principled stance. Please note that Arun Majumdar of Stanford, who was to be our keynote speaker today, will shift his remarks to tomorrow's scheduled technology panel webinar, giving us two keynote speakers for tomorrow. EPRI, the Electropower Research Institute, is unequivocal in its commitment to progress towards a more fair and just society. The tragedy and ensuing events of the past two weeks here in the United States has stirred deep emotions, the motions of shock, grief, and outrage for many of us, our families, our friends, and our colleagues. We acknowledge the pain that many people feel. We applaud instructive efforts to raise consciousness, heal wounds, and create positive change. We ask that you join us now for 30 seconds of reflective silence, and we will proceed with our session. Thank you for your understanding. Thank you. Thank you for your patience. And let's go ahead and get started. A few housekeeping items. All of the, everyone who is an attendee, you are on mute right now because of the number of attendees. There are two ways to interact and ask questions. The recommended method is to submit your questions via chat. So if you look at your web, the panel at the bottom, you'll see a chat bubble that's highlighted there on the picture. You can click on that, and that will allow you to ask a question, and we will be moderating that and will refer to that for questions as we go on. You can also identify yourself and raise your hand virtually, and we can unmute you if you wish to ask a question verbally. But I would recommend the chat feature with your indulgence and understanding. We are recording this webcast, so your participation is your consent to this recording, and we will be posting the recordings as well as the presentations from this particular panel session as well as the other sessions for this week on both the EPRI and Stanford University websites, and that information will be shared to all of the registrants. So our objectives for this week are to present first a vision of a shared integrated digital grid. And this is a concept that we have tried to explain in some of the descriptive material about this webinar series, but there are a number of different aspects to it. As you see there on the figure on the right, part of the essence of this vision is around enabling the dispatch of customer-cited resources, customer DERs, if you will, in a way that's fully integrated with utility operations and planning to become assets to enable flexibility on the grid and flexible dispatch. And in that process there are a number of concepts that are related from transactive energy to the implementation of AI and embedded technology, both on the customer side and devices as well as upstream to the utility. And issues of data security and the common information model are all concepts that I think will have bearing in this. So we want to present and exchange ideas of what this vision of a digital grid looks like. And really to convene here, experts literally from around the world to help to articulate their visions of this integrated grid and digital grid concept, identify their gaps to, identify gaps to achieving this vision. We have identified principally this notion of needing to have an enabling data platform as a critical enabler to actualize an integrated and digital grid. But we want to discuss that in some more depth and in different angles. Also to explain and explore utility requirements and plans from utilities in the U.S. yesterday and the European perspective here today. And then to discuss technologies both commercially available in the market as well as some future technology options to help bridge those gaps. Ultimately what we wish to do, and this is for both our perspective at Epri and Stanford as the co-host is to inform a robust research roadmap and a collaborative initiative industry wide to advance along this path to developing an integrated grid. As I said, Epri I think most people on this call are probably aware but just to take a minute. We're an independent nonprofit research organization. We do research on every aspect of utility operations, generation, power delivery and the end use of electricity for public benefit. And our work is aimed at advancing the safety, reliability, efficiency, affordability, health and environmental sustainability of electric service through collaborative research. Speaking on behalf of our colleagues at Stanford, the Bits and Watts Initiative within the Stanford Precourt Institute of Energy is our co-host. This is a major Stanford initiative and they're focused as it says there on digital innovations for the future electric grid. And they, in their multidisciplinary group, advance business innovation policies and supporting customer control and end user technologies to recast and rethink the relationship between consumers and the grid. So at the bottom you see there are shared workshop goals to convene these experts from around the world, including our distinguished panel here, to develop a research roadmap that's developing a standardized data platform that can enable the shared integrated grid vision with a aim towards connecting and interfacing customer distributed resources with the grid. And just one quick thing before I turn it over to Mark McGranigan. We do have, as I mentioned, a webcast at the third of our this week series tomorrow. And it's at the same start time and we will have two keynote speakers. Our moderator will be Nicola Peele-Molter from VMware and we have another great panel set of panelists. And going forward beyond that, we're very proud to unveil a weekly summer webinar series that will begin on July 1st. And these will, again, be 90-minute sessions with, you know, looking at various aspects of the digital grid. And the first three of these are noted here, but we're planning additional ones into the summer. So there will be more information and the ability to register for these events. So I just wanted to make sure that you were aware of that. And the postings of these workshops this week, as well as the panel, as well as the series into the summer, will be available on the EPRI and Stanford sites. With that, let me turn it over to Mark, maybe to say a few words and to say a few words about our moderator, too. So Mark, it's over to you. Okay, Omar, thanks a lot. And I appreciate everyone joining today. We had a very good session yesterday with some U.S. perspectives on this topic of digitalization and connected customers and connected communities, which I think is one of the, you know, top research and demonstration priorities for really understanding the way the future grid is going to work and making communities and customers an integral part of that. We've been talking about a concept of an integrated grid for a number of years now and kind of define that as making local energy optimization part of the global energy optimization. And I think the integrated grid is really all about this challenge of integrating all the way down to smaller customer equipment as part of optimizing the way the grid operates. It's a huge opportunity and it's really a requirement as we go further in integrating renewables and with increasing need for flexibility of grid operation and through the process of electrification of transport and heat and other elements, it's only going to become more important to make all these pieces work together. I was part of the ETIP S-Net working group for that Mahir Chebo and Miguel Sanchez-Fourney are leading and that's been a great focus to help bring together thoughts in this space, especially we added to the function of working group four, which is on digitalization. We added the part of our title of connecting the customer to the function that we're trying to accomplish there. So I'm really delighted to have Mahir kind of leading the panel of a European perspective on this topic today. I think it should be a very interesting discussion. I look forward to helping out with the questions and I'm sure I'll help myself and learn from the process. So I think rather than take up a lot of time myself, I'll just turn it over to Mahir. Mahir was in charge of this space in GE digital grid for a number of years and now is headed in a new venture still close to GE and helping companies with their strategy in this space going forward. So we look forward to continuing to work with Mahir and Mahir, I'll turn it over to you for the further introduction of the panel. Thanks, Mark. That's a great opportunity to be here. Fantastic brands, Iprai and Stanford and I hope you will have another good session tomorrow. I heard that the good session yesterday was really good. Just give me a second. So I would like to, before I introduce the panelist, I would like to just say a few words about the initiative we have in Europe at the European Commission, which is around the ITIPSSET Working Group for Digitalization of Energy. Basically, we have evolved since 2005 when we started looking at the European Technology Platform smart grids, we have evolved from a picture which is grid focused, electricity focused, a picture which is more convergent around different sources of energy and different energy networks as well, heating, gas, electric and so on. And the ITIPSSET, which stands for European Technology Innovation Platform, smart networks for energy transition has set up a few years ago, like three years ago, five working groups who look after the overall energy value chain, grid topics, flexible generation topics, batteries topics and also digitalization topics. Now, what we did in this group is drafting what could be our recommendations about the digital technologies as enablers, about the use cases that digitalization can bring, and also we looked at the cybersecurity to make sure that everything we bring that digital is robust, is sitting on the robust platform. And more recently, we evolved from drafting recommendations in white papers, we did have like a long one for technical guys, 170 pages available on the ITIPSSET portal. And we have a shorter one for policy makers where we have mainly our recommendations, like 24 pages to present our point of views. But now we evolved as well as MAC has been very much active in this into going through what could be a big idea, a big project for Europe, I call it the Airbus of Europe, which could be a digital platform we could build for Europe, democratized, broadly used by the 250 million customers in Europe, and where you could have as simple as Googling, as simple as ordering something on Amazon, for instance, with the consumer who would have a portal who can have access to all the energy services, energy supply, whether they are in the country where he lives or she lives, or whether the other Europeans who could act in his country as well. And that simplicity in accessing the information and ordering things, and having the activation of the services or the services delivered in one click to the customer. That's easy to say, very complex to implement. We have seen it in telecom, telecom have made a lot of interoperability connections between the systems that at the end is extremely simple for any customer in Europe to change things, to order things in telecom services. And we want to come to that stage. So this is the latest work we have been doing. We went through a scouting and funneling of different ideas from the group, Max idea was one of them. And then we came to shape that into a proposal where we have had at least two or more interactions with the European Commission to see how we can execute on that big idea as an ambitious project for Europe. Now, I'm delighted to stop here the introduction about the ATIPS NAT at this stage and go through the presentations from the different panelists. The structure would be the following. So we will have a 10 minutes presentation, particularly with slides from the panelists, followed by like two minutes questions. And then I will open it in the last 20 minutes for a dialogue among all the panelists here with the after hearing everyone presentation, going through the learnings and recommendations, and then a conclusion at the end. I will start by Miguel Ponce de Leon. Miguel will be speaking about the service based architecture for IoT and will be speaking about a European Horizon 2020 project, so no project. And I think this will be at the beginning setting up the scene as well, speaking about highly technology focused project around the IoT that is very much in the context of the DSO distribution system operators or what we call as well DNOs distribution network operators. And then we will go through the other two presentations, one from Etienne Gehan, who works at NG. And finally, Peter Söderström, who works at Battenfall. The experience of NG digitalization is interesting one. And the experience about the digital hop at Battenfall and the strategy to build foundation for digital grid at Battenfall is also an interesting one. So let me start by Miguel Ponce de Leon. The floor is yours. Thank you very much for handing over. So hopefully my audio is coming through for you and all the participants. So thank you very much for taking your time to come and have a chat about the digital grid. I have great pleasure to be part of this particular event. And I'm going to get into my presentation quite, quite rapidly here. Just to give you some context, Sanyo, which is the name of the project, means dream. And you're finding us in a project that has 13 partners across Europe. As was suggested, there are both DNOs, DSOs, so it's called DSOs, also part of this particular project. And very much looking at when we talk about service orientated grid, we talk about, you know, a service platform, a digital service platform in order to be able to also use a network of the future to be able to deliver in the digital grid space. So I'm going to take a few minutes to go through my slides if I go the right direction. The context of this project, it has been running for the last two years. So it's actually been underway for the last two years. So what was reforming the project is the idea of, or this happening of grid resilience. I live in Ireland. I live in the southeast of Ireland, which is out on the western coast of Europe. We get hit by a lot of storms coming in from the Atlantic Ocean. And this has significant impact. You can see up to 5,500 damaged overhead lines. So hundreds of kilometers lines of electricity being distributed across Ireland. That much damaged 385,000 homes and businesses out of electricity. And essentially, 665 million customer minutes lost over storm or failure alone. This is having a significant impact on the grid. And so this was really an initial part of informing the team of coming together to look at, well, how could we essentially work better with putting digital as an element to help us in this. There are other challenges. And I'm not going to have to go into too much detail here that the normal challenges around storage of electricity, about the renewable energy sources and how they are predicted or more predictably added to the grid. And also the mobility of electricity through electric vehicles. We can see in Ireland is certainly having an impact on how ESB networks who are the DSO here in Ireland can essentially manage their network and how it's impacting their network. So the distributed of generated electricity is having a big impact. So complexity is the key thing. The key point here is that it's becoming a much more complex network. And so this set the objectives for our two and a half year project, Sanyo. Looking to enable smart operations on the grid. We need to unlock grid automation to be able to better get better asset utilization, especially when you're looking at bringing renewables into your low voltage and leading voltage areas, especially low voltage, how to better get better utilization and how to build up system awareness so that you can start to become an act upon the accuracy and buying grain data to make secure planning and secure use of your network. These are some of the key goals of the project. And in order to deliver on that, we've had four essential papers to the project and essential ingredients to us. The inclusion of low-cost PMUs and advanced PMUs onto the grid about bringing cloud virtualization. And this will be a major part of the topic when I talk about the architecture that we've used to be able to integrate both cloud and edge cloud into the delivery of a number of services into the grid space. They're to be able to use 5G wireless communications and be able to lab test its use with software defined radios in the grid scenario. And finally, to be able to use machine learning to be able to do data analytics based on deep learning of the network and the data that's coming up from all of these low-cost sensors and devices. And they have to apply these to specific services. So the rest of the presentation is going to take you through a couple of these topics and help tackle them in this scenario project. The first one around 5G. I mean, 5G hasn't been fully standardized yet. There are some elements that are to be built in. But for the Stania project, we really wanted to look at the low latency. The low latency is certainly as applicable to looking at frequency control on the grid. Low latency is essential. Communication platforms, they have support that is needed. Edge cloud, again, is very much an element, an architecture of 5G and how we can reuse that within the grid space. Network slicing, which essentially allows your IoT devices to have a separate network within a network, separated from all the normal multimedia and searching tools that most people will be using on their mobile devices compared to the IoT devices which are taking important data up from the network. And we need to rely on the reliability and security of 5G and how this integrates with our grid solution. What's been really big in Stania as well has been able to take innovative algorithms and monitoring approaches, ones that are in research and essentially sometimes in MATLAB and been able to bring this towards the data that's been provided by ESB networks and been able to apply it and see how it will help the grid. And also been able to show at DSO and DNO how machine learning tools that are existently there can be parsed and used over the large amounts of data that are already being gathered by the DNO and how to give you advanced analytics on this information. Again, been able to visualize key performance indicators have been an essential part of being able to show how Sonio, when using network codes for example interconnecting both from the high voltage to medium voltage and also through your low voltage areas, how to be able to keep visualizations on these KPIs as being a part and part of the project. Here is an example of the architecture that we developed on within the Sonio project and just to highlight that, you know, while it looks simplistic, that's the key part I think of any type of architecture. It's simple to explain but does solve a number of issues and I'm going to start at the top where it comes to visualization. You know, being able to visualize all of this data we found very early on is really important. But we did find that there's many different devices and many different connectivity management elements to the data that comes in. So having an integration bus is an important element of all of this. You can see service one to service n and I will explain these in the follow-up slides where I'll talk about power quality and I'll talk about flip circles. But just to show other parts of the architecture that are important, databases, both from the perspective of having time series data as well as rational and data bases, both are actually needed within the architecture and you need to be able to integrate both. Using your communications, 5G being a key item and then obviously the electrical grid and being able to interconnect to it has been an important part. You do see labels there talking about Docker, you've been able to deploy on the cloud half the state, one of my big points for this presentation is being able to use Docker containers as software system in order to contain what you build and be able to deploy it whether you're using Microsoft Azure or Google Cloud or AWS or any other type of cloud provider. I think having Docker and Docker containers has been one of the big revelations in our deployment. As I suggested earlier, I was going to cover a little bit about the actual services themselves. We broke them out into system-aware services like state estimation, power quality and power control and also moving towards autonomous self-healing. We really wanted to look at this topic and looking at fault, location, isolation and service restoration has been an important part. Being able to look at those storms when they come in, being able to identify where the faults are, being able to send the crews out in a timely manner has been really, really important. I've been able to show how we can build algorithms. Just to show you the field trials within this project is based right across Europe from Ireland where I'm based and we've done a number of field trials there but also in Germany and Aachen. We've also been in Estonia and also in Romania. So we've been reapplying some of the algorithms to be generic enough to be applied across a number of different scenarios. This was also really important for us to show that the architecture could stand up. What we've really found is when you look on the right at the left-hand side of this particular slide deck about the conventional approach, I think most of you would know SCADA systems, would know ADMS systems that are in place and measurement devices and I think the industry has been used to single-vendor product procurement cycles and certainly what we found with building Aachen and dealing with a number of DSOs that it's going to become a multi-vendor environment where service procurement is going to come to the fore. You're going to have to have a digital platform that's going to be able to support that. So there's going to be multiple algorithms and multiple implementations that you need to be able to bring in. There's going to be multiple suppliers and multiple devices and you need to have a slide deck to support this. I'm nearly there with the slides. There's apparently for a second. Although I'm not going to go through each word on this slide, but I think you can see where we view today's distribution grid, where we see the next distribution grid having more accurate sensors, more monitoring data, more conventional algorithms, where the future distribution grid, where they're going to be more open, more distributed. So we have to have a scalable and modular IoT set of platforms and digital platforms in order to support those. Just to say that in Ireland itself, we did test both the flitter, the power quality, and the state estimation services. We reused data that was already available within ESB networks. We deployed some new sensors and were more intelligent about where we deployed those sensors. There were times when it would have been nice for the algorithm to have 200 sensors deployed within 100-mile radius. It just wasn't economically feasible to do so. So we've been able to be more intelligent about how many sensors and where we placed them, but still had the granularity of being able to have the machine learning algorithms still provide us with insights that became a very important item. Okay. I've probably taken enough time with my set of presentations. I'm open for questions and I'm also happy to receive anything in the Q&A section too. Thank you. So don't worry. We had a time because the introductions were short at the beginning. Just questions here on the 5G. Do you believe the use cases you have implemented here on 5G are among today the pilots of the most advanced, for example, in Europe? Absolutely, because look, we've had access to the Ericsson, have been one of the partners in the project, and their E-Node Bs, so the actual devices for 5G, they are fresh out of their R&D labs. So we've been able to gain access to that technology to be able to test what we'd be doing with our digital grid. There's very few or next to know that I know of that have done such tests. Okay. And if you can summarize like one, two, three things, between without 5G, the way from these OT systems, data systems, and so on, and the measurements, how we were collecting the data through which communication channels, and why the 5G is bringing... Yeah. So the big word I'm going to start with the word distributors. And beforehand, everything was centralized. Everything was essentially centralized. Centralized dedicated links, centralized computing systems. What we've shown through Sanyo is that you can reuse the value of the internet. Remember, the value of distributed systems, disparages connected elements, but you're able to bring them together to give you intelligible insights. That for me has been the key thing, and being able to show somebody like ESB networks who are DSO that you can transition towards that world of a distributed digital solution. Because again, they would have had a worry that how can we move towards this type of technological roadmap, this technological architecture. And we've been able to show through Sanyo that you can do this step by step. It doesn't have to be a one big buying solution, that there are ways you can transition different services and different elements of your network to move you towards this architecture. So did you make some calculations about the cost? What's more cost effective in terms of like network operations, implementing the 5G with this distributed control, distributed communication versus what we do today? I couldn't. Now there is a report that we have from some of the partners in place. Now I don't have it to hand, but we were able to show that look, since you're leveraging, we'll say, cloud computing services. So you're not having to have dedicated elements built for you. You have to re-leverage services that are kind of commodity-based services that immediately essentially drives down your cost on certain parts of what you would be spending on before. So again, the key proof for us is that you were able to use these commoditized products, these commoditized services within your digital grid solution. Now we do have some key data around this and we do have a report that gives you, we'll say, deeper analysis on what I've just kind of given you an antidotal overview of. Hey, Mahir, could I ask a question? Sure, of course. Mark McGranigan. Hey, Miguel, this is a very interesting project and yesterday we had a really good presentation from John Hughes from Amarin talking about private LTE applications for managing the distribution grid and that's gaining a lot of support. Matter of fact, ESB Networks is one of the ones really interested in that technology. I know this project kind of focused on 5G, but does private LTE as a communication infrastructure fit as a starting point for implementing some of this service-oriented architecture? It certainly does. I mean, you can see me not in my hand here because, you know, essentially, look, we would have also put in part of this, you know, your cloud infrastructure that's in behind. So the radiations of having private LTE is, again, a way of transitioning you towards then being able to use 5G when it's fully available. Of course, you don't strictly have network slicing as you have to do it in a specific way with private LTE. It's more readily available when you have a true 5G, but, again, you can basically see how it's going to fit into your roadmap in the longer term. But any of the way that we've done the flisser, the power quality, and the status estimation services are still developed in the same way, the platform that you're putting in your cloud environment, it can still be done in the same way. Any other questions? Etienne, Peter, or Omar would like to ask? No, and I'm monitoring the Q&A. No other questions submitted so far, so we can go ahead and continue. Thank you. So the last one from my side is, what is the next step for this? Of course, this is big on the agenda of telecom digitalization, the European Commission, the 5G is really a big thing, and there is a lot of expectations out of it. What is the next step after this Sonya project? So, funny enough, we've just started it, so we just kicked off the follow-on project in April, so the 1st of April for this year, which is about flexibility in the grid and how, once you have better monitoring on the grid, once you have a little bit better control over what's happening on your grid, that you can offer more flexibility and what I mean by flexibility is offering of other renewable sources onto your grid and how you will be able to distribute that. That's a project that we call Edge Flex, and it's been able to broker both the financial side of that as well as the physical distribution of that energy. So, that is essentially where all of our team are now heading towards as a follow-on from this work in Sonya. Perfect, thank you. We'll come back to you then in the last section of the panel when it comes to the discussion across all the panelists. Let me jump now to Etienne Guillain, who works at NG, as a digital innovation officer. He will be talking about the digital transformation strategy at NG. We have seen some announcements about big investments by NG on the digitalization. We have seen a number of projects and so on. We are very much keen to listen to where is NG today and what have been achieved and the next plans. So, Etienne, the floor is yours. Thank you very much. I hope you hear me well. Yes. You hear me? Okay, good. I'm delighted to be here in this panel and thank you for the invitation and for the opportunity. So, maybe a couple of words of context you alluded to Myher about NG. We are essentially an independent power producer and a retailer. We have very, very little network activity in electricity. Of course, in gas we are major there, but today I'm going to speak as a retailer essentially. So, as you said, Myher, NG recognized the need to operate some transition and to go towards the energy transition a couple of years ago and we started by this investing from our centralized power production capabilities, essentially moving out of coal power production, but also many gas power production that were less and less profitable. Overall, we wrote off about 15 billion euros of our assets, which of course showed on our share price heavily, as you can imagine, but at the same time, we also divested and reinvested about the same amount, 15 billion into new assets, in some of which are digital assets, digital platforms, ways to interact with our customers. The fact is that we recognized that we were no longer going to supply all the electricity that our customers were going to need. Some of them would be produced locally, but that didn't mean that we had to disappear from the picture, of course, and we wanted to service our customers with offerings and caring about their new needs, caring about their assets, what they wanted to do with it, what they wanted to exchange, possibly with neighbors or how they wanted to interact with the grid. So with this context, what I tried to do is to show how we thought about all this with a picture showing that it is about technology, technical sophistication, but it is also about business modeling and customer engagement. You have to find the sweet spot between those three dimensions, and I didn't put their regulation because regulation appears everywhere from a grid code to a market organization and so on. So starting with each, starting, sorry, with a technology, I will go in all dimensions showing what we have experimented and the conclusions we reached. Technically, we have many things, and Michael showed that it's still progressing, but basically we have a lot of elements already at our disposal. We have IOTs, we have inverter-based, let's say IOTs with PV panels and batteries potentially. We have communications and some level of local computing or edge computing. The question is what would you like to afford, and we experimented with in a number of places. In some cases, you can have many different devices more or less connected. What you do need to create some local energy management is some active assets like PV or batteries, but not necessarily in every home that you want to manage, but at least you need to have smart meters or sensors that make old analog meter into smart meters, and often you need some local computation capabilities. One important thing is to realize that the installation of all those equipments is not trivial at all, and you usually end up with a typically four-hour setting of additional batteries or additional sensors, and that is a very important feature for the future of the operations. You also have to care about your architecture in terms of data collection and management. You will find different situations, consumers that are only consumers, they have nothing, but they still want to have a special relationship with you through an energy community, for instance, or local energy management scheme. You have consumers that have only PV, others have PVN batteries, and as Miguel mentioned, you do need, I call it, a data access point manager, which is the equivalent of the bus that was mentioned before. And then you can think about various layers, which you can organize into various markets, which can be provided by one or different parties as long as the interfaces are taken care of. Now, we did experiment a number of the technical solutions from the simplest to the most sophisticated. One of the lessons is you can do things very cheap and not too complicated, for instance, providing interfaces with Raspberry Pi-based gateways. That led to, for me, for us anyway, to about typically 7% of data losses. And then you just need to think whether those losses are acceptable for your activity. In our case, it was not necessarily random losses. It was more disconnection for a period of one or two days, and the rest was quite reliable, which means that you lose one or two participants all the time within your community. And that might not be a bad thing. But of course, if you look for providing frequency regulation services, you might want to opt for that simple gateway. You need a much more reliable setup, which typically would cost installed $500, for instance. And then the question is, what kind of business activity, business model, can support that kind of investment? And there are many options again. You can sell or lease the hardware. You can call third-party installers. You can use proprietary or open-source solutions. You can do a lot. What we experimented was a kind of the simplest business model you could think of is simply extending self-consumption and home energy optimization on a single home into a collective self-consumption and providing additional proportional self-consumption. But even in this simplest business model, you end up with a lot of interactions, a lot of contract transactions involved with money or data. The main learning that we got is that, in fact, this simple business model of collective self-consumption allows you to improve the proportion of energy that is consumed locally from 40, typically to 65%. And the additional gain that you obtain by managing collectively all those clients is not necessarily a lot in terms of dollars or euros. And you need adaptation to the grid fee structure, maybe. Or you need to combine that additional value with other value that you create because you have created a community. And so you must move transversely, horizontally into other types of services. And that brings us to the questions of how you engage customers and how you prepare for commercial scalability. And here you have the usual information feedback needs or tricks about gamification. You have to provide trust. But also you have to think about how to scale automatically by providing ways of viral expansion from customer to customer almost without your being involved into any marketing effort, for instance. So social networking is important there. And we experimented with a simple UX or human-machine interface, providing nudges, providing information about its own activity in one single home, and its role into the community, providing some rewards with the points accumulated for good behavior, meaning consumption or production at the right time for the community, which means for the grid as well. The main learning here is the main issue, really, with customers is to explain and present the fair share of value creation. Any doubt about the distribution of value creation ruins your efforts. You can do that using a market mechanism if you want. It's not necessarily always a case, but that allows a sensation of fairness. If you are able, sorry, to educate all your customers with those market mechanisms. One small anecdote is in one of the experiments we allowed the customer to opt out for many reasons, including legal reasons, not forcing anything on any particular customer. And we had again this number, 7% of not available capacity or flexibility capacity in the case. That meant that if you wanted to offer 100 kilowatts or flexibility at any one point, you had to connect with 107 kilowatts, of course, to be sure to provide what you promised to the grid or the market. And that number was reduced to virtually zero if you just include some kind of gaming to the point that when we had a technical glitch and a customer was disconnected, he was calling us, accusing us to not letting him having the chance to participate and win the reward. So you can do a lot with very few features of your platform or software. So all these shows that you can do a lot of things today. You have to adjust for the profit and you have to think a lot about how to transform technical capabilities into real offers that customers will accept. In a nutshell, we are convinced, of course, and we are working in that direction, that local energy management in one way or another will very soon be the norm rather than the exception in any energy supply contract. It will be aggregation or energy community ready. As it has started to be green and renewable already. So with this, I will stop here and welcome your questions. Thank you. Thank you. Another very interesting presentation. I read an announcement from RISCOP and this is about the cooperative energy saying by 2050, 37% of the renewables will come from cooperatives or things that we might call maybe energy communities. And energy community is now a task force that has been set up at the European Commission where there is probably 10 different categories of what we would call energy community. So for instance, there is like people who invest some money and that's it into renewables, which could be for a community. There is also what you are presenting. There is also a models where you could have any customer who say I could be the facilitator of the energy community. If I can give an image with the hotspots, for example, when you have a wireless connection, somebody could say I would like I would become like the hotspot and then you can plug yourself on my on me and then share what he has from presuming, for example, from renewables with the others. And you have players, one of them is a member of the working group called Greencom Networks who have a full platform solution as well for the energy communities. So this is really an interesting topic. Now my understanding from what you said is Ra, the model is much more interesting when it comes to presuming if an individual prosumer joined forces with the others and then having a much stronger profitability or voice being like a hundred prosumers together rather than an individual prosumer. And this is what you tested here with this model. Yes, in a sense, you're right. But I would add that if you only have prosumers, you will run into problems as well because you will have a period of time with access production. So what you do need is a balance community or set of customer acting coherently. I mean, I'm making the distinction between community because it gives the sense of belonging to some organization or even institution and collective behavior that is managed by maybe a third party through, for instance, the platform that you mentioned. Anyway, what you look for is very complementarity between members of that community. If you only have producers, you have a problem. If you only have consumers, you have a problem. You need people to consume during the day when the prosumers that are not at home, not consuming, are nevertheless producing when there is sun at midday. So even you need a mix between residential customers and at least small commercial customers that consume during the day, basically. Yeah. Yeah, this is why the platform I was speaking about should combine different software solutions. For example, one which is for flexibility, another one for billing, another one for IOT and so on and the smart meeting. So the combination of these will manage what you are talking about is while some consumers are not at home and presuming, others should be consuming. Otherwise, you get into a problem of access generation versus the demand. It's an interesting topic. Now, the categories you are talking about are like a building where you have, I don't know, 20 tenants, for example. Is it an agricultural land where you have also farmers, for example, and they built an energy community from solar panels there or wind turbine? So what are the configurations here you looked at? We looked at communities of one type you mentioned, which is the collective investment into a fairly large production asset on the side of a village, for instance. Okay. That's one option. This is quite simple because it's just a common investment and equity sharing. The more complicated option that we tested is the neighbors in one street that only to, let's say, agree on the scheme of redistribution of value created. And there are mathematical models that are totally exact and that are totally non-understandable by people and therefore producing a lot of questions. The simplest way is to slice the day in 15 or 10 minutes interval and see who is producing and who is consuming at any one moment and rewarding people on their marginal effect on the, let's say, equilibrium of the load curve of the total community, for instance. And then the third one is, let's say, the vertical community within the apartment building, for instance, with a single asset on the roof or again on the side of in the street. It is different not for technical reason, it is different for business involvement because there you have usually intermediaries, owner of the apartments and so on. Yeah, the co-property or the syndicate of the company. Social housing. Social housing, facilities managers. Facilities managers, yeah. So, sorry. So, any question from Mark Etienne or Miguel? I think Mark had a question. Yeah, I had a question just in terms of the relationship between aggregators that would be directly interfacing with resources at the customer and community level and how they would coordinate or maybe it's, you know, maybe the retailer is an aggregator, but the relationship between the aggregator and the retailer in offering services to the TSO and DSO and I know that DSO Flexibility Service is kind of new. Mark has a market for that, but how does Anji feel about, you know, the relationship between the retailer and the aggregator in connecting to these distributed resources? There are different roles that must be played and they can be played by either different entities or the same entity can play more than one role. So, that means that interfaces must be set up and quite explicit. We are quite fortunate in a sense that in France where we did some experiments, the regulator forced us to consider that each individual participant to a community had to have the right to contract with its own preferred retailer. Right? So, in other words, I said differently, it's not because you set up an energy community that automatically you become the sole retailer for that community. And that forced us to immediately imagine that we could play a role of community manager, helping the different customers according to their technical sophistication or market readiness and still provide information to the DSO and so that in this case in France the DSO can relate what happened within the community to the different retailers so that the difference between local production and the rest of the electricity need could be correctly calculated. In a sense, in France, this kind of community is essentially taking the readings from the meters and correcting those readings according to any kind of rule that the energy community has agreed collectively to set up to take into account the activity of its members. And therefore, each member of the community only pays to the retailer what is left to be paid. Overall, of course, the Kirka flows are always the same. The decisions of value sharing within the community doesn't change what enters or leaves the community in terms of energy. So, the retailers are not affected. To come back to your question, we believe that the different actors need to take into account the different potential interactions and interfaces and the very important interface, of course, is with the DSO and that might have been one of my questions to Mike Miguel is about what kind of provision in terms of API interfaces, for instance. His project is imagined to have support from community managers or aggregators or market operators if the local exchanges are managed through market mechanisms. So, Etienne, I might come in for a second since it is a question. Funny to say that because, yes, as part of the Sonio project, we have to find a set of APIs to allow to gain access to two aggregators and actors like that into the types of data information in order to do this. So, this is something we have considered through the Sonio project and are looking to reuse through our Edge Flex project. And this is part of the code as well where you have this flexibility marketplace where the DSOs and DSOs plug in and where through this flexibility marketplace you connect as well to the aggregators. So, we and G are not looking at becoming the sole player around communities. We might have to adjust, for instance, if open source, very good open source, I would say, technology stacks emerge that becomes adopted for one reason or another by many users. In that case, we would need to become a helper around those solutions and stop maybe our own developments into a platform supplier. It doesn't matter. I mean, what we want is to service our customers and help them make the most of their assets. Look, Etienne, I need to jump to Peter, but I think that was very interesting. I will come back to you. Maybe I will leave you with a question. You can take the time to think about it when we will come at the end of the dialogue. Is that now an established business? Is that a big business for NG or not? But reserve your answer to later on. Peter, then you are the head of the digital hub at Buttonfile and I guess you will be speaking about the strategy and demonstrations that are providing the foundation for the digital grid at Buttonfile. So you have like 10 minutes and hopefully we can still have time at the end for a dialogue. Thank you very much for the introduction. Yes, that's correct. I'm working for Buttonfile, specifically at Buttonfile distribution. Sorry about that. Just to give a heads up on Buttonfile, it's a company acting in the northern Europe of all parts of the value chain, but Buttonfile distribution in specific is operating a grid in the northern half part of Sweden basically and in Berlin. We also have some network in the UK. So I won't spend more time about Buttonfile, but this really sets the scene and of course also affecting the challenges we are facing when addressing the future of parts of the science. We'll go into our strategy, I will talk about the different parts of our digitalization distribution. But before going into what we are doing, I would like to mention a few words about something what's different now than before. The evolution of our energy system has always happened and it is nothing new. There has been big challenges before. There is new areas for certain, but the major difference for now that we haven't seen before is the difference of time. Everything goes so much quicker right now. We are seeing demands from society going in a scene that we haven't seen before. We are building new houses in just a few months, establishing solar farms in 30 days in a space that's not really seen before as part of the evolution of the energy system. Before going into a lot of what we are still struggling with of course, and there is no clear solution, I do think we need to keep in mind that our real challenge here is time. We need to match the need from society in the evolution of our energy system, electrical system, meeting the sustainable goals, and it needs to be done in a pace that we haven't seen before. Now looking, I see something has happened with the slides. I don't know why the slide is looking as they are. Formatting may be PC Macintosh or something like that. Yeah, I'll move on. I guess most of the text is still there. Basically our strategy for addressing the digitalization of our DSO is oriented around four topics. It's the digital grid. We are seeing things like digitalization of our substations. We have the low voltage monitoring going all the way to the delivery point for customers. We have IoT systems. Just to give an example, we have the eyes build up around our lines. In just 10 years, that's drastically changed due to climate change. It puts a completely new demand on understanding how the eyes build up. We have microgrids. Of course, you can see that the energy communities are starting to build up. But independent of that, we are seeing connected microgrids being increased at the drastic pace. And the change behavior of these microgrids is, of course, changing the needs from the overall grid. Then we have the digital worker. The work environment for our office workers, but also field workers, is drastically changing. We are getting the support of, for instance, drones. We have virtual reality or augmented reality that helps us get efficiency in our processes to a degree that we haven't seen before. In the data analysis that support our workers are, of course, AI support increasing. You can see that we are basically moving away from an old-fashioned way of doing maintenance to go into not only condition-based maintenance, but also preventive maintenance, which is a completely new level of doing maintenance or being proactive. And we have the tool of a digital twin. Even if that is just in its early phases, I strongly think that the possibility to actually create a digital version of your grid is a great enormous possibility to understand what's really happening in the future demand. So apart from the digital grid and the digital worker, we can look also at the digital customer. In that area, we can see new demands of our new need for tariffs. Of course, the tariff setting is different from country to country. In Sweden, we have the luxury of actually being able to set our tariffs ourselves. There is, of course, some boundaries, but basically, we can decide the structure and the layout of the tariffs by ourselves. I know that's not a luxury in all countries, but it, of course, gives possibilities to the meeting. We have had, for 15 years, smart features roll out 100% in Sweden. We have the earlier first generation of the meters. We have our values. We are now going down to the possibility of five minutes' values through the rollout just happening as we speak. But we are seeing that it's not enough to meet in the future. For certain customers, not even five minutes' values is really sufficient to be able to support the needs in the UN system. So we are talking about real-time matters, meters, not for all customers, but there is something we need to manage. The electrification of the transport sector is another one. Electrical roads is a great challenge. The indicative demands are huge. And, of course, then we have the direct customer dialogue, talking about digital channels, voice bots, using your Siri or Google Home or whatever to actually communicate with your utility verbally. The final part, and, of course, that needs to be digital as well, so digital dis-business models. Maybe that is not that easy to understand, but it can translate into flexibility. Here we are seeing the increase of use of bilateral contracts for demand and generation. We are seeing the test. We are running a couple of tests in different parts in Sweden with local markets, local capacity markets, services from energy stores. It's increasing in a rapid speed. But we can also see flexibility created from a change view of risk or preferably reduced risk in operating your network. That can actually create dynamics. And finally, to some degree, there is also the possibility to use local generation. So all of this is pointing at areas where our strategy takes us, where we are doing, in some part, innovation and some part, implementation at the moment. Just to give two examples, we can see on the left-hand side, the low-voltage monitoring. I think being well in the way of the 21st century, our customers are really expecting to know what service we are delivering to them. They don't expect that they need to call off if they have problems. For six years now, we have put people into operations using, of course, the smart meters to get the knowledge all the way to the delivery points at the customer size. And what we are using is the event coming. And we are talking about between one to five events on average per customer per day. So we are talking about millions of events that actually our operation center should act upon. Of course, that requires some analytics and some AI in order to be manageable. But the benefits are quite huge. The trust is increasing. We have an improved, of course, quality of service. We're finding problems faster. And there is a safety aspect also you can identify by moving. Another example is forest clearing. We have a lot of overhead lines that requires clearing of forests around the line. And this is just an example of automating an existing process where we are today is selecting 2D, 3D, laser scanning and camera graphics of aura lines. And we're not doing it yearly. Every part of the line, but we are doing 13,000 kilometers per year in data collection. And this covers areas with 100,000 real estate owners that need to be interacted with. And in total, we also give a rough picture. We're talking about around 100. This is something, of course, that you need a digitalization of your process in order to be efficient. And now I know the pictures are very small. But it's just an example of our system where the AI is automatically identifying trees that need to be cut. It even outlines the roots, the fieldwork it needs to take to get to each individual tree in the most efficient way and how we can dispose of cut-down trees. This is maybe the simple parts of the digitalization. You take something you already are doing and you do it more efficiently by digitalization. You're slightly extending it in some way. The more new part of this is going into the flexibility, which is really a big game changer, I would say, for the energy system and something that is going to be a real struggle for different stakeholders. Not only for the DSO, but all the stakeholders, including the customers, making this not only possible to do in single pilot, so demonstration, but also on a wide scale. But before going into one of the examples showing, okay, where are the challenges, about flexibility, because that's not uncomplicated at all. I think when most people are referring to flexibility, they are really thinking about the market-based flexibility. They are talking about demand response programs. They are talking about somebody buying services from an energy storage, for instance. It's market-based. It's something there is in market actors that need or possibly also a market platform that you're acting with. Of course, you can also include bilateral contracts instead of a more open market. But I think that's what people mean when it comes to flexibility. But for the DSO, this is not the only flexibility that's available and being chosen from. We have the two other types of flexibility that has the equal value for the DSO, but it's not connected to the market. One is the connection-based flexibility, and here we are talking about connection agreements, where there could be limitations in such an environment. That makes completely sense. But there is also the part of the tariff. The tariff in itself, of course, there are some limitations. You can't change it, at least not the structure every day or every month. But still, it's a slow-moving, maybe, but it's still a type of flexibility that is not going really to the market, but it's still affecting the available flexibility. And the final part of flexibility is not involving the customer at all, or not involving the market at all. And that's more the technical flexibility. And a very typical example of technical flexibility is you change the risk level, you operate your network. It's not an on-off. It's not that you go to a certain level, and then you run it on that level, and then it's something just likely a small kilowatt is added, everything breaks down. We are operating the grid, of course, with a safety margin. But with increased knowledge about our network, with increased sensors and data collection and calculations around this, we can, of course, give in certain scenarios run the network closer to its safety margin. And that is an available flexibility that's dependent on the risk appetite of you as well. It's not involving the customer. So all of this needs to keep in mind when talking about flexibility. Not everything is market-based flexibility. Now, the next slide is really the best slide I've ever seen. The best slide ever created. I haven't done it, so I can't take any credit for it, but I'm still reusing it because I think it's so great showing the challenges we have with flexibility. And that's not only market-based flexibility, but all types of flexibility that we are currently facing. This is a slide coming from the EU project. Now, I see the logo for the project has disappeared, but it's from the EU project coordinate and interesting project that's looking into enabling local flexibility platform. But the interesting part about this is that when you're looking at the marketplace as such, where you actually are trading the flexibility, this is the circle in the middle of this slide. And, of course, that's important that you need to have an interface to it and it shouldn't be marginalized. It is important, but it's just a small part because what needs to happen in order for flexibility to be usable before getting to a platform or a connection agreement or a technical flexibility that you want to use. Well, there's a lot of arrows going into that circle in the middle. Looking on the left-hand side, you can see it from the DSO perspective. You need to start to take, can we do flexibility planning? If at all, it's the skills in your department that you actually can't plan your network with flexibility as an integral part. I can't speak for every company, but I can say for us, it's a real struggle to get that into something that's usable and something that you can risk mitigate. Then you, of course, also have the operating flexibility. When it's built, you operate your network with flexibility. This is a new way of operating your network, which is custom to just work with your existing network, the grid, the copper network to do. But in order for them to actually be successful, we come to one of the most important things that's happening right now. That's the possibility to do grid forecasting. How can you forecast your grid? Here, there is a lot of projects going on throughout the world, throughout Europe. There is a lot of projects or companies involved into this, but you need to get something practically workable in order to do the first work. Finally, when you go into, now you're getting closer to the market, you need, of course, the coordination without the network company here and the DSO specificity. That's coming from the left-hand side. From the right-hand side, it's basically the same, but now we're coming in from the customer side. It's not given that the customer knows its own processes when we are talking about or industry knows their own processes for your company. How flexible are you? This is the journey they need to take. They need to understand what type of flexibility product is really working for us. What can we do? How can we commit with a lower risk? Forecasting is also a part of them. They need to forecast. How can we optimize our key processes in order to deliver on why we actually exist, but we also see, can we make money out of being flexible? Finally, going closer to this, when this is actually happening, when we have the product defined, when we know the forecast, we also, of course, need the way that we can verify that and we're coming in. Peter, we have for the whole session like five minutes left. I don't know Mark if we can take more time beyond the six 30 here, CAT, but I would like to keep some time for also some questions and maybe a bit of dialogue. For sure, and a good part is that was my last slide. Okay, thank you. Let me maybe ask the question and then it's going to be an open question to everyone. What is the driver behind the three digital example projects that you are presenting here? Where the starting point was? Was it happening in the grid team? Was it coming like a top down? Was there like a pain point? Or did we need for every one of these cases to justify the business case? Of course, the first one was in Ireland, was the European funded project Horizon 2020 and so on. But still, you need also to explain why we need to do this project. So maybe Peter, you can. Yeah, I can start. I think, of course, each of the digital projects shows up with different drivers and from different origins. But I think overarching for this is that you can always trace it back that we have a society and customers that that is moving in an increased speed with new demands. And I think it's it's deciding and customers that's really putting new demands on the electricity and the energy business as such. So and for me, this is this feels good. It's really those that that really should use the system we are creating that that is actually putting the demand and say, okay, you need to improve, you need to pace up. I have a data center. I want to connect it. I can't wait three years to get that connected. I need to be able to connect it in much quicker time. So I think I think it's society and customers that's really driving this when you're looking real. Thanks. What about Miguel and Well, funny enough, I'm going to agree with Peter. I mean, very much in agreement with what he's just mentioned. I mean, obviously, we had a practical driver again, when it came to the complexity of the network and to be able to look at it. But again, just looking at how you can if you get better visibility of the grid, what will that mean about flexibility? And then the social and the the market forces that are now bringing that onto the market and DSOs having to be quicker in that responsive. We need to open up their data just to say, I saw there was a couple of questions in the Q and a section asking about how open was the data available, say for Sonia, or how open will others outside of the utility themselves? And certainly, that's what we're also driving to the because you have to be able to provide that data to other people to be able to deliver on all the services in this flexible world. So yeah, that's certainly what we found as we've transitioned through the years and through the projects. In our case, being more retailer than as I said, DSO, it's quite simple. It's the differentiation with respect to other retailers. The old digitalization allows us to provide services to smaller and smaller customers that we used to provide already for many years to big customers, meaning, for instance, energy services aiming at lowering energy bills and energy saving commitments and things like that. So that starts with home energy management and energy optimization. And very quickly, you try to grab a few more percent of efficiency, I mean, in voice efficiency. And therefore, you start suggesting to group people together and create communities because it's even more efficient and so on. So yeah, thank you. So another question. Every is like a global organization. We might have people as well in the U.S. attending the call. What do you believe in these topics that were presented like local energy community, the model, which is interesting for retailers, the flexibility definition of all the projects that button file is doing, smart mentoring and so on, and also the 5G and the grid project. Are we more advanced than in the U.S.? We have best practices here. Are they replicable as well to the U.S.? For example, when it comes to hurricane and then predicting the storms and so on with analytic solution. Of course, if you go to Florida, next era has also a lot of these needs. So tell me, who is more advanced than others and who can learn from others between Europe and the U.S.? I think we all can learn from each other into this. I think it's dangerous to say that on general, that's one part of the word and the other, because I don't think it's true. I think for certain parts, there is maybe Europe or even certain countries in Europe that is really ahead the rest of the word. I'm pretty sure and certain that there's other areas where the U.S. is ahead. So I don't think you can speak generally. Okay. Are you cooperating with U.S. utilities, for example, on some topics? There is an exchange, yes, for certain topics. Yes, and I think that that's the way you need to deal, because in many cases, there is different priorities. There are different parts that are more important. For us, for instance, it's the weather affecting our overhead lines. That's really something to be struggling. But within the Vasanfallas company, going to Berlin, where we often see networks where we basically do not have any overhead lines at all, this is for sure not a major challenge, and their priorities are different. So I think that should be confident. Yeah, so definitely the vegetation, it's an interesting use case everywhere, what you have done here at Vasanfallas for like 13,000 kilometers. That's interesting in Scandinavian countries, interesting in Portugal, in Greece, in Italy, also a lot of places, a lot of states in the United States. Miguel? Yeah, again, there was just one additional point, but there is a question in there about the fact that market settlement is happening days, if not weeks afterwards. And again, a key thing for me in all of this in sharing the data and open data is having access to real-time data, essentially. Given the amount of renewables that we're going to be bringing in as well. We are also collaborating with people in the US around this. So again, maybe just to address maybe your question, I think being able to host these webinars and being able to share in this way and being able to share across the Atlantic between both sides, I think we'll both learn from the progress that we can make towards this digital grid. Okay, and Etienne? Well, we are present in the United States. We have a commercial operation there. We, for instance, manage university campuses, which we consider as a local energy community because they are various buildings with various functions and different departments most of the time. So they can be considered individual customers anyway. So the regulation is different, but we do microgrid management and the energy community over there as well. Yeah, well, I can say, for example, on the standardization of the smart grids, there was a good cooperation between Samsung, like Etsy, who were working on the Monday for 90, which is the standardization of the smart grids in Europe closely with NIST. NIST were part of the committee, for example, and permanently exchanging the standardization experience in the US with what was done in Europe. And definitely, there's a lot of global solutions when it comes to SCADA and so on that are being implemented worldwide as well with some tailoring here and there to consider the market structure, for instance. And of course, the flexibility solutions. I mean, NL, for example, has acquired an American company when it comes to demand response management and flexibility as well. So there is, of course, a space for cooperation, particularly in the digitalization of the grids. And honestly, I learned a lot today as well from these presentations. There was not overlap. These were different topics and involving different stakeholders as well. And I think the audience who attended the session today can reach out to you. We have your contacts here and continue. There is space for cooperation or for questions. Continue directly. From myself, thank you again for EFRI and with the help of Stanford here to organize this session. It was very interesting for me and I hope it was also interesting for the audience. Maybe I give it to Mark and Omar to close the session. Thanks a lot, Meher. Three great perspectives, very interesting to me personally, and I'm sure everyone got a lot out of it. So thanks for that. And I think we're going to make this a continuous dialogue with an ongoing series of discussions. So I hope we have an opportunity to continue the collaboration across the ocean as you kind of headed us down that path, Meher. Thanks a lot. Thank you. And with that, let's close a fantastic session. Thank you, Meher, Etienne, Miguel, Peter. We will post this webinar recording and presentation once it is set up and we will send information to all the registrants, as we said earlier. And if you can, we have our third of this week's webinar series tomorrow at the same start time. So we invite you to join us for that and our continued webinar series into the summer. So on behalf of the Stanford Pits and Watts Initiative, Epri, our esteemed moderator and distinguished panelists, thank you so much. Have a great rest of the day. Thank you. And stay safe. Thank you, everyone. Thank you. Thank you very much. Thank you.