Catalyst with Shayle Kann - The early days of AI on the grid

Episode Date: March 8, 2024

The first wave of digital grid infrastructure in the U.S. didn’t quite deliver on its promises. More than 100 million smart meters have rolled out across the country, buoyed initially by billions in federal funding. But instead of using them for exciting things like time-of-use pricing and automated demand response, utilities used them for more mundane things like automated billing, according to a whitepaper from Guidehouse.  Could the new wave of AI-based grid tech be different? In this episode, Shayle talks to David Groarke, managing director at the energy consultancy Indigo Advisory Group, who co-authored a forthcoming Latitude Intelligence report on utilities and AI. David says that AI is showing promise so far. Unlike the first wave of hardware-focused advanced-metering infrastructure, AI leans heavily on relatively cheap software and data. He also says that AI’s capabilities are advancing quickly (“doing pressups” as the Irish say) by improving algorithms, handling more tasks, and improving efficiency.  David and Shayle cover use-cases and other topics like: Wildfire management, using data from cameras, lidar, and satellites Customer propensity modeling, including detecting EVs to aid with infrastructure planning Automated and personalized communication with customers Predictive maintenance of substations and other grid infrastructure, using data from, for example, computer vision to detect corrosion and reduce downtime Optimizing transmission capacity by moving from static ratings of transmission lines to real-time ratings Whether incumbents or startups are leading the development of these AI-based solutions David’s take on whether AI’s impact on utilities will be revolutionary or incremental Recommended Resources: Latitude: Welcome to the smart meter’s second act Latitude: AI is simplifying complex decisions for utilities Latitude: Seven ways utilities are exploring AI for the grid Latitude: Could AI-fueled weather forecasts boost renewable energy production? Catalyst is supported by Antenna Group. For 25 years, Antenna has partnered with leading clean-economy innovators to build their brands and accelerate business growth. If you’re a startup, investor, enterprise or innovation ecosystem that’s creating positive change, Antenna is ready to power your impact. Visit antennagroup.com to learn more. Catalyst is brought to you by Atmos Financial. Atmos is revolutionizing finance by leveraging your deposits to exclusively fund decarbonization solutions, like residential solar and electrification. FDIC-insured with market-leading savings rates, cash-back checking, and zero fees. Get an account in minutes at joinatmos.com.

Latitude Media, podcast at the frontier of climate technology. I'm Shail Khan, and this is Catalyst. In that last round of investment, it was generally large platforms. It was all communications infrastructure and hardware. So I don't think the original investments played out, but I think that is exactly why AI is interesting in the sector right now. you know, low-cost solutions with high OI's is what the name of the game of the sector is now. All right, time to gear up as the AI hype and or reality wave crashes through every sector of climate tech.

This week, utilities. When utilities need flexible capacity they can count on, they turn to Energy Hub. Energy Hub works with more than 170 utilities, coordinating over 2.5 million devices to manage 3.4 gigawatts of flexibility, built for the moments when utilities can't afford uncertainty. Energy Hub builds and operates virtual power plants that utilities actually stake their grid planning on, coordinating EVs, batteries, thermostats, and more through a single platform built for utility scale.

Predictive, verifiable, and designed to perform when it counts. Learn more at energy hub.com. Trillions of dollars are flowing into clean and critical infrastructure, but those investments aren't driven by technology alone. They're shaped by markets, by policy, by capital, and by the institutions that connect them. I'm Alfred Johnson, CEO of Crux, and host of a brand new podcast, Critical Capital.

Each episode, I talk with people deploying capital, shaping policy and building the clean economy. Tune in as we unpack how progress is actually made. Listen to Critical Capital on Spotify, Apple, or wherever you get your podcasts. I'm Shail Khan. I invest in revolutionary climate technologies at Energy Impact Partners.

Welcome. Well, you already know this, I suspect, but let me just confirm for you what is happening and has been happening inside the boardroom of, I think, basically every large company in the world, which is at some point the company leadership in the board ask and attempt to answer the question of what AI will mean for their business. Is it a threat? Is it an opportunity? What should we do about it?

And I think in some sectors it's pretty obvious that AI is going to transform big businesses, like if you're in the legal profession, for example. in others, it sure seems like AI should play a big role, but it's less obvious exactly how and when. And that's how I would categorize AI for utilities in the power sector today. They operate an extremely complex system that is full of data that could be leveraged for various versions of optimization, cost reduction, decarbonization, efficiency, reliability, all the things that you want. So there probably should be something big there. But as we all know, everything in electricity is more complicated than you think it should be. That has, in fact, been my primary learning after having spent a good part of the last 17 or so

years thinking about it. But there are things happening here, and it could be transformative. You just have to look closely. So let's see what's what. David Gork from Indigo Advisory Group has been working with utilities on AI since before it was cool. So I brought David on to kind of run through the state of affairs with regard to AI and the power. sector today. Here's David. David, welcome. Thank you, Shale. Let's talk about AI in the electricity sector. I haven't talked enough about AI probably on this podcast today, so we're going to make up for it today by using the term like three or four hundred times in the course of 45 minute conversation. We're doing our best. Yeah. It's, you know, it's a topic that everybody's talking about in every

sector. So let's start with like why this one. And maybe talk a little bit also about why maybe not this one. But starting with why this one. Like why is the electricity sector, or electricity and gas, I suppose, the world of utilities. Why is it an especially interesting one with regard to the possibility of new applications using AI? It's a good question. And it really is kind of the biggest question that folks ask is, how does AI apply to the electricity sector?

It's maybe not something you equate immediately, right? But I think the thing to say, Chelle, is that there's been a ton of... investment, right, in IT infrastructure and OT infrastructure and, you know, all of these smart meters, we're at over 70% smart meters in the US. There's all this, you know, digital infrastructure, and this is from years of vendors and rate cases and so on. So there's an appetite right now to kind of, you know, prove out some of those promises from over the years and, you know, utilize all of this data and whatnot.

So I think that's interesting. I think what's more interesting is the actual problems that AI is trying to solve in the sector. And, you know, if you kind of stand back and you look at, you know, what the power sector is trying to do and take a leading role in decarbonization, and you think, right, well, if AI is supposedly, is a set of computational processes that can perform a task at human intelligence, I mean, the scope is actually enormous. in that regard. So AI has been put to use on various parts of the grid across the whole value chain. There's a whole set of use cases, but primarily it's been pointed at, you know, a set of problems. And it's using all of that data that's been kind of gathered over the last number of years.

And, you know, when we look at use cases, they generally relate to top line issues in the sector. And that's where the activity is. So I definitely want to run through some actual use cases. I think that's often the thing that's lacking in these conversations. But before we get to it, I do think there's an interesting point in there, which is you and I together lived the world of smart grid, you know, 10, 15 years ago, whatever that was. And that was pre this AI wave.

We were talking about quote-unquote AI a little bit at the time, but I don't think we meant the same thing that we mean today. And yet we were doing all these, quote, smart grid investments nonetheless. You know, is there a case to be made that like all those investments, as you said, smart meters at 70% penetration now, all the IT and OT infrastructure, all the DER management systems, all these different things that got implemented

to varying degrees, that it turns out now we have the capability to leverage them where we didn't before, or at least leverage them better where we didn't before, or is it more like, actually, we're using them for exactly the things that we were intending to use them for, and now it just happens,

we have a new set of tools that we could do more with. It's a really good question, right? I think the short answer is that a lot of the promises of, let's say, the first wave of digital infrastructure didn't really prove out in the US, right? I mean, just the cost of infrastructure needed to deliver power is nearly equal to the cost of generating power itself, right? Right. If the point was to reduce T&D costs, we've failed. We've failed. And not only that, you know, OPEX costs are up about, you know, 14% a year.

and there's been a host of other challenges that the sector is going through. But I think this proving ground is a long time coming. What's happened, though, I think is that AI has been doing press-ups in the background, right? And that's the interesting thing, right? So algorithms have got better, right? They can do more tasks. They're more efficient. We have all that training data, right?

That first round of investment. And that's like multimodal data, right? It's energy, electrical, customer, weather, and gases, visual text, a whole host of data. So that training data, those advances in the algorithms. And I think the solution decline, cost decline, where, you know, vendors can leverage open source tools like TensorFlow or Pai Torch and develop applications kind of quickly. I think the cost to innovate has come down. Whereas let's say in that last round of investment, it was generally large platforms.

It was all communications infrastructure and hardware. So I don't think the original investments played out, but I think that is exactly why AI is interesting in the sector right now. You know, low-cost solutions with high OIs is what the name of the game in the sector is now. I'll note that you said AI has been doing press-ups in the background, and I just learned that press-ups is Irish for push-ups, or maybe push-ups are American for press-ups? Let's go with the latter, yeah. Yeah, push-ups, press-ups, yeah, pull-ups even, right?

But it's definitely been working out. We can all agree on pull-ups, yeah. All right, so I think it's probably fairly self-evident like why there's some opportunity. We have to define what that opportunity is, but some opportunity for AI in the power sector. It's an enormously complex system, with an enormous amount of data and incredibly challenging optimizations alongside that.

And so there must be opportunities somewhere in there. But before we get into those use cases, let's talk about why it's actually going to be tough. And in some ways, why that first wave, as you said, hasn't played out as expected. Like, why is the electricity sector a hard one? Why is it not the first sector? Like, there are lots of sectors wherein AI is completely transformative already today. I think of the legal profession, for example. Why is electricity probably not one of those?

Right, yeah. I think it owes the primary reason why we're not on a path towards full automation and there's been a spade of use cases launched. Just modeling grid physics accurately is incredibly difficult, right, for an algorithm. So, you know, solutions on power flow obviously need to take into account, you know, Ohm's law and Kirchhoff's law and accurately predict or make real-time decisions. So there's a limit to where AIs can be deployed currently for real reasons. I think just the complexity of the system, if you look at a Sankey of any utility, right,

and you've got the data sources and you've got the data management platforms and the visualization platforms, they are some of the most complex diagrams that you could witness. So it's not an easy operating environment for solution providers or for solutions to be embedded. And look, I think there's the typical industry issues around just rate-based new technologies, right? But more broadly, just for some of these use cases, the data's not available, Shale. It's not frequent enough, right? So we don't have the millisecond granularity that's needed for some of these use cases. So that makes it all a little bit difficult.

I think it's also, you know, there's a torturous sale cycles for startups, which can be off-putting after a couple of years. I mean, people tend to give up. Torturous is a literal word in some cases. Right. Yeah. And that's not helpful to innovation or to start-ups. So what you're seeing is incremental change by kind of the grid giants over time. And, you know, a few use cases and vendors kind of making it through the ranks.

But I think all of that, I think the complexity, I think the cybersecurity components, I think they're just the grid physics. I think all of that makes it complex. And I think, you know, when we talk about use cases, you'll see that some of the use cases play out across areas that don't touch on those grid physics. Yeah, I think it's this like, you know, enormous opportunity meets incredibly high barrier to entry problem, that basically everything in energy tends to fit. face. Right. And just to add to that complexity, it's also, you know, there's so many problems

within the sector around folks retiring. So, you know, 50% of the workforce is going to retire in the next 10 years. You can see that happening right now. So you have this kind of field technicians are leaving, right? And, you know, there's other priorities, right? You know, building out transmission infrastructure, different capital programs and so on. So, you know, utilities are having to balance kind of their IT investments with some very weighty challenges that they've been given. So there's a juggling act going on too. Let's talk for just one minute about what we actually mean by AI. I think the term gets thrown around a lot.

It has gotten thrown around a lot. And it means different things to different people. You know, there's the LLM, chat GPT version of it and all this generative AI is a new phenomenon. but AI sort of interspersed with machine learning and other tools. It's all over the place. As you think about it specifically in the power sector, what are the subcategories of AI that we should be thinking most about? Yeah, that's a good question.

And that's the very start of a conversation with a utility or with any stakeholder in this industry. Defining AI is, you know, it's an interesting question in terms of how you frame it. I think one way that we've been looking at it is defining it through a set of capability. So you can understand it from a business perspective, right? So looking at machine learning and predictive analytics, right? So these are algorithms that detect patterns or can make decisions. That's one set of solutions.

Computer vision. So these may be cameras that interpret or understand the visual world, right? Infrared cameras and so on. Another set. Natural language processing, interpreting text. voice, human language, robotics, anything that's doing a task in a robotic nature that a human would do, right? So perimeter security by a robot would be classified as AI. I think, you know, areas like digital twins, so these are, you know, virtual replicas that move beyond things like anomaly detection into scenarios and decision making.

You mentioned LLMs. We're seeing some application of LNMs. across the industry, but usually at the enterprise level, around regulatory documents, around, you know, generating text or guidance, less on the actual power sector side of the operation itself itself. I think we look at distributed AI shale, which is, you know, embedding that intelligence at the edge. So in the sector, remote substations, that mightn't have strong communications with HQ. You know, that's a big component.

I think there's less mature parts of AI in the sector too, right? There's an emerging area, I think, that the sector's really interested in called Explainable AI, which is really explaining how a decision was made. I think that's really important from a regulatory perspective as we move towards automation. But really, it's just a set of capabilities. And these capabilities are embedded in existing applications, right? They're in new applications. And these are the things that have been doing, the pull-ups jail, right?

These have got better, they've got cheaper. Utilities have got more comfortable with them, and vendors are using them more and more. Virtual power plants are becoming a reliable way for utilities to manage capacity, but enrolling devices is just the start. What really matters is confidence, knowing those resources will perform when dispatched and being able to prove it from the control room to the living room.

Energy Hub's platform handles the full picture, from near real-time forecasting, locational dispatch, and the kind of rigorous verification that holds up when regulators, grid operators, or leadership ask, did it deliver? Easy enrollment creates momentum,

proven performance builds trust. That's why more than 170 utilities rely on Energy Hub to manage over 2.5 million devices delivering 3.4 gigawatts of flexible capacity. See what that looks like at Energyhub.com.

We're living through a profound economic shift, and energy sits at the center of all of it. Trillions of dollars are flowing into power plants, transmission lines, battery factories, data centers, but the future of energy isn't shaped by technology alone. It's shaped by markets, by policy, by capital, and by the institutions that connect them. I'm Alfred Johnson, CEO of Crux,

the capital platform for the clean economy. Join me for my brand new show, Critical Capital, as I talk with people deploying capital, shaping policy and building projects. Together, we unpack how risk is priced, how incentives are structured, and how progress is actually made. Listen to Critical Capital on Spotify, Apple, or wherever you get your podcasts. Okay, so you started to get into some applications, but I asked you to pick like three actual applications, real use cases of AI in the power sector that you think are real and immediate and interesting. So let's talk about those. Pick your number one.

Okay, well, I think if, to look at something contained, right, and something that's really new, that's it, been in rate cases in the last 12 months across the country. And for good reason, right? So wildfire is obviously a frequent kind of event now for utilities, particularly in the West Coast. And, you know, PG&E in 2002, 22, had a 1.3 billion dollar rate case, right, for, um, for the wildfires that happened that year, over a three-year period. And so anything that plays into that space is going to be a welcome event for utilities, right? Anything that helps vegetation management, which utilities spend billions a year on collectively, is interesting.

So AI for wildfire management, right? Very discreet area, and we're seeing a lot of deployments in that. So what you have there, right? You've got new data, right? So you've got cameras, right, that are in fixed positions. and they're monitoring images of vegetation, right?

You've got drones that may be taking images of density of growth and so on. You've got historical GIS data. You've got LIDAR, which is obviously treaty modelling. So you have these new sensors and data has been deployed, and you have external data that's used in the mix, like satellite imagery, weather data. And you combine all this data together. You know, when you're collecting these high-resolution images and you're predicting the vector of growth to vegetation or you're seeing changes over time or positioning of vegetation in relation to infrastructure, you can do some pretty cool, you know, predictive analytics.

And I think, you know, what we're seeing there is that, you know, you can better deploy field crew, your existing maintenance personnel and so on. And that sees immediate results. So I think why that's interesting just from an AI perspective is it's a really discrete use case. You've got new sensors. You've got better algorithms. It doesn't really interfere with core operations. I mean, it's tied to core operations. But I see why the discreteness of it is valuable here because you can imagine saying, okay, well, this isn't like, you know, grid reliability short of the wildfire itself, impacting grid reliability isn't going to be affected here.

Yeah, and it's also got like a broader benefit. So the PG&E rate case, so what they ended up building was this fire potential index with a whole series of partners. So they used all that data I mentioned, right? And they used 30 years of historical data. And they're sampling that several times a day. And that gives them a pretty good, not real time, but, you know, a pretty accurate insight into potential threats. And it's relatively low cost, right? These are SaaS platforms with some sensors running through the millions over years of agreement.

So, you know, that's where the high value comes from shale. But yeah, discrete. Doesn't touch operations, not overly political and can roll it out pretty quickly. All right. So that's discrete new data, new data sources, sort of fits the mold for like things that could apply new technology or in this case, AI, various stripes quickly. Right. What's number two?

Number two, I mean, if we look at the customer side of things, which we'll get to, We'll get to some of use cases on the power flow side, but customer propensity modeling, things like EV detection, what's really been interesting is like over the last 10 or 15 years, utilities have really waxed lyrical about owning the customer relationship and monetizing all of this AMI data and, you know, it's a fixture as a panel at events and so on. I think that's actually happening, right?

So in terms of how propensity models working for utilities at the moment, So they have the AMI infrastructure, we said there's over 70% deployed. They have obviously their customer information systems, all the records of a customer and their interactions and so on over time. And then they have external data. So this would be data like socioeconomic data or weather data or property information, size and type. And what a utility can do with that is some interesting propositions for customers. So you can perform non-intrusive load monitoring, which basically disaggregates the total energy signal of a smart meter.

And you can look at various characteristics of that, using machine models, and get something like an EV charging signature and say, okay, we've detected an EV. Maybe you compare that detection with some other data and validate that an EV has been charged. That's actually really important for utility because that changes the nature of the customer relationship. You can be very direct. You can do clustering and segmentation and get customers on new tariffs and own a little bit more of the relationship on the EV side. And that whole propensity modeling piece around clustering and segmentation just using all of that data, utilities can be more purposeful about how they plant EV charging infrastructure. based off those decisions, but also their grid infrastructure.

So that non-intrusive load monitoring, using external data, identifying consumption and selling new products and services is happening now. And that's live at utilities, right? Just like the wildfire example, there's live at Duke and Southern California Edison and so on these EVV detection segmentation use cases. I think that's pretty interesting. It's real value from the smart meters that we talked about earlier. Yeah, so maybe what's distinct about that from the wildfire use case is not exactly new data sources, right?

Like that AMI data has been around for a while, as have all the other sources of data that you described, it's maybe the capability to weave them all together and the desire to do something with it, and particularly the arrival of EVs on mass, at least for some utilities, maybe catalyzes some actual action on this. Yeah, I think so. And there's upstream benefits, right, for how you plan your graded transformers and you look at rolling out new infrastructure and so on. Scott benefits both side of the meter, which is always appealing to utility. And it's about owning and transforming that customer relations a little bit, because there's

more you can do with that, right? You can transform how you communicate with the customer through more automated, personalized messages and just communicate and own that relations a little bit. We all know our relationship with the utility is the only time we think about the utility is when we get a bill. But this is a little different. And utilities are communicating more frequently with their customers than that. So, yeah, exactly, Shale.

That's why it's interesting. Again, you know, somewhat discreet, right? Not dealing with operations or power flow or any of those complex matters as yet. Right. Okay, well, Powerflow you mentioned we'll get to one of those. For number three, can we talk about something that is in the sort of core operations world of power flows? Yeah, so, yeah, there's two here, right? I think in core operations, the first one I mentioned, because this is highly important for utilities, right?

This, you know, substation, acid management, it doesn't light everyone's world and fire, but there's huge dollars associated with this, right? So, you know, you've got vibration sensors, partial discharge sensors, gas sensors, temperature sensors. You've got inspection drones and robots, and they're looking for, you know, wear and turn, equipment. They're looking at SF6 gases. they're looking at insulation breakdowns and so on. And what's interesting is, again, the algorithms have got so much better in the past five years. So these computer vision platforms, right, they're analyzing video feeds from fixed cameras and they're able to detect corrosion, right, on these assets.

The machine learning platforms, which can be built on some of these open source tools and then made proprietary are able to look at some predictive measurements. maintenance strategies, build it into existing digital twin software, where you're moving beyond anomaly detection into digital replicas and scenarios and so on. And I think that's a huge value use case for utilities, right? It can't be understated, right? You can reduce downtime of some of these critical components by like 30, 50%. You can extend the life of it. Transformers, right? They cost anywhere from 100,000 to a million. They take forever to be delivered these days.

I think all of that, that's pretty critical work-free utilities. And, you know, the whole system space here and the point applications and the integration, utilities have got really good at this. And this is where, when we talk about the proving ground of AI, it mightn't be cool to, you know, somebody from outside the industry. But I think those types of use cases, are pretty interesting, particularly at the substation level. Again, though, it's complex.

There's lots of protocols. I.C. 61A50. There's lots of complexity in here. But this is, I think, where we've seen the most maturity in IT, O-T-A-I-type work. But there are, I don't know if you want to touch on the power flow, too, piece, Shale. I think, yeah, right, quickly. I think, so let's say transmission capacity optimization, right? is an interesting area, right? So you've got lines that you're using to get renewables from,

you know, from up north, down south, wherever the low pocket may be. And in the past, you know, to look at the level of congestion or the heating or the sag of the line, right, which causes a traffic jam effectively, you know, utilities used to use traditional static rating, right? So that was, you know, based on consumptions and environmental conditions and some external weather. What they're doing now is moving to using near real-time data, right? So they're looking at the current, carrying capacity of overhead lines, right? And I think they're, you know, these are live these solutions. They're able to deploy non-intrusive centers.

So you've got LIDAR solutions that are monitoring the line. These are video cameras. You could be using mechanical data, look, vibration centers and you're losing real-time weather data. And that really gives you better control over how you write power. And it helps remove the bottlenecks, relieve congestion, maximizes current infrastructure. You know, the business cases here are really strong.

Perhaps, you know, the FERP requirements and the regular direction are still emerging. But I think an area like this, this is right in AI's wheelhouse. And it's non-intrusive and it can add kind of, you know, pretty, immediate value to increase transmission efficiency, for example. So that's pretty discreet, I think a pretty impactful area of AI right now. Yeah, I think, okay, so hopefully we've sort of convinced ourselves and everybody's listening that, like, there are, you know, these things are not, they're not as mind, obviously mind-blowing as all the generative AI stuff is today, but there are, for everyone that we mentioned, there are probably 10 that we didn't

mentioned, right? There's just like a million discreet little opportunities to leverage AI and new techniques and new tools and new data sets for to improve this system. The other question, though, of course, you know, especially if you're in the world of startups and venture capital and so on, is like, okay, but who's doing all this? You alluded to this before in the first wave, you know, a lot of what ended up happening, you know, it was Schneider and Eaton and GE and Siemens and ABB and the typical large utility vendors who ended up rolling out most of the new technology, there were exceptions, but that tends to be how it goes in this sector. As you see this new wave of innovation coming, who are the suppliers delivering the new solutions? And is it any different

from how it has been historically? That's a great question. I think it is a little different, Shale. I mentioned that the cost to develop these solutions is kind of collapsed a little bit, with these open source libraries and frameworks and packages. And we were purposeful in this work to look beyond the conglomerates and the grid joints to really get a temperature check on the market. And what we found was a series of startups here, looking at discrete problems, right? And across all of the areas where we mentioned it's the most pressing, even across the use cases, we mentioned there's a series of startups.

So there's a lot of activities. We looked at, I think, 350 to slow deployments since 2001. And we had a focus on startups. And I think what we found was that startups received $1.5 billion, right, from 80 rounds of funding related to those deployments since 2021. And so that's a pretty sizable capital inflow. Well, wait, just to be clear, that's the venture funding that went into those companies, or that's the money that came from customers?

Venture capital funding that went into startups since 2021. So I guess what that tells us is that there are startups being funded to take on these challenges. Doesn't yet tell us if they will win. No, that is, yes, I think the journey of a startup in this space might look very different than maybe other market shale. But yeah, it does not tell us that they will win. But it does tell us that there's a healthy venture market. There are startups.

There is data. And there is something different about this point in time in terms of access to data. That is not to discount that ABB, Seaman, Schneider, Electric, Oracle, aren't doing incremental changes to their products and picking up startups along the way. I think what's more interesting is where the startups are occurring. And they're probably, and they are more so on the grid edge. I think most of the startups we found were playing in the DOR integration space or EV charging management place. We saw some startups in the enterprise type of use case space, even looking at regulatory documents with LLMs, for example. As you go down the list from the grid edge and the customer into core utility operations, that becomes more grid joint focused, you know, around fault and the outage management detection and so on.

But, you know, where there's change or urgency for solutions, there's markets being created. They're typically organized around traditional utility imperatives shale. That's what we found. I can't say that there is an enormous new market emerging because of AI. As a result, I think AI is kind of riding the wave within the sector. Obviously, you need policy changes and business model changes and whatnot for complete reinvention in the sector. But there's quite a few startups. The deployments we looked at,

there's successful use cases that are scaling across the U.S. I guess final question, if we take a slightly longer term view here, the promise that lots of folks have gotten excited about, generally people who are new to the industry but are in the tech world, look at electricity and at some point get very excited about this idea of full optimization and automobiles. of the electricity network, you know? And they then run in like hard into the brick wall of reality at some point as they're

trying to figure out how to actually implement that full automation. You could take a few different views today as to how that plays out. One is that, look, this is a pipe dream and let's be realistic, it's never going to happen. The second is maybe it will happen, but it's going to happen via a series of incremental changes. There's no like blanket solution who comes in and automates the. entire system from generation to the customer, but we're doing it piece by piece now and maybe thanks to these new technologies, we can do it faster. The third is actually, yes, I think something will fundamentally change and the opportunity will be there to, like, I don't want to say

rip and replace, but execute a fundamental transformation of how electricity is delivered in a relatively short period of time. Like, of those three options, Or a fourth one that I haven't thought of, where do you sit? Right. Yeah, good question. I think there's possible and preferred and potential futures across all three, right? And they're all beholden to different characters. I certainly think that it's potentially possible. We're a few years away from being a few years away from that, though.

So I think the timeframe is really important. I think we could look if there was the right investment and the right technologies. And we can talk about some of the challenges to get there. Your third option is possible. It could be a reality. I think number two, I think the incremental change over time will lead us to a system that eventually becomes very automated with the support of regulatory infrastructure and the support of proven technologies through rate cases and so on. I think we'll get there.

But Shail, it's really important to underline in this that, you know, the data right now is not available, right, for, let's say, real-time decision-making across the grid, right? What we have is AMI, right, which is minutes to hours. We have SCADA, which is, you know, seconds to minutes. And then we have PMUs, which are kind of millisecond frequency and granularity. I mean, for that future you're talking about what we'd need is like 100%. nodal determination, right? And you need a lot to get there. I mean, right now, most of the work is on like post-event analysis after the fact model validation. And I think to get there, it needs a lot of investment. And that's where, you know, your third option would come in.

And I think the things that need to happen there are quite large in terms of deploying PMUs and having better interaction between the digital and physical layer of the grid, you know, taking into account OMS law, as we said. And I think that regulatory innovation, I think the, you know, even communications infrastructure would need to greatly increase along with those PMU deployments. I think we talked about explainable AI. I think AI would have to explain how a decision was made to a regulator. It couldn't be fully automated.

I think there's so many steps to get there. But I think like let's, even the few use cases we touched upon, they're pretty significant jumps for how the sector operates today. And it mightn't be like the tech industry where you could see a huge leap and it's phenomenal and you can chart the growth or something. I don't think that this counts from the level of transformation that's actually happening right now. It mightn't be, you know, front page news at fast company though. But it's still interesting. But yeah, I'm going to stick with my answer being the second incremental change. I think Shale is where we'll end up. It's probably the obvious answer.

But it would be interesting if a few years from a few years from now, something really turns over. David, thank you so much for doing this. Really appreciate the time. Yeah, of course. Great. Thanks, show. David Gork is the managing director at Indigo Advisory Group. This show is a production of Latitude Media.

You can head over to Latitude Media.com for links to today's topics. Latitude is supported by Prelude Ventures. Prelude backs visionary is accelerating climate innovation that will reshape the global economy for the betterment of people and planet. Learn more at preludeventures.com. This episode was produced by Daniel Waldorf, mixing by Roy Campanella and Sean Marquan, theme song by Sean Marquan. I'm Shale Khan, and this is Catalyst.