The utility sector, at long last, is modernizing. And in many ways, it appears to be making up for lost time. New digital technologies and capabilities are unlocking unique opportunities for utility providers across the grid. And while sometimes the image this modernization evokes is in new, shiny hardware like meters, generators, and more, some of the most powerful upgrades are taking place on a software basis. In particular, artificial intelligence is bringing about more excitement than ever before for decision-makers at utilities.
But artificial intelligence, or AI, is sometimes difficult to grasp, and when not properly informed, people can be negatively misinformed. Ensuring members of the Energy Central Community can count themselves among the properly educated is one of the desired outcomes of our Network of Experts, and so we’re excited to welcome to that network someone who is greatly plugged into the world of utility AI. Sean McEvoy is the SVP of Energy at Veritone, and he was also recently installed as the latest expert in our Digital Utility Group.
And to start that process, Sean was gracious to participate in our Energy Central Power Perspective ‘Welcome New Expert Interview Series.’’
Keep reading to learn more!
Matt Chester: Sean, we’re thrilled to welcome you to the circle of Energy Central Experts. Let’s start broad to introduce you to our community: what is your role within the utility sector and how did you find yourself there?
Sean McEvoy: I ’m the head of the Energy Division at Veritone, which includes everything in terms of data science, software engineering, and other functions such as marketing, sales and implementation. We are here to bring artificial intelligence solutions to the energy market, and that was a natural fit for me.
I've been in the software industry pretty much my entire life from being a software engineer at various companies from Symantec to Dell, IBM, and Quest software and now Veritone. As Veritone branched into artificial intelligence solutions for green energy, that jumped out as where I really wanted to get into in my career in terms of being purpose driven. So, it was the perfect combination for me.
MC: Talk to us more about the AI use cases in the utility sector. What are some of the common applications and where is there opportunity for AI integration that perhaps people aren’t thinking about?
SM: Some of the most common use cases in the green energy field are around predictability. This is mainly because with solar, wind and hydro, these renewable resources are intermittent, meaning they are less reliable and reliability is always the top concern for utilities. AI is bringing predictability to green energy, not only in terms of how much you're going to get and how long you're going to have it for, but then also there's predictability in the markets in terms of how utilities want to balance their supply and demand on a day-to-day basis. And then also with independent power producers who have battery storage and want to be able to play on the wholesale market.
So, we bring a lot of data science and forecasting in those areas, in terms of grid-level and ISO-level demand forecasting, power generation, congestion, and demand. After that, you get into more complex kinds of areas such as optimization, which is around the control of the physical assets on the grid. We're able to build digital twin models of all the different hardware assets that are on a grid. we can take solar, battery storage, inverters, and transformers and ultimately optimize them. We can predict how they're going to work under what conditions, what is going to increase their life or decrease their lifespan, and how to optimize their output and performance.
One of the areas where we’re also headed is fleet-wide optimization and electrification optimization. You have your chargers, you have cars that have batteries. So, battery optimization for us is a key area in terms of how we're able to optimize at an aggregate level. With chargers and batteries, you're now talking big optimization problems, which encompass large amounts of data. If you can just imagine in a city all the different chargers, what their capacity is, when we need to charge, as well as what are the peak times, what are the off-peak times, when can we extract back energy into the grid, etc. That's where AI modeling and optimization really comes into play.
MC: When you’re looking to advise a utility partner about how they should be looking at AI solutions, what are some of the mistakes you come across in how the decisionmakers are thinking about this area?
SM: The most common one I would say is AI is coming to take my job, for example. Which we then explain that the human fits in the actual loop. Usually, the human is in the loop doing the task. What we really want to do is deploy AI and then move the human to the point on the loop where they're more in a supervisory role and where they can actually do more. Sometimes, once they get used to it, they let it run autonomously. For example, at Tampa Electric our Veritone iDERMS solution is now running autonomously. But it didn't always start as autonomous. It started with somebody in the loop, then somebody on the loop, and eventually they moved out of the loop.
But you really must get into conversations about digital transformation with the people that work at utilities and then talk about what is possible. You start with forecasting, and you help people understand what an AI model can do in terms of predictability for renewables. And then we go one step further and we talk about the digital twin model. And we start talking about the physical asset in their own terms. And then we will show them how we can make a digital model. And then we bring up a dashboard so that we can show the inside of the ramp rates changing, the temperature changing, the power output changing. As things happen dynamically on the grid, the AI in return may have a different cause or a different path that it's going to actually take.
MC: And do you find that when you're dealing with a new client, different utility characteristics impact their approach? For example, do different sized utilities have different priorities or do you notice differences between IOUs versus municipalities?
SM: Oh yeah, definitely. The IOUs have to be a lot more risk averse because of what they're dealing with is at a much higher level and the impact is much higher. So, we take baby steps all the time because our utility clients are dealing with FERC rules, NERC rules, and more. Even our last couple of utility projects, we went through full commissioning exercises because the stakes are high for them, and they want to assure new technologies are running properly before they add them to their day-to-day operations. Everything has to be on premise because they're worried about security and cyberattacks.
So, you even have to take a different approach to deploying the software based on the customer -- whether it's an independent power producer, a co-op, or a municipality, which are actually kind of quicker to adopt. And I'm not saying they're less cautious, but they're quicker to adopt because they’re at much smaller scale. And then they're also able to think about putting some of it on premise and some of it in the actual cloud. So, they tend to be more open to different types of technologies as well.
MC: Why did you feel compelled to get more involved in the Energy Central Community? And what value do you hope to bring to your peers on the platform?
SM: For us, it's about education, being available to people when they want to understand what AI can contribute to the utility sector, to the renewable sector, and to all elements from behind the meter to front of the meter. When I became a member on your site, one of the things I opened myself up for was questions. And for me, there's so much in terms of what I want to get out of Energy Central. You have the different rulings, for example. What is FERC Order 2222 going to do to decision making? What we are seeing is more independent power producers are now looking to do value stacking under batteries and play into wholesale markets, which was not done a lot before.
So how do you plan to use your battery in that sense, or how do you plan to dispatch or bid into markets? And then the ones around transportation, electrification. What are energy players thinking and how can I learn from them in terms of their strategies moving forward so I can try and position better where AI aligns? And again, because there's just so much happening in this space, the more people that I can talk to and leverage the better.
________________________________________
Thanks to Sean McEvoy for joining me for this interview and for providing a wealth of insights and expertise to the Energy Central Community. You can trust that Sean will be available for you to reach out and connect, ask questions, and more as an Energy Central member. So be sure to make him feel welcome when you see him across the platform.
The other expert interviews that we’ve completed in this series can be read here, and if you are interested in becoming an expert then you can reach out to me or you can apply here.