Empowering energy transition with AI
The fact that AI can be applied in multiple disciplines makes it a powerful technology. Our guest, Sreedhar Sistu, VP Customer Offers at Schneider Electric explains that while AI can create damage when used with wrong intentions, it can also solve critical problems. One of them is growing energy demand – finding solutions to decrease energy consumed in buildings and industry is high on Sreedhar’s agenda.
In this episode he explains what AI can do to optimize energy demand taking multiple factors into account, such as weather forecast, building occupancy and its characteristics. He highlights that AI can reduce energy consumption while maintaining the same comfort level of occupants. Later, Sreedhar explains what are the key factors to implement AI successfully. Education comes at the top – it is important to understand technology to adopt it at scale.
Listen to the AI at Scale podcast
Listen to Sreedhar Sistu: AI for energy, AI for real challenges episode. Subscribe to the show on the preferred streaming platform (Spotify, Apple Podcasts).
Transcript
Introduction
Gosia Gorska: Welcome to the AI and Scale Podcast. This is a show that invites AI practitioners and AI experts to share their experiences, challenges and AI success stories. These conversations will provide answers to your how do I implement AI successfully and sustainably? How do I make a real impact with AI? Our podcast features real AI solutions and innovations, all of them ready for you to harness and offer a sneak peek into the future. Hi, this is Schneider Electric AI at Scale Podcast and I’m Gosia Gorska, the host of the show. I’m thrilled to introduce my guest, Sreedhar Sistu. Welcome, Sreedhar.
Sreedhar Sistu: Thank you, Gosia.
Gosia: Sreedhar is Vice President, AI Customer Offers at Schneider Electric. He’s currently working on applying artificial intelligence to address energy management, industrial automation and sustainability at scale. So you have large responsibilities, end to end ownership of AI offers, strategic planning, working with key customers and other stakeholders, and the magic keyword that appears in your job description is at scale. So Sreedhar, let’s start with this. What does it mean to you and how does it resonate with your current position?
Sreedhar: Thank you, Gosia. And thank you for recognizing the keyword. As you said in what I do, I think the term at scale is really key for us because when we set out to do AI at Schneider, the first thing that we considered was how to do this. Not as small pockets of innovation, which by the way, most large companies do. And it results in innovations and people understanding how AI works and what it can do for them. But to truly realize value of AI, you have to think about it at scale. And more importantly, when we think of a company of the size of Schneider, if you do not plan for things at scale, which includes, by the way, to really kind of give it something more concrete. Doing it at scale requires a lot of discipline. You need process, you need right kind of tools and technologies. Without thinking through all of that, it becomes a science experiment. So I think we are quite happy to say that this is at the foundation of our thinking, but I’m happy to go further into that, how we think about it.
AI and Energy: The Role of AI in Energy Management
Gosia: Yes, that’s correct. It’s visible across our strategy, our approach to AI. Definitely. And one of these areas in which we are scaling AI Solutions is related to energy. So I know that last year you had the opportunity to testify before the House Energy and Commerce Committee of American Congress. And I was wondering when I saw this information. It is a sign that AI is no longer an area for early adopters. It is really for everyone to play right now. And the biggest governments, but also the smaller players, they are now interested in how they can use AI. They see the potential, the opportunity. So. So if you could expand a bit more about this, what was the speech about exactly? And also if you are stressed or.
Sreedhar: Not, I’ll answer the second one first. So absolutely not something that I had in my mind that I would be sitting in front of US Congress to testify on this. But I think on a more serious note, it’s really interesting to see the interest from our policymakers on a topic such as artificial intelligence. So my sense is two things. One, AI has become common parlance, so most people now see it in some form or the other. And secondly, when you think of a domain such as energy, which is critical for us, for both the economic progress and general well being of people, I mean our modern civilization is built on top of energy. What AI can do is really an important angle to look at. I mean, there are many ways AI can help in the society, but I think energy is a domain where this is going to become a significant role. And I’m very happy to see that the policymakers were also focusing not only on AI in general, but AI specifically applied to energy. Which by the way, is an area close to my heart.
Gosia: Yes, definitely. And this is one of the areas that actually brings some positive thinking, positive feelings to me because we see different examples and different applications of AI. I recently saw the announcement that basically there are applications already that are cloning your voice. You can do your own avatar. And then people react sometimes with excitement, but also they have some negative feelings about it. They are, they are a bit scared how we will differentiate between what’s real and what’s fake. And I think energy is the opposite example because here we really apply AI to solve some specific challenges and I would really like to give us some examples. What are the areas? What is the role of AI for energy?
Sreedhar: Yeah, I’m glad you mentioned, you know, the comparison with the problems like deep fakes that AI can make possible and the application to energy. I think it’s important to emphasize that AI is a general purpose technology like many things that have come before. And it can do damaging things to society and it can solve some really critical problems and it’s up to us to make sure that we apply to the right area. And energy management is a timely, complex problem to which AI can provide some really interesting solutions. So if you think about broadly on the energy side, the challenges that we are facing are broadly increase in demand for energy. And this is natural with the economic progress, more people getting into the modern way of living. So there is an increase in demand and then there is a change in the way energy is being produced and distributed with the advent of solar panels and wind turbines and everything. And then I should correct myself, it’s not windmills, it’s wind turbines. Windmills have been around for a long time. And then we are also looking at new kinds of loads that are coming in, such as electric vehicles, more data centers. So if you look at this whole complex ecosystem, it is not something that humans can easily manage in real time. So the examples that come to my mind are, let us take the case of managing a complex building HVAC system. Right. I mean, buildings consume about 60% of energy in today’s world.
Gosia: Yeah. So we talk about air conditioning and heating, right?
Sreedhar: Absolutely. Yeah, you’re right. So buildings consume a lot of energy to both heat and cool. And it’s because you have some really predetermined rules to heat and cool buildings. What AI can do is AI can take multiple factors into account, including weather, occupancy and the characteristics of the building. So we can really try to reduce the energy consumption while maintaining a comfort level for occupants or whatever the building is doing. If it’s a factory, maybe it has different characteristics. So this is an important example of application of AI to energy domain that’s going to result in significant savings.
Resiliency, Microgrids, and Adoption Challenges
Gosia: Yes. And especially that people who are managing the energy at the building level, sometimes they are not the energy experts. Right. It would be great to have a tool for them to optimize the way that they use the energy coming from the clean energy sources and then how to charge the electrical vehicles in the most efficient way, right?
Sreedhar: Yeah, 100%. I mean, number one, people who are at the buildings level, they are overwhelmed with all the things they have to do at the building. I mean, there are. It’s not just the energy. They have more real time problems to address with the occupants. And more broadly, I mean, buildings are pretty complex. It’s. These problems are not something that humans can solve easily by themselves. This is where we need technology like AI. So absolutely, this is an area where, where it is not really about, hey, this is something that humans were doing and we want to do it this is really about augmenting humans, enhancing the ability of facility managers to manage the building more effectively.
Gosia: And there is one specific case that really got my attention. It’s about extreme weather conditions. So I know that we already have solutions to better manage and be more prepared. And it’s really related to the way that we can more accurately predict certain weather events. So could you maybe share with us some of the examples and explain how this solution works?
Sreedhar: Yeah. So we offer a solution called microgrids. And as you know, this is an area of growing importance. I had the opportunity recently to speak with a number of partners and the overall ecosystem microgrids. The interesting part is people say that AI is really critical to make microgrids work. And for those of not familiar with microgrids, a microgrid is really a culmination of different energy resources. If you are a building or a small factory, you might have some energy coming from solar panels or have a battery storage, you have a generator. Because first of all you want to have resiliency in the case of extreme weather events, that you have an unexpected weather event, you lose power, you want the system to be able to work independently, at least for a period of time. Same thing can apply to a data center which we rely on heavily. So the first part is to provide this resiliency and this microgrids powered by AI can ensure resiliency. Then the next step is how to optimize, how to make sure that we are using the energy that is produced locally, PV or wind, or we store energy when it is cheaper so we can use it later on and ultimately reduce peak demand on the grid. This is absolutely critical to make sure that grid works at proper levels without having this ebb and flow.
Gosia: Yeah, that’s correct. So I wanted to come back a bit. You mentioned this event recently where you had the opportunity probably to discuss with some of the customers, like facility managers, utilities. So what’s their approach and what are their feelings towards this kind of solutions? What’s the adoption? Are they read to experiment?
Sreedhar: Yeah, so I think it’s the overall response that I see from the customers that I work with is I think there is a bit of a caution to be very transparent, but there is a lot of excitement. We are working with a few customers, one in North America and one in Middle east, where we can demonstrate clearly the value that the AI brings, whether it is for things such as optimizing desalination. These are plants that consume a lot of energy and we know that we can optimize the chemical Dosing and then manage the membrane life and then also reduce the overall energy consumption. They see the benefits. Obviously what is important for all the customers is how to integrate this AI capability into the overall workflow. Because AI by itself is not a complete solution. It’s a capability that needs to become part of a workflow which then changes the game for our customers.
Gosia: Okay, I see. So it means that some customization is also needed. What else is needed from customers to use AI solutions, how they can prepare.
Sreedhar: So I am a huge proponent of education. I think it’s really key. And in my interactions at the US Congress, as well as some of the things that I am part of, I also work with Massachusetts state government, their AI task force. What we try to do is it is really important for people to understand that AI is your ally and understand its capabilities and not be swayed away by what we see in the press, that, okay, AI is going to, I don’t know, take away jobs or AI is going to take over everything. It’s a really useful technology. The more people understand and the more people experience it, the easier it would be for our customers to adopt it.
Generative AI and Future Energy Considerations
Gosia: Okay, I see. Yeah, I come across this opinion that actually the fact that we have such negative headings in the press and so negative examples that are brought to our attention about AI is that some people basically don’t even have opportunity to see these positive examples and that this is drawing some people away from AI, that they don’t want to learn about it more, they don’t want to be part of this innovation, be part of the consulting groups about foundations that are actually inviting different people to be part of this, of this innovation and really to bring some insights, also bring some fears and, and some possibilities to conversation to the point that I even saw this opinion that if we continue just to show this negative side of AI, probably some of the women engineers, women in tech, they won’t be that eager to follow up this career in AI because they are usually driven and they are more interested to follow specific applications that are, for example, in healthcare or maybe on the climate change, more related to the society itself, more related to ethical examples of the AI usage. So I see a lot of sense in what you’re saying. And do you have any recommendation then for people who would like to upskill and be prepared and maybe start to be engaged in this topic?
Sreedhar: Yeah. So I think it’s not superlative to say that it’s the transformative technology of this generation. We have seen how Internet transformed our lives today. You cannot imagine Life connectivity without Internet. And I think AI is the next frontier and it is very crucial for people to understand what this can do, whether it is in terms of learning. We can talk about Gen AI later, but you know, this is something in the common parlay now, how to interact with it, making the best out of it. I’ve recently come across a book that I finished reading. It’s called CO Intelligence. Okay. So really thinking about AI as a companion that’s going to work alongside you and you kind of learn to use the intelligent agent. I think making this part of curriculum and also making it part of everyone’s day to day work. I think the more we promote education, the more we promote awareness, both through formal means and informal means, I think we’ll be much more ready to harness the power of this technology.
Gosia: Yes, definitely. I’m great that you mentioned Genai because indeed Genai is part of the headlines that we see everywhere in the press. So could you maybe share us a little bit about how we are using Genai? Are there any opportunities to use Genai in energy?
Sreedhar: Absolutely. I mean, I don’t think there is any business area that will not be touched by Genai. And just to kind of put things in perspective, right. I mean today we use Genai as a way to generate text, summarize information, ask questions, et cetera. And one thing that you see is take the case of energy ecosystem. It is not just about forecasting or predictive maintenance, which are all heavily data centric and analytic AI, but there are many areas where we have lots of unstructured information, which is where generative AI excels. You have whole bunch of texts that people need to follow, maybe there are manuals that people need to understand and even prepare submissions to meet regulatory compliance, et cetera. So there is a lot of unstructured information that is embedded in the workflow, in the energy management. And I think Gen AI is going to have a significant role in easing pain out of some of those processes.
Gosia: Okay, so like analyzing, contextualizing the data that is available somewhere like in the systems, but also to be able to interpret probably some of this information. Okay, and how does it compare with analytical AI?
Sreedhar: So the beauty is these are complementary things, right? I mean there is no competition between these two things. In fact, fundamentally the techniques that are used in Gen AI come from the foundations of networks that were built for analytical AI. But I think broadly where this is going to differ is with analytical AI we are going to be doing lots of things with the structured data, sensor data that comes from them. Energy system is full of sensor data that comes in. It’s a complex grid, lots of equipment that emit data. And Genai is going to play a complementary role both in interpreting this information at large and facilitating the communication collaboration between people involved in this ecosystem.
Gosia: I see, thank you for this summary. And then the natural question that comes in, and that’s a question I ask to our Chief AI Officer Philip Rombach, who was also a guest of our podcast, is that okay, that’s great, we are saving energy, we are more efficient with this AI solutions. But what about the energy that we need to supply to AI models?
Sreedhar: It’s a fair question and I think it’s important that we are asking the questions at this time, not as an afterthought because we didn’t ask this question of what is the energy consumption of watching a YouTube video. Okay. It turns out watching a three minute video on Netflix or YouTube also consumes a lot of energy. In fact, the energy is probably not all that different from what we might use for an analytical AI model. So this kind of helps put in perspective that there is a cost to the things that we are doing. In the case of AI, my informed opinion and my conviction is that number one, this is going to be a net positive. Yes, it is going to consume energy, but I think that what it saves would be orders of magnitude larger than what we would consume. The second thing is a lot of the emphasis on this energy guzzling model. AI is coming from generative AI in the energy system. We are still going to be using lots of analytical AI. These models are simpler, not as energy intensive, but provide enormous benefit. So net net. I think yes, they are going to be consuming energy, but they are going to be providing so much more benefit that I think it is a cost that is worth paying.
Gosia: Yes, indeed. So it looks for me like in the future we will need even more education about also the energy savings that are related to the usage of AI. Like today we say that if you are not in the room, you should shut off the light, you shouldn’t use the water while you are brushing your teeth. So I can imagine already telling to my daughters that hey, you know what, let’s not watch this movie or let’s not send this email because actually let’s go for a walk and it will be more, you know, carbon neutral activity. So definitely I see this as a future education location for everyone.
Sreedhar: Yeah, yeah. I mean, you know something that the example that you gave of water faucets turning off etc is a good One, because we’ve learned over the period that we can optimize the usage of water. But there are. It’s also worth noting that, for example, data centers, you know, that we consume a lot, you know, we see consume a lot of energy. Their energy footprint has not changed significantly in the last 10, 15 years because of the continuous improvements in, you know, how we optimize their usage. And I’m very confident that with the AI we will find ways to manage the energy footprint of them.
Wrapping Up: Recommendations for Facility Managers and Final Thoughts
Gosia: Thank you Sreedhar. We are coming to the end of the episode. It was a really insightful conversation. The last question that I have, it’s maybe related to the facility managers or sustainability officers who are looking for this kind of solutions. They would like to prepare their buildings, their facilities for the adoption of AI solutions. What would you recommend to them as a, let’s say free first steps, what to focus on first, how to prepare.
Sreedhar: As you know, buildings and facilities are one of the largest consumers of energy. So there is a huge opportunity here. I think to me, like any technical problem, I would start with really the pain points that they have in managing these facilities. If it is in fact dealing with alarms in the building, that is an annoyance for people to do. We can start with a solution that AI can help to mitigate the pains with alarms or if, if the pain point is the energy bill that we want to contain some energy in that we can start with some optimization of the HVAC systems I think identify, but not too many. I think we want to have some really concrete scenarios where we can apply, see the benefit that really results in the conviction that people can then take on more and more use cases and ultimately some really smart and autonomous buildings.
Gosia: Yes, that’s a great thought. Thank you so much, Sreedhar. It was a pleasure to have you.
Sreedhar: Thank you, Gosia. It was great talking.
Like what you hear?
- Visit our website to discover how do we transform energy management and industrial automation with artificial intelligence technologies to create more sustainable solutions.
- Listen to the other episodes of the AI at Scale podcast.
- Read more about AI for energy transition in this e-guide.
AI at Scale Schneider Electric podcast series continues!
The first Schneider Electric podcast dedicated only to artificial intelligence is available on all streaming platforms. The AI at Scale podcast invites AI practitioners and AI experts to share their experiences, insights, and AI success stories. Through casual conversations, the show provides answers to questions such as: How do I implement AI successfully and sustainably? How do I make a real impact with AI? The AI at Scale podcast features real AI solutions and innovations and offers a sneak peek into the future.
Add a comment