Utilities may not be known for being first off the mark when it comes to adopting new technologies, but the power of data — particularly real time data — has many considering where AI could streamline their operations.
We sat down with Carlos Nouel, who leads Transformation Programs at National Grid, to talk about why he’s all in on the next generation of advanced metering infrastructure, what’s slowing the power sector’s adoption of artificial intelligence, and why everything comes back to better, faster data.
This conversation has been edited for brevity and clarity.
Stephen Lacey: What are the technologies that you’re currently focused on?
Carlos Nouel: One that’s pretty meaningful for us is advanced metering infrastructure. And part of the reason why that’s important for the AI discussion is that it all lays on the foundation of data.
When you think about the amount of data we’re going to have with the meters we’re putting in — the ability to have that sub-second sampling that enables us to do real time load disaggregation — it’s going to put us in a position where AI can actually become really powerful. And it’s not just from a consumer perspective, but also on the grid side. Think about taking all that data to build models that forecast what your load would actually be at any given time.
The second part of it is rethinking using customer data to start engaging them and to help lead them through this energy transition. And now is the opportunity for us to start leveraging all of that.
How can we get algorithms that say “Stephen, you’re a great candidate for EV”? The model has looked at their usage, the model knows your profile and can say this is a good EV rate for you, so you actually can save money.
The third piece — which is probably not as exciting, but is actually really interesting — is billing. As we move to a world where we’re going to be doing different types of transactions, we might end up doing peer-to-peer transactions, where consumers are also producers. How do we ultimately produce a bill for you that actually meets all those requirements?
And then the last part is creating products and services for customers in a distributed system, where you as a consumer become as important an asset as a transformer is today. Again, a lot of that is underpinned by having good data, being able to learn from that data really quickly, and getting insights that it would take a long time to produce ourselves.
We do think that this transition to AI is going to happen, and that we’re part of it.
SL: But will we be able to make use of this data? Do you think the next era of grid edge technology, partly driven by AI, will be different or better than the last decade?
CN: I actually do think it’s going to be different. In the past, we just had 15-minute interval data, but as an industry claimed we had real-time information. The reality is, when you’re 15 minutes behind — and you might even push it every hour at best, or even every eight hours — that’s not real real-time.
But now we actually are going to have real-time data. So that, for me, is a fundamental difference from what we’ve done before, because it allows you to quickly react to a signal that actually is happening in real-time.
Distributed grid computing power will also be important. It’ll give us the ability to have the data in real time, be able to process what is actually happening, and send that signal back to the broader system. I’m looking forward to when this technology is deployed, because every meter will be a computer, every meter will be a sensor that can provide us real time data on what’s happening on the customer side and the grid side.
SL: What’s your impression of how utilities are evaluating technologies with AI embedded now that we are right smack in the middle of this new hype cycle?
CN: I think utilities will probably be skeptical. It’s just the nature of the industry, right? We don’t like to take on too much risk, because at the end of the day, people still want to wake up in the morning and flip the switch and it needs to work.
For the most part, for the last 100 years we haven’t had a lot of problems. But as the grid changes, we’re now having a lot of problems that we didn’t have before. And I always feel that problems are the mother of creativity. We cannot keep doing things in the same way we used to, because that’s just not going to work anymore. I think to some degree, that’s going to push the industry to say “How do I navigate the transition while still keeping the reliability and resiliency of my network?”
Lisa Martine Jenkins: What do you think is the timeline when it comes to these technologies? When do you expect wider-scale adoption?
CN: My optimistic side would like to say a year or two. I think my more realistic side says it’s probably a bit more — two or three— when you start to get into more meaningful applications.
Imagine if you could shrink the length of a call center call. That’s real dollars, so I think this is one where you can iron out the problems really quickly, and actually start to show the real value of the technology.
But the proof is going to be in the pudding. As soon as we can start to show that this data can make a fundamental difference, I think it’s going to make the adoption quicker.
Carlos Nouel will be presenting at Latitude Media’s Transition-AI: New York conference on October 19th. See the agenda here.
