Securing Power Grids: AI Innovations – Kaitlyn Albertoli, Buzz Solutions

This is Textron tv. Hi everyone. I'm Bonnie Schneider, sustainability contributor to the Textron Group, and today I'm pleased to join be joined by Caitlyn Al Alberti, the c e o and co co-founder of Buzz Solutions.

She's here to discuss the, their cutting edge AI powered solutions for power substation surveillance and equipment condition assessment. Caitlyn, it is a pleasure to have you on. Uh, we, we'd love to hear more about your work at Buzz Solutions and the problems that your technology is aiming to solve.

Absolutely. Thank you Bonnie, so much for having me. I'm excited to be here and share a bit more about Buzz.

So tell us, yeah, I'd love to just dive in if that's okay. Yes. Please tell us about Buzz Solutions.

Yeah, absolutely. So at Buzz, we launched about six years ago from Stanford University, um, originally launching into the utilities market. Uh, at Buzz we're providing a AI powered analytics platform analyzing visual data for infrastructure inspections.

We started off in the transmission and distribution infrastructure space, analyzing data from drones, helicopters, fixed wing and ground-based, uh, imagery. But we have now expanded our offerings to include substation market as well, particularly given the, the climate that's been happening around substations coming under attack and, and being under threats. So, uh, we're now delivering, uh, as, as you mentioned, and automated solution providing alerts for, um, any sort of threat that, uh, may pose on substations as well as equipment monitoring.

So how does that work? For 24 7 monitoring, We're deployed on fixed camera systems. We're working with a partner in this space, uh, to deploy on the, the cameras that they have at substations.

We're taking in all of that visual data that's being collected, and we're providing real time analytics for that data, analyzing any sort of, um, variety of different types of detections, actually looking at the perimeter of the substations and any sort of threats that may be posed to the perimeter of the substation, looking at any sort of the components that are happening, any sort of the components within the site itself, uh, and looking at any sort of failures or damages that are happening on those components. And then we're able to, as I was mentioning, analyze it in real time and then send out those alerts to the relevant teams so that by the time they send a maintenance crew or uh, someone out to the substation, that they'll know exactly when and where to look for potential issues. Well, that makes sense.

You know, we do have physical attacks on power grids, unfortunately, on the rise. How does the AI technology that you use mitigate that risk? Absolutely.

Uh, we are seeing unfortunately that, um, not just substations, but infrastructure more broadly has been under attack and, uh, there are many more threats even that are happening and risks that are mitigated before there's an attack. Uh, but with our, a solution like ours, we're able to provide those real time insights. So many of these substation sites are in highly remote areas, and even if you have a camera that's, let's say out there monitoring the site, you still have to have a person look at that camera data 24 7, analyzing that, that information in order to see a threat or, or catch, um, an issue that's happening.

Uh, with our solution, we're really helping take that person who, you know, instead of sitting behind a behind a camera looking at the data, um, we're taking that human resource and saying, okay, you need to go exactly to this site and this location. And when by sending those, those automated alerts, uh, for us it's, you know, fraction of seconds where we're able to tell a utility, Hey, you know, we're detecting a risk that's happening here. We're detecting that a component is malfunctioning here, we're detecting an intruder, or we're seeing unmarked cars enter the substation.

And, uh, with all that data, we're then able to help provide that to those relevant teams who can then send out the right crew at the right time or, you know, maybe able to alert authorities, uh, whatever's the, the necessary protocol. Have you found that these individual threats to substations can have a much greater impact on power utilities? Absolutely.

Uh, one substation has, when one substation is out, it has the potential to impact thousands, if not tens of thousands of households. And that's just one substation. There's tens of thousands of substations across the US and, uh, you know, we need to take the utmost care of not only maintaining those, those, uh, substations and upgrading that equipment, but also monitoring for any sort of, you know, either bad actors, intruders, things like that.

Uh, we are seeing that the, the number of threats that are being posed on substations is skyrocketing. And so as such, we need to take the necessary protocols to, uh, reduce any sort of risk. What are the challenges that you face implementing AI into this solution for the power utility sector?

It's a great question. Uh, there's many, you know, many challenges of course, that, you know, we face when, uh, deploying at a, uh, at a site and providing real time insights. Um, many of these, uh, for many of these utilities, it's, these are new systems and bringing in our artificial intelligence or bringing in automation is an entirely new process.

And so being able to work with education and coaching and guiding these utilities on how to best use this equipment is one area where we're investing a tremendous amount of time. Um, the second thing is we're continuously increasing the number of detections that we are providing because there are so many different types of detections. And in order to train, um, an algorithm to have success at doing, uh, you know, a great job at detecting that particular type of anomaly, we need the data sets in order to do so, and we need the, the right data sets in order to train a successful algorithm.

So that's one area where we're continuously investing as well. Um, and then the last thing is, uh, you know, it's all about the right messaging. We find because our solution is really here to help support field workers, help support maintenance crews.

Um, a big part of our detections that we're also building in now are around safety and helping provide linemen and field technicians safety as well. And so we want to be a very positive force that supports these linemen and out in the fields, maintenance workers out in the fields. And so, um, being able to properly demonstrate that, you know, we're a helpful tool.

We are AI for good, we are a really, uh, strong use case to benefit those teams is another area where, you know, we spend a lot of time around that messaging. That's a really good point about safety, especially, uh, now, you know, in the parts of the US during hurricane season, and we have threats from power utilities with that, and of course the, the cyber threats that we mentioned earlier. But how do you view the role of AI in shaping the future for power utility security?

Do you see this as a, a greater trend going forward? Absolutely. There is so much potential for, uh, AI to provide a really positive impact.

Uh, we see that there's so much work that needs to be done right now in upgrading and modernizing grid infrastructure, especially as we have these really exciting but aggressive goals of onboarding renewables at such fast rates and electrifying, um, so quickly. That poses a lot of stress and strain and pressure on the grid, and it poses a lot of stress in and pressure on those linemen and field technicians. We see that the role that AI machine learning all of these technologies, these automation technologies can play is they can provide the right insights at the right time so that these linemen and field technicians can spend more time out in the fields really solving the biggest problems as opposed to sitting behind a desk analyzing this mass amount of data.

Because humans should be focused on the more reason and action-oriented tasks, while AI can take over the, um, you know, the, the basic mundane tasks that really aren't suited for, for a human in the first place. And the second thing I would add to that as well is, uh, we see that AI has a really negative connotation in, in so many ways. And, you know, AI can have some, you know, some not as positive use cases, but particularly in this industry and in this space, uh, we're seeing that there's a massive wave of retirement that's happening of these very highly skilled and, uh, very talented, uh, linemen field technicians and maintenance workers.

And yet there's a massive increase of work that needs to be done. And so as the shift is happening, we need to ensure that we're supporting these, uh, line crews, uh, and field technicians with the right tools to help them be even more efficient and more successful in their jobs. And that's where these types of solutions can bring real value.

You know, you mentioned something I just wanna ask about the renewables. You know, oftentimes we have different places in the world leaning more into renewables, some leaning less, but it seems that there, there's a hybrid approach of, of, you know, leaning into different types of renewables when, when they possibly can. Has that made it more challenging or like how, how can you talk about the relationship with that integration of a hybrid situation where maybe it's partially dependent on renewables and, and partially not, and where AI can kind of bridge the gap with that?

Absolutely. Um, a big challenge that we see in this space is, um, demand energy storage and being able to properly make the most of those resources. And that's where these data aggregation platforms, um, these, you know, energy management, resource management platforms, um, like the d e r derms platforms can provide a lot of value because, uh, we're able to more closely monitor, um, our energy consumption as well as the energy and the power that's being generated so that we can more efficiently and optimally utilize that power.

And as we are making this transition to renewables, and especially as we're in that transition phase, being able to get those near realtime insights is critical of understanding, um, you know, where we're doing well, where we need to improve, and you know, how, especially the shift in demand, uh, is going to impact the necessary generation that we're going to need to successfully onboard renewables at scale. Well, um, Caitlyn, as you're looking forward to, what's the head for Buzz Solutions? Uh, what are you working on now and what can we expect in the future?

Absolutely. Um, a big one is this substation use case. So for us, being able to deploy on these fixed camera systems and provide more, uh, more insights, faster to, uh, the utility and the utility customers, it's a big area where we're heavily investing.

We just, we just launched that news, um, just a few weeks ago, so we're quite excited for the growth in this market. We also continue to build on our offerings in the transmission and distribution space. We are coming into storm season and unfortunately as we're seeing in Canada, you know, wildfire season is already here.

And so for us, that's a very top of mind. Um, top of mind work is helping utilities reduce risk of storms, uh, wildfires, and also help reduce the number of, um, unscheduled outages or really un unneeded outages. So, um, that's top of mind for us and it's going to be a very busy couple of months and very busy couple of seasons here coming up.

So we're working on two, two different, but, you know, very exciting, um, kind of growth potentials with our product and we're looking forward to supporting utilities, um, as they are going through some of these challenges, whether it's weather related or whether it's, um, unfortunately attacks on the infrastructure. Caitlyn, Albert, c e o and co-founder of Buzz Solutions, very informative and timely considering, uh, what's been going on in, uh, the weather of course, and unfortunately the, the threats that you mentioned. So we appreciate you sharing your new technology with us.

Thanks so much for having me. I really appreciate it. Great.

Well stay with us on Techstrong tv. We're gonna have a lot more coming up after this.