How innovation is changing the way utilities approach vegetation management.

and applies AI machine learning to provide a risk-based approach to vegetation management. Welcome, Kareem, how are you today? I'm doing very well. Thank you for having me, Adrian. It's a pleasure to be here and excited to talk about this topic. So maybe let's start with why is vegetation management so important to utilities? What are the challenges that you see there? Yeah, vegetation management for those not in the industry seems a little boring, but it's actually one of the most important factors causing outages that people experience, as well as wildfires, which, you know, for those of us that live in the West like I do know that's a major challenge. It's also one of the largest annually recurring cost items for most utilities to go out and trim and manage the vegetation. So it's both a source of a lot of risk to both the reliability and safety of the grid, as well as a major source of cost for utilities to sort of manage and deal with the vegetation risks that are out there. So a large topic that a lot of people don't think a lot about. So I read a recent stat from the DOE where they estimated the impact to the economy is around about $150 billion a year in the US alone. So with so much at stake, what is the traditional approach that utilities take when approaching vegetation management? Yeah, it's a great question. So historically, a lot of utilities have relied on what's known as cycle based trim. So basically, they divide their network up into parts and they go out and trim it on a cycle every three to five or six years. And they do a portion at a time. And that's worked well in the past. You know, it provides a lot of predictability and reliability to the costs and level of effort that a utility will do. It also gives them some assurance that they'll visit all parts of their network. And historically, that's worked relatively well, given some of the given relative consistency. I think we're entering into a period now where there's just a lot of exponential change due to climate, labor, inflation, and other things that are making it pretty challenging to sort of continue with that sort of pure cycle based approach, sort of which which also relied on largely, you know, sort of staff based patrolling of lines either on foot or perhaps aerial patrols to sort of augment that. But it led to a lot more anecdotal information and really a heavy reliance just on that cycle. So maybe we can take a deeper dive. You know, at Distribute Tech earlier this year, I think it was in May, there seemed to be a big buzz around vegetation management, more so than I've seen been going there for the last 20 years. Why now? Is there something changing that is all, I mean, there's much more, there's a lot of buzz around vegetation management? What's your views on that? Yeah, I think there's a couple of factors at play. I think one is some of the things, you know, I alluded to a moment ago around just the amount of change we're seeing in the environment. I think this summer has been a great example of just unprecedented weather events, you know, rainfall, ice storms, not this summer, obviously with ice storms, but this past winter, but you know, just major wind events. There's a lot of increasing amount of activity that's happening due to climate change, which is causing vegetation to interact and have more impact on the grid. So just more trees falling, more outages, obviously that's also manifested itself in a lot more wildfire activity. I think a lot of utilities took note of sort of PG &E's implication in some of the fires in California and the tremendous liability they suffered. I think they just settled for the Dixie fire and the Kincade fire this last year for over $55 million, which is real money. And so I think that combination of wildfire liability and just the recognition that the conditions on the ground for vegetation due to drought, you know, extreme rainfall events, disease are just changing so rapidly that the sort of status quo and the techniques they were using in the past just aren't really keeping pace. And I think that's really, you know, gotten a lot of utilities to focus on and think about how they need to update those operations and really take advantage, I think, of some of the digitization and sort of data-driven approaches that are permeating other parts of the utility business. That's really interesting. You know, you touch on digitization, which is a key theme we see across all workflows in the utility. And also analyzing data, there's lots of data available to utilities today. Could you share with us what you are seeing and how Overstory actually approached data, a data-driven approach to vegetation management? And the second part of the question is, what's the real benefit of that approach to a utility? Yeah, absolutely. So a couple of things that when we talk about data and digitization, I think, you know, vegetation management is one of the last areas of utility operations that have really started to go through this transformation. And I think the ability to get data on your network and really not just data per se. So not just there's a tree here and it's healthy or it's dead and sort of leaving it at that, but really turning that data about trees and vegetation into real risk-based insights. So one of the things that we do at Overstory is actually not just give information about where trees are and whether they're healthy or not, but actually then calculate sort of at the span level, a risk score that helps interpret and understand whether that's likely to cause an outage or an issue for the utility. So it's about actually interpreting the data into sort of relevant metrics, and then the ability to sort of aggregate and view those at a scale that's appropriate. So, you know, I mean, there's, you know, they get reports all the time from linemen or customers will call in and say, hey, there's a, there's a dead tree that I'm concerned about, but it's hard to weigh that risk against the a hundred other risks that are on the network. So I think the real promise of digitization and being able to sort of be more systematic about your data collection is that you can actually start to look at the system as a whole and say of all the risks that are being faced on the network of all the different places you have potential for encroachment or fast growing species, that might grow in or, or, you know, dying trees that may fall in, how do you start to stack rank and prioritize? And so that's one of the things that, you know, we help to do is actually get down to that sort of risk scoring on a per span level, and then allow utilities to sort of look holistically at the network and start to stack rank and prioritize, you know, which, which are the most, most valuable segments. And I think the real benefits of that are, of course, the cycle based approach has a lot of value. But now you can start to say, Hey, maybe there's some segments that are, I don't need to trim this year, even though they were on the schedule. And there's these other ones that really have had, you know, major changes over the last year, but I want to swap in instead. So the ability to sort of adjust and make changes to your plan based on the real, the actual on the ground conditions is becoming increasingly important, especially as, you know, labor markets tighten, costs of doing a business are going up, fuel costs are going up, all those things, the ability to make sort of more informed and better decisions about where where utilities are spending their resources, is critical to keep pace with the challenges. So it sounds like the data driven approach and using AI and machine learning to identify risk is something that can live co coexist with more traditional methods, but accelerate or provide a more efficient approach. Is that is that the right understanding? Yeah, I think absolutely. I think there's there's a lot of other, you know, these are more traditional approaches out there typically being, you know, foot patrols with with trained arborists or aerial patrols, either with, you know, arborists and personnel in the vehicles or, you know, using unmanned vehicles where you're collecting LIDAR or multi or hyperspectral imagery or others. And I think those are, you know, still have a lot of value and utilities can find a lot of value in those. I think that, you know, where we think about it, or how some of our customers have started to think about it, is, you know, doing those types of, you know, intensive patrolling either by foot or be with with aircraft tends to, you know, it's costly, and it's time consuming, because you actually have to go visit every single area. And then it takes a fair amount of time in post processing to get that data back. And it can give you very accurate, highly accurate information, but it's hard to get that systemic view. And so I think what we've seen some customers start to do is start to use satellites and other things to get the sort of full system view and start to make real risk based decisions about where to focus their effort. And then they can complement that with some of these traditional techniques, where it really merits the additional costs, and you need that additional precision to make an impact. Yeah, so listening to what you're saying, so, Tay, the system wide risk factor really resonates with with a lot of utilities that I talked to. How, how frequent could a utility use a satellite data approach versus the, you know, the, the, the traditional fixed wing or helicopter approach? Yeah, it's a, it's a, it's a great question. So the beauty of satellite technology is that, you know, the satellites are up there up there and orbiting the earth at all times. So you can get satellite imagery, you know, I don't want to overstate it and say you can get it as, you know, as quickly, you know, as daily or weekly, but in theory, you could, you could scan multiple times per year, every quarter, every couple months. If you want to keep track of real time changes, especially if there's, you know, disease outbreaks or things like that, you want to track. And I think that compares very favorably to, you know, flying fixed wing aircraft where, you know, generally you, you go fly a segment and then you've got to wait a few months before you can get the data back. With satellite imagery, you can collect very quickly and you can task it and schedule it. So you can do it multiple times throughout the year. And then the data processing is pretty quick on the back end. So you can really shorten that cycle, cycle time to get insights. And you can do it, you know, as, as frequently as you need in essence, I will say in practice, a lot of our, a lot of the utilities that we work with find there's a good cadence of doing it, you know, sort of twice a year. So they get sort of an early in the season and late in the growing season. And that gives them a good chance to sort of understand any changes as well as sort of verify the work that's been done throughout that season and really set them up for planning for the next year. It sounds like a reducing the cycle should have a very direct impact on reducing the risk and improving system reliability. What I'm, what, how do your customers measure the ROI from adopting new innovative techniques like satellite imagery and machine learning? How are they measuring the ROI in these techniques? Yeah, it's a great, it's a great question. And maybe it's best illustrated with, you know, a real, a real world example. So we work with a customer, actually, it's an IQgeo customer as well, called Show Me Power in Missouri. And they were running a large side trimming project that they wanted to do. And they were paying, you know, I think the cost per mile of side trimming, side trimming, for those of you that don't know is essentially, you know, you have a large right of way, and then you're trimming vegetation off of the side so that it doesn't grow in or fall into the lines. And they were running a large project cost about $15,000 per mile to do that trimming. And with over 2000 line miles, they were really looking at what's the best 7% or so that they could trim that year. And using the satellite based imagery and the insights derived from it, we were able to guide them very specifically on here's the spans with the most square footage of vegetation that you'll be able to trim. And that actually allowed them to sort of prioritize the best segments and saw over a 64% increase in the amount of vegetation they were trimming for the same dollars, based on a prioritization using the satellite imagery versus sort of their status quo methodology. I think there are sort of similar methods that they found, or similar sort of results they found with their hotspotting program, where, you know, previously, they've been relying on anecdotal methods to go out there and, you know, reports from linemen and stuff. And they were spending wasting a lot of time sending crews out there. And they were able to essentially go from two crews spending all year trying to track and respond to hotspots to doing that with one crew within a quarter of the year. And that saved them over, you know, $600,000 roughly over the course of the year, just in those reduced labor costs. So I think they're able to really quantify both in terms of efficiency of, you know, how best spent, you know, how are their dollars being best spent, but also in terms of actual raw dollar savings. The other metric that a lot of our utilities think about is things like other metrics like SADEE and, you know, SADEE is a measure of reliability and what's the outages that they're likely to experience and how many, you know, how does that impact customers? And then we've seen that even doing things like keeping their line miles that they're trimming the same, they're able to improve their expected impact on their per their reliability metrics. And so they're able to quantify efficiency in that way by, you know, sort of what's the SADEE impact I get per dollar spent on vegetation management. So we've had customers look at it in all those different ways and really start to quantify their return on investment or the return on number dollar spent. Yeah, that's a really good case study for both over-storing IQG or over-customers really been able to leverage the benefits of a progressive approach to vegetation management. One final question for you, Karim, to wrap this short chat session up. If the utility you're talking to today, they're interested in transforming their process around vegetation management, are there any best practices as is there any recommendations in how they should approach it? Yeah, I think there's a couple. I'd say one is, you know, let the data do the hard work for you. And I think really focus on your key business metrics. So if your key business metrics are things like saving costs, understand what those are, and then think about using the data and how you can optimize for those end goals. And I think sort of a similar point to this, or maybe a tangential point is about really focusing on the change management aspect. I think sometimes when utilities think about shifting to a data-driven approach or more of a risk-based approach, it can start to make them a little bit uncomfortable about the amount of change and just the change in operations. And that really, there's a way to do this where instead of wholesale changing your cycle-based trimming plans or wholesale changing out your processes, you can look at sort of making a hybrid approach where you're optimizing around those and making changes that are, you know, maybe it's 10, 20% of your segments that you're swapping in and out based on the data. So you're still getting the benefits of your traditional workflows, your traditional cycle base, while also reaping the benefits of the more data-driven approach. So I think there's sort of attending to that change management aspect, and then really tracking back the impact you're having to your core business metrics are really the sort of the two best practices that we've seen. Well, thank you, Kareem. It's been great talking to you today and learning more about the data-driven approach. And thank you very much for helping us on our IQ industry chat today. Excellent, Adrian. Thank you again for having us. And it's been a pleasure talking with you. Thank you.