Wildfire Mitigation: Cutting-Edge Insights, Tech and Research
July 20, 2022 | Webinar
Wildfires present an all too real threat for countless families, businesses and communities. We went beyond the headlines for a deep dive into the latest digital technologies and eye-opening research that are powering insights into wildfire mitigation. From data-driven analysis at the property level to a new science-based wildfire mitigation designation for homeowners, our all-star panel of experts shared their latest work.
Presented by the Travelers Institute, the MetroHartford Alliance, the Master's in Financial Technology (FinTech) Program at the University of Connecticut School of Business, the American Property Casualty Insurance Association, the Professional Independent Insurance Agents of Colorado, the National African American Insurance Association Colorado Chapter, the Insurance Institute for Business & Home Safety, Zesty.ai and the Risk and Uncertainty Management Center at the University of South Carolina’s Darla Moore School of Business.
An open laptop sits on a desk. Next to it is a mug that displays the Travelers umbrella logo. Text, Wednesdays with Woodward (registered trademark) Webinar Series.
JESSICA KEARNEY: Good afternoon, and welcome to Wednesdays with Woodward. Thank you so much for joining us. I'm Jessica Kearney, Assistant Vice President at the Travelers Institute. And I'm so pleased to be guest hosting this week's program, standing in for Joan Woodward.
Our team here at the Institute, which is the public policy division of Travelers, began this virtual education series at the start of the pandemic to help us all navigate through these uncertain times. Since 2020, we've taken on some of the biggest challenges, and importantly, the biggest opportunities in insurance, risk management and leadership. And today, we're diving into wildfire risk and understanding the latest mitigation strategies. Before we get started, I'd like to share our disclaimer about today's program.
Text, About Travelers Institute (registered trademark) Webinars. The Wednesdays with Woodward educational webinar series is presented by the Travelers Institute, the public policy division of Travelers. This program is offered for informational and educational purposes only. You should consult with your financial, legal, insurance or other advisors about any practices suggested by this program. Please note that this session is being recorded and may be used as Travelers deems appropriate. Travelers Institute. Logo, Travelers.
I want to start by extending our-- really, our enormous thanks to the groups that you see here on screen. I want to call them all out by name. You can see there's many of them who are helping and supporting this program today.
Zesty.ai, the Insurance Institute for Business and Home Safety, the Master’s in Fintech program at UConn School of Business, MetroHartford Alliance, the American Property Casualty Insurance Association, the National African-American Insurance Association's Colorado Chapter, the Professional Independent Insurance Agents of Colorado, and the Risk and Uncertainty Management Center at the University of South Carolina's Darla Moore School of Business. Thank you all, thank you for your leadership. And welcome to all the members and networks listening in today to those groups.
Text, Wildfire Mitigation: Cutting-Edge Insights, Tech and Research.
So as we sit here today at the height of the summer heat, wildfires are in the headlines and they are top of mind. We are dedicating the next hour to discussing the very latest technology, research and solutions that are helping to power the next generation of wildfire mitigation strategies. I cannot say enough about our lineup of wonderful speakers today. They are truly leading the industry and we are incredibly grateful for their time that they're spending with us today.
So, first, some introductions. Roy Wright is President and CEO of the Insurance Institute for Business and Home Safety, or IBHS. He is a leader in climate adaptation, building science and disaster mitigation.
A former FEMA official, Roy brings together stakeholders, really, from across industries and specialties to tackle some of the toughest disasters facing communities in the U.S. From wildfires and hurricanes and tornadoes to hailstorms, he oversees the research and risk reduction efforts that are creating stronger, more resilient homes, businesses and communities. Roy is a recognized resiliency shaper who has dedicated his career to breaking the cycle of human suffering that strikes families and communities in the wake of severe weather.
Attila Toth is founder and CEO of Zesty.ai, a property risk analytics platform powered by artificial intelligence. Zesty.ai offers access to precise intelligence about every property in North America for insurance and real estate customers. The company uses aerial imagery, building, permit, transaction and weather data combined with AI to turn more than 200 billion-- with a B-- data points into insights about property value and risk, such as exposure to catastrophic events.
Before Zesty.ai, Attila served as Senior Vice President of Worldwide Sales at C3 AI. He was general manager at SunEdison. And early in his career, he was a strategy consultant at McKinsey. Attila holds an MBA from the Kellogg School of Management at Northwestern University.
And Eric Nelson is Senior Vice President of Enterprise Catastrophe Risk Management for Travelers, with responsibility for coordinating the evaluation of risk. He partners with our business units to develop pricing and underwriting strategies related to catastrophes. In conjunction with his role at Travelers, he previously served as chairman of the board at IBHS. Eric is an advocate for coastal mitigation strategies and has testified before Congress on the benefits of strengthening building codes.
Roy, Attila and Eric, thank you so much for joining us today and sharing your expertise. To begin the conversation, we're going to have each of our speakers provide some welcoming remarks. And then we'll bring everybody back together for some moderated discussion, and importantly, take your questions.
So you'll see the Q&A feature at the bottom of your screen. Drop your questions in now throughout the course of the program, and we'll get to them shortly. So, Eric, over to you. Why don't you kick us off?
Smoke billows from a hillside. Text, Evolving Risk Landscape
ERIC NELSON: Well, thank you. Thank you for that lead-in. And I appreciate both Roy and Attila joining me here today. And so I'm going to kind of set up the conversation here today by just talking about the evolving risk landscape.
So as we think about this, without action, you see it in the news constantly about the wildfires, the droughts. Without action, we face, really, an unprecedented increase in wildfire risk. And it's really-- It's not all about the weather. And so as we think about-- As we move to the next slide and we think about this--
I was raised in the '70s. A kid in the '70s. Boy, I saw Smokey the Bear on every kid show there was.
Smokey the Bear points his finger on a poster.
"Only you can prevent wildfires," that's what was drilled into you as a kid in the '70s.
And the stat was really startling when someone on my team brought this up to me, was 86% of U.S. wildfires are started by people. And so let's just put the weather aside for a minute and just focus on ourselves. What can we do differently?
And so you think about the major fires that have occurred. A "gender reveal party causes 45,000-acre fire in Arizona." That's really astounding. "Garbage truck unloads burning trash causing a 1,000-acre fire." "Sky lanterns cause an 800-acre fire."
There's so many of these examples that occur. We have to drill into ourselves and into our families and our children's and our communities that we have to take more prudent action at preventing wildfires.
Wildfire Statistics. Charts show California temperatures and precipitation, acres burned in Western states, California Drought Index and population growth in high-risk areas.
But the weather does come into play, as we see the worldwide temperatures. And I'm in Hartford today, but I live in Massachusetts. It's a warm week. Across the U.S., we're seeing warm temperatures. But really when you look at the trends, California, as an example, is hotter and drier today than any point in the last 120 years. So the temperatures are getting warmer. The acres burned in the West are increasing. And so the chart on the right shows that. And the drought index are becoming more frequent and more severe.
But on top of the weather stats, it's also, we just have to say, more Americans are living in higher risk areas. And so if you look at California, specifically the high-risk areas, population between 1990 and 2020, population up 92%. That's a big increase, considering the U.S. population over that period is only up 33%. So people are really moving to those high-risk areas, and we have to really consider that. What can we do? And Roy’s going to give some great examples in a few minutes about that mitigation of that risk.
Maps of the U.S. show drought conditions and forecasts. Text, Western Drought Expected to Persist through Summer.
Drought conditions as we look at today, on the left-hand side, you can see the U.S. drought monitor. Pretty significant droughts going out throughout the West. And the outlook looks like it's going to continue. So we're in for a tough, tough wildfire season.
A bar graph shows U.S. Industry Wildfire Losses, Industry Economic Losses. A sidebar lists Factors Impacting Wildfire Losses. Text, U.S. Wildfire Losses Significantly Increasing Over Time.
Now let's turn over to losses for a minute. So we talked about what people do, we talked about the weather trends, now let's put it together and just talk about economic damages that have occurred across the U.S. And it's really been on the upswing. The chart on the left looks at economic damage by decade. And you can see the really benign wildfire losses in the '90s and the 2000s. Then the next decade, we see a rather significant uptick. And then in the latest decade, we're on a pace to almost double the last decade. And so it's not just about the weather, it's population growth impacting.
Here's all the factors that will impact those wildfire losses-- forest management, fire suppression policy, season expansion. So when we look at the weather, predominantly looked at events in the summer, now moving into different seasons with higher winds. Inflation. We all know we're feeling the bite of inflation, but inflation, even before this recent period, we've always in insurance had something called demand surge. So when there's a big event that happens and there's hundreds or thousands of buildings and homes that have to replace, we typically will see demand surge post that event.
Legislative and regulatory impacts. Now there is a good sound public policy. I don't want to make that sound negative. Public policymakers look out for the interests of their consumers. And at times, that regulation could either increase or decrease potential losses. In a lot of cases, the protections put in place do increase loss costs for the insurance industry.
Evolving firefighting capability and capacity. Firefighting capability has been great over the years, but we have to recognize there's a finite amount of capacity that is available. And the last one is utility service line management. We've seen the impacts on utility lines and how those, in many cases, have started the fires.
Pie charts show percentages of U.S. Industry Catastrophe Losses. One shows losses from 1980 to 2021. Wildfires 11% since 2017. The other shows wildfire loss by location. California 72%, Other Western States 21%, All Other States 7%.
Now let's talk about the geography and let's put this in perspective for wildfires compared to other CAT perils. If you look at the last-- 1980 to 2021, you look at that period, where do we see most of the economic damage? While hurricanes, tropical cyclones make up a huge portion of that. Severe storms, winter storms.
Wildfires have made up about 6% of those losses historically. But, on the uptick, since 2017, it's now made up 11%. Big increase. And then what's the states-- well, California is predominantly the geography, but we see a lot of events that do occur in Western states. So that's something you have to be aware of, is where do the events occur?
Text, Public Utilities Have a Role to Play. An illustration shows an electric power grid. Another illustration shows right of way and safety zones between power lines and trees.
Couple more slides. Public utilities have been really talking about ways to mitigate the risk. And so they have a big part to play in this. And the California Public Utilities have pledged to bury thousands of miles of power lines. And that’ll significantly reduce the risk. In addition to that, they've committed to a pledge to remove hundreds of thousands of trees around the power lines. So those two effects are going to reduce the risk.
And my last slide is to say insurance companies have a role to play. And so insurance companies have a role to play. I'm very proud to say that Travelers is one of the founding members of the IBHS Research Center, but the Research Center is funded by the insurance industry at large. I think the stat is close to 90% of the industry participates in IBHS. Travelers, we're also advocates for federal and state actions. We're part of the BuildStrong Coalition.
It's also insurance companies' roles to educate consumers pre- and post-event. We also provide consumer incentives for appropriate actions. And then at the end of the day, we are here to make a timely response, if an event does happen, and have appropriate claim payout.
So with that, I was just setting the stage to lay this out. We're going to turn it over to Roy to have a more in-depth conversation about that wildfire research, that cutting-edge research that's going on at the Research Center. Thank you.
Text, Climate Resilience, Resilience Pathways and the Future Building Environment.
ROY WRIGHT: Thank you, Eric. I appreciate it. Hello. It's great to be with you all joining this Wednesdays with Woodward series. It's always a treat, even if Joan is off today. As Eric just mentioned, IBHS is the research arm of the property insurance industry. And Travelers is one of the leading insurers that invests in our work, along with modeling companies like Zesty.ai that you'll hear from in just a little bit.
Researchers monitor a controlled fire.
But for more than a decade, the insurance industry has been asking us at IBHS questions about wildfire and its interactions with the buildings where we live.
In a laboratory, a monitor measures wind speed and time as a fire burns.
Now our research is shaping our understanding of the wildfire peril, its impacts and the mitigation strategies.
A house burns in a laboratory space.
The growing impacts of weather on communities is undeniable in the face of climate change. Things like this, once confined to movies out of Hollywood, how appear every season just now in our news feeds.
The remains of charred buildings stand in a residential area.
We don't have a choice. Ignoring the changes will only exacerbate these outcomes.
Burned trees stand next to the remains of homes.
I do fear that in educating people about climate, we have simplified the complicated maybe too much.
A Google Earth map shows a view of the planet from outer space.
Reduce, reuse, recycle, bike to work, buy electric vehicle, green energy.
At the 30,000-foot level, we almost all agree that the climate change is happening, which honestly couldn't be said a decade ago. But when climate comes barreling through your front door, simple doesn't answer the questions.
The map zooms to the roof of a home in the Western United States.
We are not powerless. We may not be armed with one singular answer, but our climate change problems-- There are undeniable actions that consumers can take. We have choices to make. And while we make individual choices, these decisions really are interconnected. They're trade-offs, but we do know how to bend down that risk curve.
So let's look at homes in wildfire-prone areas. Each home must have an impeccable home ignition zone-- full stop. Nothing combustible in the first five feet around the home. And typically, that means rocks or concrete near a house, but just bare earth is fine. And I was explaining this to a reporter in Denver first of the year in the days right after the Marshall Fire. And repeatedly he asked me, "But won't those rocks and pavement raise the temperature of my home and drive up my air conditioning bill?"
I honestly didn't quite know where he was going with the question, so I leaned in.
The landscaping around the home alternates from soil bordered by rocks, to rocks, to grass only.
He was talking about heat islands and the effects that heat-radiating pavement have on the climate-- more heat and probably more heat-trapping carbon dioxide. The reporter's suggestion: less concrete and more trees. He's not wrong, but not everything that is good and right perfectly lines up. We have energy efficiency standards that can help offset this and we must find other places away from structures to plant our trees because we need them too. These are the choices, the trade-offs.
In California, we swapped our lush green grass for mulch to be drought-resistant, proving we can change the aesthetic norms around our landscaping practices. But the troubling irony, that drought-friendly solution is in and of itself a wildfire program. Whole yards are now essentially full of matches just waiting for a hot glowing ember to land and ignite, beginning that fiery cascade of damage that can spread from house to house.
So what do we want homeowners to do?
A digital drawing of scales and dice connect to a safety shield and a dotted line that ends at a home.
Understanding the risks and the trade-offs is something the insurance industry is really good at, but the average person struggles. They struggle to even understand what a 20% chance of rain really means for their day. In our current times, probability in return intervals, that's the language insurance speaks. Those aren't the vocabulary spoken at the kitchen table.
A home ignites in a laboratory.
So at IBHS, we're looking at this because we have to make resilience more accessible for homeowners. This is here in our test chamber, there being both sides of this duplex is being hit by those wildfire embers. At the point of ignition, it took about seven minutes to get the first ignition inside of the next 60 seconds.
This is how fast that fire begins to move through. It begins to spread under the deck because that too is dry organic material. And we see the smoke coming up through the deck, as well as there on the other side. We begin to see it move up the side where it will affect the windows and crack them and get up into the attic space.
A timer shows the fire reaches the edge of the roof at around 40 seconds.
This all plays out inside of one minute.
These kinds of pieces are difficult. And it was in the face of this that we said, we have work to do. We have work to give homeowners a pathway forward. And that's why last month, IBHS launched the Wildfire Prepared Home, the first wildfire mitigation designation program for single-family homes in California.
Text, Wildfire Prepared, A Program of I.B.H.S. A suburban home shows with a checklist. Roof, Choose a Class A fire-rated roof maintained clear of debris. Choose noncombustible gutters and downspouts. Building Features: Install ember- and flame-resistant vents. Ensure 6-inch vertical noncombustible clearance at base of wall. Defensible Space: Create and maintain the home ignition zone (0 to 5 feet), including the removal of branches that overhang this area. Clear and maintain the underdeck area; enclose low-elevation decks. Maintain yard clear of debris. Replace combustible fencing within 5 feet of the home.
Wildfire Prepared Home brings together the set of mitigation actions that must all be taken to meaningfully reduce wildfire risk. It's rooted in more than a decade of experimental research here at IBHS. In doing so, Wildfire Prepared balances achievability and risk reduction.
So we look at the roof, specific building features like vents and the base of exterior walls and the defensible space elements that are needed. Resisting that initial ignition is so key. The five-foot home ignition zone is based on research we did right here. Together with that six-inch vertical noncombustible zone, we reduce the risk of ignition without even have to replace all the siding on the home.
From the research right here at the chamber, embers we watch collect at the base of walls and in that side, that first five feet around the home. But each one of these elements isn't enough by itself. It really is an all of the above. This system is how we will reduce the impacts of wildfire on our communities. Climate change, it's here. And now we need to stop just admiring the problem and actively address its daily impact on our lives.
We know that these mitigation pieces work because our research, the field investigations, and collaborative studies with folks like Zesty.ai, this is how the science comes together and meets the needs of consumers.
A hand holds a metal ball that reflects the sun. Text, Twitter, Royster Wright. Disaster Safety.
I think the challenge in front of us today is we need to embrace these holistic solutions to address our growing risk.
We need to build for the present and for the future. Companies like Zesty.ai and what they're about to outline for us, Attila is going to show us how you can enable some of those technologies and help us navigate this really complex, interconnected set of decisions.
So with that, Attila.
Logo, zesty.ai. Text, AI for Insurance. Harnessing AI to Assess Risk to Every Building. July 20, 2022.
ATTILA TOTH: Thank you so much, Roy. You're always a class act to follow. Thank you for this informative presentation. Jessica and Eric, thank you so much for the invitation. My name is Attila Toth, and I'm the CEO and founder of Zesty.ai.
Zesty.ai is a property risk analytics platform. We use modern technology, as Roy has mentioned, to protect the livelihoods and communities from natural disasters. Our focus is on wildfires, but also on severe convective storms, wind events and flood events. We partner with national and regional insurance carriers, MGAs and reinsurers to build the next-generation climate risk models.
So what is this artificial intelligence voodoo, you may ask. And let me try to simplify this.
A zesty.ai webpage shows a satellite in space, a person stamping papers. Text, Eliminate the Guesswork from Underwriting and Rating. Proprietary AI Applications for Predictive Analytics. Structured Property Insights Captured through AI. Aerial Imagery. Permits and Transactions. Weather Stations. I.O.T. Sensors.
At the basic level, we use unique data. And when I say unique data, data that has not been available five or seven years ago; for example, aerial imagery from low-flying aircraft or ubiquitously available satellite imagery, building permits, transaction history, weather station data, and also data from sensors such as flight sensors, for example.
Then at the first layer, we use artificial intelligence, such as, for example, computer vision, to understand the key characteristics and risk factors associated with every single property. So our focus is very much property-specific. How complex is that roof? What is the completion-- what is the condition of that roof? How much of the roofline is covered by dry brush? Those are some of the questions that our computer models answer.
And at the third level, we basically build predictive climate risk models. And this is one of the models, our wildfire risk models that I'm going to talk about in detail now. Actually the name of the model is Z-FIRE.
A map shows areas around Santa Rosa, Sonoma County, Highways 101 and 128.
And let me start with just painting a picture of the 2017 Tubbs Fire. So this is the event that was the aha moment for us getting into wildfire risk mitigation, wildfire risk modeling.
What you're seeing here is Sonoma County in California. And with red, you are seeing the so-called high fire hazard zones as designated by Cal Fire and the California Department of Insurance. With blue, I'm showing about 300 or so dots that are properties that burned or structures that burned where humankind expected them to burn during the Tubbs Fire. And here, I'm showing you about 5,000 properties with blue dots that completely took the entire industry by surprise.
Multiple blue dots are outside the red area.
So this is the vast majority of the almost $10 billion of insured losses. Just think about what happened in Coffey Park.
An aerial image shows residential rooftops. The rooftops change color.
So this is when we thought that looking at this problem differently with the help of technology could be very beneficial. So we are leveraging aerial imagery, as you guys can see, pre- and post-wildfire events. We have trained computer and vision models, just like the human eye, to see which property was fully destroyed, partially destroyed or left intact in the wake of a wildfire.
Then we also look at what are the individual property characteristics-- vegetation density in multiple defensible zones, as Roy has mentioned, distance to the WUI, slope of the parcel, building materials that drive the susceptibility of an individual structure to wildfires.
As I mentioned briefly, our approach is rooted in science, but it's also rooted in understanding what happened in the past.
An image of flames engulfing a home. Text, Z-Fire (trademark). The World's Largest Wildfire Loss Database.
The past is not always 100% predictor of the future, but if you can build the largest loss database at the individual property level from 1,500 individual wildfires, dating back about 100 years in California, about 20 years outside of California, then you have a better chance of modeling the future.
From space, we come to a grid. Text, Regional Wildfire Risk. Very High Risk. Moderate Risk. Low Risk. Very High Risk. Very High Risk. Moderate Risk. Very High Risk. Very High Risk. Very High Risk.
So with that, we have divided up the continental United States into a grid. In this grid, every cell is about a 250 by 250-yard cell.
A Very High-Risk cell opens. Text, Risk of Wildfire 8 over 10. The key factors increasing level 1 risk are regional landcover class, low distance to high wildfire hazard potential area, and high wildfire potential score.
For every cell, we assign a probability. And yes, Roy, we talk to insurance companies more than we talk to homeowners. So we talk probabilities. So the score that we assign to every single cell is a probability of inclusion in a potential wildfire. It's an annualized probability. And basically we translate that into a score to one to 10.
Property-Specific Wildfire Risk. Risk of Property Loss, 4 over 10. The key factors decreasing level 2 risk are vegetation density: zone 2, high year built, and low vegetation density: zone 1.
Then we zoom in into every single one of those grid cells and we look at every single property. What is the roof made out of? How much is the slope that this property is nested on? How much vegetation does this property have in multiple defensible zones? And then we assign a vulnerability score, which is a dependent probability. So if the property happens to be in a wildfire, what is the expected damage rate to this property?
So bringing this together in a California example where there are 10 million properties give or take, on one axis, we are looking at the risk of wildfire. So this is a more of a regional view. On the other axis, we are looking at the risk of property loss. And then we have this very granular view of risk where we can confidently say that 88% of California properties are not problem children. About 12% is where we need to zoom in.
Wildfire Risk Breakdown. ~10M properties in California. Chart shows Risk of Wildfire and Risk of Property Loss. Low risk of wildfire, 88% or 8.8M properties. High hazard, 12% or 1.2 million properties.
And even there, not every property is created equally.
I want to flip up a quick example here in the Hollywood Hills, which is the area of the 2018 Woolsey Fire. So I keep zooming in on this property or seeing the high-resolution aerial image on the left-hand side pre-fire, post-fire. And here's where you can observe the power of property-specific risk mitigation.
So all those things that Roy spoke about-- defensible space, building materials up to code-- can be easily observed in this picture pre- and post-fire. Again, the point here that even though you might fall into an area that has higher risk for wildfire, not every property is created equally because property-specific mitigation matters.
So how do we know that this voodoo artificial intelligence model can give us useful insights?
A bar graph appears. Text, Z-Fire Wildfire Risk Score Distribution. Top 10 Wildfire-prone States, 2021. Assigned Wildfire Risk Score. The values range from Very Low, about 57 percent, to Very High, about 5 percent.
So this model is characterizing all the properties, so each and every property in the wildfire-prone states, into those risk categories that I'm showing here. As you see, most of the risk falls into very low and low. Just think about it. Downtown L.A., you should not be concerned about.
A second bar graph appears on the right. Text, Z-Fire Relative Risk Multiplier of Wildfire Inclusion, Top 10 Wildfire-prone States, 2021. The values range from Very High, 110x, to Very Low, 1x.
On the right-hand side, I am showing the distribution of risk. What you are seeing here is the relative risk multiplier of wildfire inclusion. So what does it say? Even though most properties are very low risk, they fall in the very low and low risk category, the differential-- the risk differential between very low and very high is 110x. That's what you want to see about a good model.
A map of the U.S. shows widespread regulatory approval, categorized by state as Adopted for Rating and Underwriting, Pending Approval for Rating and Underwriting, Adopted for Underwriting.
Also, there is regulatory traction with the model. Multiple states have adopted the model for rating and underwriting, including the California Department for Insurance, as part of rate filings by carriers. There are a couple of states that are pending filings, and there are a couple of states where the model is being used for underwriting.
So let me bring it together into mitigation options, right? The model is looking at mitigation options that are under the owner's control, right? So vegetation management, the different building materials, that Roy was talking about, are all under the owner's control. Some of them are easier said than done. Some of them are cheap. Others are a little bit more expensive.
But there are also variables, and we have to recognize this, that are not under the control of the individual property owner. For example, building density, the weather, fire response, distance to historical wildfires. So I just wanted to highlight this, that there are things that homeowners can do to mitigate, but there are also things that we just need to learn to live with in those models.
So let me bring this together into what are some interesting takeaways, having looked at this data for the past five, six years? Number one, vegetation is crucial.
Text, Approximately 47% reduction in aggregate risk. What does the data suggest about relative value of mitigation efforts?
Just by doing the effective vegetation management, you can cut your risk almost by half. Second, construction standards. Very important to bring up a property to new construction standards. About 12% of aggregate risk reduction opportunity. And 2008 seems to be the magic year for building codes.
And then on the other side, also tell me things that don't really matter. If you're an individual homeowner and has a wood shake roof, you might be concerned. But at the aggregate level, replacing all wood shake roofs in California will not drive that much mitigation because there's only less than 0.1% of wood shake roofs so the impact would be less than 1%.
Thank you so much for your attention. I'm looking forward to your questions.
Text, For more information, please visit www.zesty.ai. Contact Atilla Toth, CEO and Co-Founder. Atilla Toth at Zesty.ai. 1 4-0-4 6-6-8 2-5-4-3.
JESSICA KEARNEY: Well, Eric, Attila and Roy, thank you so much for those presentations. I think we covered a very good swath of information related to wildfires. So I'm going to invite you all back for our moderated portion of the conversation. But before we dig in a little bit deeper, we want to hear from the audience. So we're going to have some live audience polling.
Let's bring up the first question. We want to hear from you and we want to get a sense-- we've got over 1,000 people who are joining us from across the country and beyond. So you can see our first question. How confident are you in your knowledge of wildfire mitigation strategies and your ability to help educate others, right, because I know a lot of us on the line are in those roles and in that capacity.
So the answers are rolling in. It looks like there are about two-thirds of people-- I think we can share the response here. About two-thirds of people who filter into that middle bucket of somewhat confident. We have 17% who are very confident. 19% who are not confident at all. So clearly, we have some room on the education there and then looks like some room for about two-thirds of us who are only somewhat confident.
Any thoughts from the group here on, is that what you were expecting to see?
ROY WRIGHT: Seems right.
JESSICA KEARNEY: Right on there.
ERIC NELSON: I think because we have a lot of insurance agents and brokers joining and a lot from the Western states, they live this all the time so-- but anything we can do hopefully to provide more insights helps them.
JESSICA KEARNEY: All right, perfect. So let's move quickly on to our next question. We just heard in the presentations from Attila and Roy on this. Would you be willing to change the landscaping around your home to mitigate wildfire risk?
And this is interesting. I think we're about a 50/50 split between yes, I'll do whatever it takes, and, yes, to an extent. Almost no one answered no. Hopefully, primed by the presentations that just preceded this. Roy, any commentary on that? I know you talked about consumer acceptance and talking to consumers in a way that will get them to actually act. Can you comment on this?
ROY WRIGHT: Yeah. We're watching this play out. There's so much that needs to be done. And when you show people and says, you've got to move the hedges beyond five feet away from your house-- even the point of some consumers out in California I was visiting with a couple of weeks ago, it was a brand-new construction. It was just six months old, and they were very reticent to pull back from the landscaping that was up next to their house.
There's this attraction to it, yes, for something that's been in your yard for 30 years, but I think it's actually this behavioral shift is maybe the hardest thing in front of us. And again, just to highlight some of the final things that Attila was showing in his presentation, not only do we see this in the lab and in the field, the aggregate-modeling work that plays out continues to show how important that vegetation and those building characteristics are of the home. It's what differentiates some of those highest risk versus mid-levels.
ERIC NELSON: Jessica, I was thinking back to the '70s again when white rocks were popular. You think about going around your house and you'd see the landscaping. It was all the white stones. And then it changed over time to different color stones. Now it's mulch, right? So that's what we have to do as a society, it's getting those aesthetic trends and changing those.
JESSICA KEARNEY: And Attila, you mentioned, a pretty staggering statistic in your presentation that just the vegetation itself could reduce risk by half. So those are pretty eye-opening and we'll dig in more into those. I think we've had a lot of good information on all three of your presentations. And obviously, the theme for today is looking at the cutting-edge research and technology. We're going to go there.
Before we go there, Roy, I'd like to pull you in and I just want to step back and look at this from the human perspective. You shared all that great video footage and really gave us a sense of some of the scene and what it looks like. Being so close to this, can you share what it feels like to step into a community after one of these wildfires has devastated it?
ROY WRIGHT: It is. I've worked in the disaster space for a couple of decades and came to IBHS and was very immersed in the wildfire pieces. And that morning when the campfire moved through Paradise, California, I began getting texts because both my parents and my brother lived in Paradise and ultimately lost their homes in those fires. As a point where your worlds collide, where the research work and the things you talk about and then--
I use this line in my remarks but I says, climate change came barreling through the front door, and it's exactly what we experienced. In both cases, they were insured. They were able to move through their claims experience. And financially, they had the pieces. There's still a really, really long road to the recovery of the community. And this is this piece that I think all too often like, well, I have the right amount of money I need. And that is an appropriate place to put attention.
But by taking these collective actions inside communities, we actually change the trajectory about whether or not the community is going to be able to withstand it. I think that-- we focused on the parcel side, but the real game changer is when dozens of parcels in a given neighborhood all meet these standards. That's the point where we avoid that cascade-- the technical term is the conflagration of losses that hit in that space that really just rips apart communities in a really, really difficult way.
JESSICA KEARNEY: Yeah. Thank you. Thank you for sharing that. And I think that's a good segue. And Eric, I want to pull you in here. You talked a little bit about-- so we're talking about looking at the parcel level and the community level. And then you look at the scope of the risk, you talked a little bit about season expansion in your presentation. In recent decades, severity, size and number of wildfires across the United States has increased. Can you talk a little bit more about that season expansion and what we're seeing and how that's evolving?
ERIC NELSON: Yes. Yes, no problem. I'll go back to the video that Roy showed for a minute. You saw all the flying embers in the lab hitting the house. And those flying embers, that's the problem, is that flying embers are what's causing most of the wildfires and most of the structures to be destroyed. And so when you think about the summer moving into the fall, especially in California, you move into these weather patterns with these Diablo and Santa Ana winds, high-winds periods.
So now you could have embers flying a half mile, quarter mile, a mile. And so those are really flying into areas and creating really large events, much bigger, more severe events than we've seen before. So weather pattern and drought conditions expanding into the fall really create a tough environment for us to fight these fires.
JESSICA KEARNEY: Great. Thank you, Eric. And now-- so let's shift gears a little bit and let's talk about the tech and the research, which was the headline for our program today.
And Attila, you gave a very nice preview and look behind the curtain at Zesty.ai and just wanted to hear a little bit about the story. You mentioned an aha moment and how you brought this company together. Can you take us back-- I believe you founded the company in 2015. And we'd love to hear why you thought this was ripe for innovation and why you thought there was so much possibility here.
ATTILA TOTH: Absolutely. Thank you, Jessica, for that question. So started the company in 2015. And the glamorous startups right away fumble in a business model and they hit gold and they are off to the races. It did not happen to us. For the first two years, we were a couple of dedicated folks who were committed to doing something about climate and climate risk. We believed in the technology, but we did not have a business model.
So 2017, as I mentioned to you-- so for the first two years basically, we build those computer vision models where we taught algorithms to think like humans do and basically recognize roofs, recognize vegetation, and we did not know how we are going to monetize this. And 2017, as I mentioned, was the aha moment because we have seen the data that I showed you. We have seen the devastation of the Tubbs Fire. We went and we walked Coffey Park.
And that's where it dawned on us that with the technology that we have been building at that time for two years, we can better explain this type of climate disasters at the root, at the individual property level, right? Then we dug a little bit deeper. We are not insurance people. We dug a little bit deeper and we show that there are insurance models out there, but those have not been updated for decades, in some cases, for three decades.
We also understood that those are regional models, not property-specific models. We also understood that those models are taking limited number of variables into consideration. So that's how this idea came about putting artificial intelligence to good use here. And how about pointing this business to property and casualty insurance? And the rest is history.
JESSICA KEARNEY: That's fantastic. Thank you. And Eric, can you jump in there too just from the carrier perspective, from our perspective here at Travelers? Can you talk a little bit about the tech and innovation happening in the recovery space?
ERIC NELSON: Yes. So if you think about the last five years, we've seen incredible advancements in aerial imagery and AI. And so many times we'll partner with firms such as Zesty, KPAnalytics, various other firms to evaluate some of their specific products, but we also have access as carriers to that same sort of imagery data. And what's great from our standpoint is learning and gaining insights and deploying tools that can help us make better decisions on the recovery standpoint.
So we can see, post major events, we can actually see the imagery of the damaged properties. And that'll give us insights into our claim deployment strategies about where are the geographies that our customers have damage. We also-- if you're an agent of the Travelers, we've also made that post-event imagery available to our agents on our website and on our portal, on our agent portal. So post-event CAT imagery is available for agents and brokers that do business with the Travelers. And you can look at your specific book of business.
So great innovations. More to come. Attila, it's just the tip of the iceberg when he talks about every time I meet with Attila, he has a new product he's thinking about. And every time we meet with our own researchers, we think of new business problems we can solve with these AI and this imagery.
JESSICA KEARNEY: So we're getting just a ton of questions from audience members in the Q&A feature and I just wanted to pull one. Kim Stankovic is actually asking, how often is aerial imaging updated?
ERIC NELSON: It depends on the--
ATTILA TOTH: Eric, is it OK for me to--
ERIC NELSON: Yeah, you can do that one. No, no. Go ahead. Go ahead. I'll add just anything on top.
ATTILA TOTH: Yeah. Yeah. I'll give you the short answer and then a little bit deeper because there are many layers to this. So aerial imagery can be updated as frequently as daily, right? But there are four key components when we are thinking about aerial imagery. Number one, as you mentioned, the frequency of update. Number two is the resolution of that image. So for example, from satellite, you're getting one-foot resolution. From a low-flying aircraft, you can get two- to three-inch resolution image.
The third aspect is coverage. So for example, can you cover rural areas or only high-population density areas? And the fourth one is cost. So when Eric tells us this is the use case I want to point aerial imagery to, then we know what is the optimal mix of those four things. So just refresh rate, you cannot look at that in isolation because there's always trade-offs to make.
So for example, for claims use cases, as Eric has mentioned, refresh rate is very important. For the wildfire use case, people don't cut down vegetation every single day. It's OK. For example, in our filings with the Department of Insurance in California, we committed to at least three times updating the models, which is a lot more than what they have seen in the past, and they were very happy with.
JESSICA KEARNEY: And continuing on that line of thinking, Attila, how are the insurance regulators embracing new technology around wildfire risk?
ATTILA TOTH: Wow, that's a very, very good question. Look, when we started this and we said we would like to get some of the models embedded in rating, usually the answer we got, "Oh, AI is a black box thing. Regulators are going to have a hard time with doing that." And we have invested a few years in deep partnerships with some leading regulators where we spend a lot of time educating each other, right?
They educated us about the regulatory constraints. Like think about California, Prop 103 is posing a lot of regulatory constraints, right? And then we educated them about the art of the possible with the technology. And after that, I'm very happy to report that, as you have seen on that map, we have been able to garner approval of rate filings leveraging this model, which once thought to be a black box model in many, many states.
And the key to that, Jessica, is transparency. Once we could show that we are not only giving a black box score, but we are also explaining the key drivers that, Jessica, can take into consideration when mitigating their home. So we are not saying, Jessica, your property is high risk, but we are also telling you, you can decrease vegetation density from your roofline and cut it by half. You can upgrade your building material. So we tell you exactly why. And that transparency is what has won regulators over.
JESSICA KEARNEY: Great. Great. Thank you. Roy, I want to pull you back. And I know you talked about your Wildfire Prepared Home designation that was recently launched in your opening remarks, and you talked about how that was backed by years of research. You showed us some fantastic videos from the IBHS Research Center in South Carolina. Can you give us a little peek behind the curtain in terms of what research is happening today, what's on your agenda, and what are the priorities in terms of wildfire risk?
ROY WRIGHT: Yeah. As we look at the wildfire risk, clearly there's something available already for consumers to take action on. Literally today and throughout this week, we are running tests around how far structures need to be separated from each other, the structure separation question in suburban contexts. Because we know that wildfire risk is about a parcel and a set of neighborhood or community dimensions. And there is a difference when you are 50 feet apart, 20 feet apart, 10 feet apart, 6 feet apart in terms of what the risk is from a conflagration perspective.
We talk so much about embers. And embers is how we see more than 90% of ignitions occur. But in a really concentrated neighborhood, once a house does ignite, it starts becoming the generation of those embers that spread across the neighborhood and you can begin to see house-to-house kind of spread. And so, so much of the work this year is looking at those community dimensions, what kind of construction types work best in that, and what should the risk tolerances be?
Likewise, some work related to community firebreaks and other kinds of actions, again, to address this wildfire community dimension that, in the same way that parcels need a designation, we probably need the same kind of thing on the community front.
JESSICA KEARNEY: And do you think, just thinking about that community-level questions, are there new lessons to be learned from recent fires or do you think maybe this is more of an implementation problem where people need to act on the recommendations we've had out there and available?
ROY WRIGHT: It's both. We are learning from the events that play out. I'll tell you that in the main what we're seeing in our post-event investigations is simply-- and not only is it acknowledging and mirroring up to what we've seen in the lab, we often see it out in the wilds of reality. It's more true. It's amplified in that space.
And so I think we have work to do, particularly on the community dimensions because when we can work more the physical sciences, that can then get handed off to the Zesty.ais, as well as the insurers, to figure out how they want to apply that into their models and the like. And you get the physics meeting what they can do in the big boxes in the sky. The biggest piece is that element related to, are people willing to take action?
You asked me earlier what I thought about that for that response? And I gave you the polite response. I don't think 50% of the people watching here today would actually take all of these actions. I think they want to want to do so, but if they had to go spend a few thousand of their own money, are they willing to do that yet?
And I think we've got a lot of work on the consumer side of the equation for people to understand, you don't get an exception for a tree that you really like. You don't get an exception for the rose bushes that grandmother gave you. All of that is still risk and it has to get pushed back.
ERIC NELSON: Jessica, can I just add to this? It goes beyond wildfire. When we think about high-wind areas-- and I live in an area with high winds at times. And you really should cut back the trees around your home and the branches over your home, wherever you live, if you have high winds and other things that could occur because you're going to get roof damage, and sometimes significant roof damage. So I think if you look at some of the materials that are on the IBHS website, you'll see that removing it-- proper landscaping is so important, no matter where you live.
JESSICA KEARNEY: I want to--
ATTILA TOTH: Jessica, if I--
JESSICA KEARNEY: Oh, yeah. Go. Go Ahead.
ATTILA TOTH: Can I just follow up on this? Roy made a very important point about availability of data for property-specific mitigations versus community-specific. We all know that community specification-- community-specific mitigation is the next horizon, but I think we also have to be honest that there is just not enough data for community-specific factors.
So let me give you an example. When we partnered with IBHS, we could go and we could identify 71,000 properties that fell within 415 wildfire parameters. We could measure vegetation density in multiple defensible zones and we could compare loss rates in various pockets for those that were mitigated and those that were not.
And we could just empirically take those findings out of a lab in South Carolina, out of the IBHS lab in South Carolina, and basically verify those learnings at a cloud scale. In terms of community-specific mitigation, fire wise and other initiatives are great, but so nascent that data is not yet available to do the same type of back testing that what we have been able to do on the property-specific mitigation efforts.
JESSICA KEARNEY: Eric, can you comment on where the insurance market is today when we're thinking about changing climate trends and the latest technological tools? Can you kind of help wrap this up? I know we've got a lot of folks-- a lot of insurance agents and brokers on the line too who might like to know how they handle that.
ERIC NELSON: Just talking about-- the first step is-- the insurance industry, we spread risk over people and over time, right? That's what we do. And then when there's a major event, we're here to help people recover. And so with wildfire and changing climate trends and changing building trends, frankly, it's difficult for insurance companies to understand the risk right now.
So that's why we're partnering with a lot of these leading-edge firms to understand moving us from just thinking about high-level kind of rating, high-level territories to really the next generation, looking at what can consumers do at the parcel level? So I'd say, we have to continue to make the investments in understanding the risk and bring those products to bear that consumers need in the marketplace. And I think over the next 24 months, you're going to see significant changes in how we all understand wildfire risk, and that will lead to different innovations in the insurance market.
JESSICA - KEARNEY: All right. I want to do rapid-fire audience questions. We're getting a lot of questions about condo owners, people who live in apartments, people who live in HOAs, homeowner associations, where there may be restrictions. Any thoughts or advice for those folks?
ERIC NELSON: I'd say that Roy is in the team because Roy and I have talked about this a lot because Travelers is one of the leading commercial insurance companies in the United States. We've talked about this a lot. We're trying to learn from the residential side and really get that's where the majority of wildfire risk is, but there's many things that you have from a standpoint of a consumer and a neighborhood how far buildings are spaced together.
Some of the science that Roy is even doing on the residential side applies to the commercial side. So I think a lot of those good tips that we have on the residential side, even on the mulch around the foundations, those sort of things, there's a lot of things that you can do on the commercial side that are similar. But Roy, I don't want to put you on the spot--
ROY WRIGHT: I'll give you a couple of pieces on this. First of all, all of the action-- the fundamentals of the action are the same on a single-family home versus a larger scale apartment commercial structure. Once a structure is more than 40 feet tall, some of those dimensions begin to change. So we talk about a five-foot defensible zone. On a commercial structure that's taller than 40 feet, it would be larger. And we're working on the specifics on that front.
I do think the upside of apartments and condos is there is a master entity that can ensure the external pieces are met. I think when we deal with homeowners associations inside single-family communities, it becomes much more of a double-edged sword.
On the one side, if you're willing to lean forward, they know how to implement rules and make everyone comply. The flip side of it is some of the things that we're requiring, particularly in the landscaping, are directly contrary to things they have written into their rules, and they are reticent to change them. But every HOA I've been around has been some version of a democracy. And if the people in the community want the rules to change, they will change.
ATTILA TOTH: Let me just add very quickly that the technology works for both, right? So there is no delineation that the model doesn't work for commercial properties. Commercial properties tend to be a little bit further away from the WUI than residential properties. At the same time, those inclusion rates that we spoke about and the vulnerability rates also work for commercial structures.
JESSICA KEARNEY: Great. Thank you. Eric, this one's for you. We have a question coming in from Sheridan McDonald. As we look toward the future, do you foresee wildfires becoming more frequent and more severe?
ERIC NELSON: Yes. Yes. I do. The reality is when you think about, especially you think about California and their property costs and their demographics, their population is expanding into these wildfire zones, high-risk wildfire zones. That, in and of itself, is going to increase the risk over time. Plus, we know the impacts of droughts. It's likely droughts are going to continue in the foreseeable future, and so we're faced with that challenge. But I'm very hopeful.
What we've talked about here is climate adaptation to significantly reduce the risk. And so if we as a country come together between the insurance industry, the agents, the consumers, the state and federal government, we should be able to bend the loss curve down significantly. So without action, it's going up a lot. With action, we can mitigate that increases over time.
JESSICA KEARNEY: And I think that's an optimistic note as we're coming to the top of the hour here. Roy, one last question for you before we close it out. Jesse Gauthier wants to know, what does wildfire mitigation look like over the next 10 years?
ROY WRIGHT: Wildfire mitigation over the next 10 years changes the default landscaping around homes, and for new developments, requires us to holistically look, particularly at how you're dealing with density of homes, which, again, home affordability is a driving conversation in this country. We're going to have to find a place for wildfire mitigation. The practices that we need for construction and maintenance meet the realities of people who need homes, affordable homes. I think that the place where the science meets the reasonableness of American consumers is one of the biggest challenges that sit in front of us.
JESSICA KEARNEY: Well, Attila, Roy, Eric, thank you so much for those insights. Thank you for your time today. You gave us a lot to think about. We had over 100 questions. I'm sorry we didn't get to everyone, but that was a really fantastic conversation. We really appreciate your time. I do-- in addition to thanking our guests, I do want to preview for our guests today a little bit of what's coming up on Wednesdays with Woodward.
Text, Upcoming Webinars: July 27, A Conversation with CVS Health (registered trademark) President and CEO Karen S. Lynch. Register: travelers institute.org.
So thanks to everyone for joining us. We have a link to our survey about today's program in the chat. So please click on it. Let us know what you thought about today's program and let us know if there's other topics that interest you or other speakers you'd like us to engage with for this program. We do go through and we look at each and every suggestion and we really appreciate the engagement.
We have one final program for the summer coming up next week on July 27, and it's a good one. Joan Woodward returns and she'll be joined by CVS Health President and CEO Karen Lynch who will share her experiences leading a Fortune 4 company on the front lines of the COVID-19 pandemic response and how she envisions a future with better outcomes for community health. So registration for that program is open now. And we're dropping a link in the chat, so please sign up and register.
As always, visit us at travelersinstitute.org for replays of all of our programs, including this one, which we'll email out shortly.
Text, Watch Replays: Travelers institute.org. Take Our Survey: Link in chat #WednesdayswithWoodward
Thank you, again, to our speakers and thanks to everyone for joining us today. Have a great afternoon. Thank you.
Logo, Travelers Institute, Travelers
Join Our Email List
Get on the list to receive program invitations, replays and more.SIGN UP NOW