Predicting Wildfires, Worker Injury with Better Risk Data

Pete Miller [00:00:03] Hi, I’m Pete Miller, president and CEO of The Institutes. We’re a global not for profit whose mission is to educate, elevate and connect the people focused on risk management and insurance. You’re listening to Predict & Prevent, a podcast that explores how technology and resiliency can prevent losses before they occur. In each episode, we learn about innovative solutions and hear from leading experts on how they are making these approaches a reality today. Welcome back to another episode of Predict & Prevent. We have a fascinating episode for you today featuring two impressive tech innovators who are transforming the homes and lives of our community through risk management and insurance. First, I’m sitting down with Kevin Stein, CEO and co-founder of Delos Insurance. Kevin is an insurtech entrepreneur with a degree in aerospace engineering. He’s taken that knowledge and combined it with wildfire science and climate risk models to develop a more accurate assessment of ever changing wildfire risk. We’re very aware that the last few wildfire seasons for the West Coast have been record breaking in a number of ways, from the number of acres burned to structures destroyed. Delos, as an MGA, aims to give property owners a snapshot of their individual wildfire exposures via its geospatial A.I. models and help provide coverage where less specialized insurers might not. We discuss Delos’ A.I. powered approach, how it works, and how it provides a better view to predict and prevent wildfire risk. Here’s our conversation. Well, Kevin, thank you very much for joining us. Certainly appreciative for you taking the time out of your busy schedule to be with us. Just by way of background, can you share a little bit about your background and how you ended up founding Delos? 

Kevin Stein [00:02:14] Yeah, thank you, Pete, and I appreciate the opportunity to come in and talk to you about predict and prevent, about the era of change catastrophes in today’s sort of different landscape that we find ourselves in. I think what you guys are doing is critical to be able to understand from an insurance industry perspective really where we’re going and how to be able to adapt and react to the changes that we see in these perils that we fundamentally need to cover for a healthy insurance ecosystem. So I am an aerospace engineer by training, did my masters in aerospace engineering at Stanford. And like every aerospace engineering master’s student, I dreamed of being in the insurance industry. My co-founder, Shanna McIntyre, her background is physics and predictive modeling. So what happened was we recognized an opportunity back in 2017 to be able to model a variety of different catastrophes. Wildfire is obviously what we started with, but we have our our eyes on others as well. We recognized an opportunity to be able to model these perils in higher spatial resolution, [00:03:23]try and drill [0.2s] down to the address, but also temporal resolution. Get an understanding of how change of climate conditions over time affect exposure. What we quickly came to realize since the start of the company was that the structure that was in place in the insurance industry to insure a variety of different types of properties in climate affected areas, in catastrophe affected areas was a structure that was going to struggle with the way in which catastrophes are going to be changing from here on out. You have the third party model company structure where a third party modeling company creates a model and sells it to generalist insurance. These journalists insurers use this information to be able to understand how to underwrite and price effectively profitably in these areas. The difficulty with this is that that type of structure works very well for static perils, but struggled to keep up with perils that are changing very rapidly. You see the specialty MGA structure, however, be employed for risks that are difficult to understand, quite complex, need a significant amount of expertise, specialty expertise to be able to understand effectively, effectively enough to price profitable risk, but also for perils that are changing very quickly. And if you look at wildfire in California as just as an example, the effects of climate change in the way in which it is affected, the perils of wildfire is not just that the catastrophe is now more destructive, but it’s actually the fact that the peril itself is fundamentally evolving at a much faster pace than ever before. 2017 is sort of really when this new type of era sets out. We were introduced to a, you know, really a new type of wildfire behavior, the big wind driven urban conflagration, where the Tubbs Fire up in wine country to Thomas Fire down in L.A. 2018 showed us that this new type of behavior can happen in different ecosystems with a campfire out in the Sierra Nevada. Obviously, we had the Woolsey Fire down in L.A. as well, similar to some, the Thomas Fire the year before. 2020 introduced us to yet again another type of wildfire behavior, this dry lightning ignition. Now, that’s something that does happen in other areas of the country, happens in Colorado, but has not happened in the Bay Area during this climate, has been been very, very long time. Therefore, sort of three out of the last five fire seasons, if you will, at somewhat of a new type of behavior. And you get yourself in a situation with these third party data models, really can’t keep up with the complexity of the risk. Therefore, we saw this in 2017. We said we need to have a different solution to be able to effectively price catastrophe moving forward, to effectively insure, profitably insure homes and commercial properties in these regions. So we created a specialty MGA that sells home insurance and commercial property insurance in areas that other groups are avoiding due to potential wildfire exposure, as I should put it. 

Pete Miller [00:06:40] So wildfires, your specialty, as you say, can you give us an estimate of the size of the market in California or what’s the addressable market? 

Kevin Stein [00:06:50] Yeah, so our analysis shows that there’s about six million homes across the entirety of the US that are struggling to find coverage because of perceived wildfires exposure. Our modeling says that three million of the six million are miscategorized, are actually quite safe and deserve to have pretty reasonable policies. The other three million are definitely exposed to wildfire. However, with certain hardening changes to the home with some landscape [00:07:23]issue [0.0s] force management with a variety of different changes. Those homes can be insured profitably as well. 

Pete Miller [00:07:32] You hear a lot about California wildfire, right? It’s certainly and as you say it, I hear you say you have temporal data, so you can watch it sort of evolve over time and it sounds like you can get pretty specific right down to an address. How does that compare to what like the state has or other modelers have? 

Kevin Stein [00:07:53] Yeah, it’s a good question. We are at Delos a partnership actually between us and an environmental think tank named the Spatial Informatics Group. The Spatial Informatics Group  work with Cal Fire, the Public Utilities Commission, the U.S. Forest Service to help create their modeling. It’s actually 100 different scientists, researchers, professors, post-docs, so they are constantly looking at this at this peril and coming up with significant amounts of proprietary data that allows them to be able to predict the science of fire and how it’s changing much more accurately. You know, a few examples. We have a fire climate model that’s based on hour by hour data for the last 11 years for all California. Through that, we’ve been able to identify which areas can have a wind driven fire and which areas cannot. There’s actually specific climatic conditions that are necessary to have that type of event. And there’s a lot of areas that look heavily wooded or have had fires in the larger proximity. But based on the climate conditions of those locations, they cannot have a big wind driven fire. We have climate projections that look out one year, five years, ten years and more. We have a lot of different aspects that go into it. We have 200 total input layers, variables that go into our risk model itself. And with that, we’ve been able to be not only a lot more predictive in terms of what has happened in the past, but our model holds up much more accurately in future coming years. 

Pete Miller [00:09:37] So you’ve done studies that said using our model, which it was, it’s a supervised unsupervised model?

Kevin Stein [00:09:45] Certain A.I. algorithms are better for certain types of problems. There’s better at solving those types of problems. Wildfire is one of those perils where more data doesn’t equal more accurate results because it’s actually a data starved peril. There’s only a handful of wind driven fires that we’ve seen. And what we’re trying to do is predict not just the likelihood of that exact type of fire, but where these fires could be in the future, where if they are in a different geography, in a different ecosystem, in a different wind system, there will be some aspects that all of the different, but certain aspects that are similar. And so you have to be able to have a model that is allows for supervision in order to make sure invalidate the fact that you’re creating a model that can accurately predict where the next fire will be, not just to where the previous fires have been, but to directly answer your question. It is a supervised model because what the models are trying to do, this algorithm is trying to do is help our professors be able to predict the science of fire more accurate. Therefore, they need the visibility into the model to understand exactly how it’s responding to the information that we’re feeding. Because the all the different training data and input data, all of that needs to be a very specifically curated set of information to make sure that the algorithm is responding to the actual science itself, a fact. 

Pete Miller [00:11:25] Yeah, just for folks listening, there’s kind of two kind of systems, right? Supervised and unsupervised supervisors, as you’re saying. Says, here’s the data, here’s specifically what it is, and I’m to supervise said, here’s a bunch of data, you figure out what it is. I mean, that’s effectively that’s Pete’s short definition, Kevin. 

Kevin Stein [00:11:46] No, it’s you’re totally right. And it works great, right? In a situation where we have, you know, this in this world today, we have way more data than we ever had before. And so unsupervised models can pick out patterns that we can’t see as humans just because it’s too large of datasets. And it can teach us some really interesting things. But wildfire and catastrophe is not something where that’s useful because you have to have that intersection between using [00:12:17]C [0.0s] an amount of data and understanding how the science of that perilous change. 

Pete Miller [00:12:22] So you talked about how your approach is unique, and I know that part of your model involves preventing risk. Talk to us a little bit about how risk prevention is incorporated into your model. 

Kevin Stein [00:12:33] Yeah, the way that we underwrite is sort of a two step process. The first step is understanding likelihood of fire at any one location and intensity that that fire would be. In the second step is saying if there is a fire, what is the likelihood that this home itself will burn down or commercial property or utility, etc. that second piece is incredibly important. And that second piece, because we’re writing, you know, insurance at scale, we’re able to understand and we will be able to even more so move forward, we’re able to understand the indicators and the price of changes to home harmony, to structure hardening. We have an adviser that was a Cal Fire chief fighting fires for over 30 years, but he also ran the group in Cal Fire that was collecting data to understand how effective home hardening is. IVHS does a tremendous job, but they’re not collecting data of, you know, thousands of data points across California about specifically which homes burned in Wisconsin for each fire and what were the unique underwriting characteristics are different between. Because that is what our business is meant to do, we’re able to collect this data at scale for really the first time and directly implement that into the price of that underwriting factor. The standard actuarial analysis with that data will get you to the point where you can say, okay, changing a roof is more important than moving that tree to open windows are twice as important or half as important as ember mesh screens. It is implementing ember mesh screens is a 100 dollar difference in the actual risk of this home will have the information to be able to do that. So what we’re trying to do here is not just start by providing coverage to homes of commercial properties that deserve them, that are safe and the industry is avoiding right now just because that rest of the industry is not the wildfire experts that we are just totally fine. But we are also trying to implement a system where we can basically work with the insurer to show them exactly what aspects, what changes to their home or property can effectively reduce loss and help incentivize that through the mechanism of insurance of pooled risk, help incentivize those changes to continually make properties across California and in the rest of the US safer. 

Pete Miller [00:15:19] So that’s really cool, right? For me, from two things that I think you’re doing that are very groundbreaking. One is I hear you say we can tell you mitigation efforts at the house level, which is phenomenal. And secondly, so you’re actually providing incentives because a lot we hear a lot and, you know, we’ve done several of these and it’s and it seems to be a challenge of how to get sort of the end user to use the you know, to use the the A.I. or to use the the knowledge that comes out of this to prevent, you know, to actually prevent things and incent incentives seem to be pretty important. You seem to have kind of considered that and taken that into account. 

Kevin Stein [00:16:04] Yeah. So the insurance as an industry, as a societal mechanism is incredibly powerful. Obviously, it provides a financial defending mechanism for all society allows us to even out risk between really catastrophic events or just certain kind of bad times and in certain good times. Allows us to take risks, allows us to create new types of technology. I know for a fact from the aerospace industry that we wouldn’t be launching rockets if we didn’t have insurance. We wouldn’t be putting new types of satellites into space, observational GPS, etc., if we didn’t have insurance. Because nobody could take that hit on their own balance sheet. So it does a lot of really important things for society. But what it also does is give a platform, a financial platform, to be able to positively incentivize these sort of actions as opposed to cost money. I think it is going to be difficult. And there’s other groups out there working to solve this problem directly with consumers. There is groups out there working to solve this problem through civil governments, and I absolutely admire their efforts. Insurance, given the role that that we play, we’re already sitting in a situation where we’re insuring these properties, and we’re insuring it for the risk that they are at that moment, if we can accurately identify and price that risk. Therefore, we have the, you know, financial platform, the financial vehicle already embedded in this industry to be able to reduce that price if that risk is reduced. And obviously it takes a lot of data to be able to get there. I think one of the reasons why it’s been difficult to implement this in the past is because we don’t have the information as a whole industry, right. These perils have gotten worse at a very rapid rate. And previously we weren’t collecting that data because it didn’t make that big of a difference to the risk, and all of a sudden it does, but we you know, we don’t have the years, decades of history to be able to look at it and say, okay, this is that difference. So it has to be sort of put together in a lot of ways. That’s what we’re trying to do. We’re trying to build an insurance provider. We are an MGA right now. We’re trying to build a trust provider that specializes in doing exactly this. And so we’re going to collect data that’s specific to our demographic of homes. That’s going to be different than somebody else that’s providing insurance in a different area that has different risks. But, you know, we are, I should specify, we were working on where could you do that building this incentive platform. And it’s going to be a lot more filled out in the future than it is right now. It’ll be much more complete as we continually get more data. 

Pete Miller [00:19:11] So you mentioned you’re an MGA. How, you know, how does that factor into like to predict and prevent has, you know, we’re focused on into an insurance solution. So how was the implications for you of being an MGA and how has that helped you? 

Kevin Stein [00:19:25] So we strongly believe that this is the right structure. Maybe an interesting way to look at it as sort of a bit of a thought experiment is to think about Cyber Risk. So the cyber insurance issue of our general liability. And pretty quickly the primary carriers looked at said, I don’t know what this is, which is reasonable. Right? The constant war between hackers and security personnel is changing on a weekly basis, probably. I have no idea. I’m not an expert at this. But I do know that it’s extremely complicated and changing rapidly. Therefore, it is much better addressed by specialty MGAs. And so, cyber came out of general liability and you had space for a variety of different insurtech to jump in who are these you know security experts. Cybersecurity experts jump in and create products that are specific to this peril, right, where you have risk mitigation services, software you put on your company systems that are able to detect threats, incoming threats. The other thing that that software is enables is the ability to gather data. What are the most common cybersecurity threats that we’re seeing in May 2023? It’s probably got a name that you and I don’t even understand, and therefore they can stay on top of the changing peril. They’re collecting that data. They understand it because they’re experts and therefore they can adapt, adapt and react to their coverage and pricing to be able to be a healthy balance sheets continually into the future, even as these perils change. We believe at Delos that climate affected catastrophes are very similar at this point. The perils changing very rapidly, and it’s much more complex than it used to be. So you need groups are really heavily specialized in a certain area. The MGA aspect, as opposed to being a specialty carrier, makes it a bit easier to be able to share risk on the back end. You can have a variety of different the insurance companies can do this too, but as an MGA little bit lighter, it’s little bit easier to put together a large scale panel, different insurance companies, different reinsurance companies to make sure that we can provide the most amount of capacity to this to segment. It allows for sort of larger market penetration. You know, we’re not going to get, we’re not going to have our carriers be very aggregated. That’s something that we make sure to never over aggregate any one capacity partner that we have in being an MGA, as is, to have many full capacity partners to make sure that that stays true. 

Pete Miller [00:22:07] Yeah, a very interesting future you present because that’s pretty fascinating. And obviously you’ve done a wonderful job for your company and for your clients. What’s next for you in your organization? 

Kevin Stein [00:22:18] Yeah, so what we’re trying to do is work with every different entity in the insurance industry to be able to solve this problem. Right? When it comes to distribution, for example, we’re not trying to compete with the primary carriers in California. We are going to these carriers and saying, hey, if you’re going to decline, only if you’re going to decline a home or commercial property because it may be wildfire exposed, then partner with us and sell our product instead. And that way, you know, you can keep if you sell our home insurance product, you’ll be able to keep the client. Maybe the client has five cars, a home in the middle of L.A., a home out in the Sierra Nevadas, and umbrella policy and a boat. Great. Instead of saying, we can give you a products for all of these, except for the whole mountain of the Sierra Nevadas, say, I can give you a coverage for all of these and here’s a partner that will give you coverage for the one to Sierra Nevadas. That  solves the problem, we believe, for carriers in California allows them to not need to worry about whether or not their risk models are exactly accurate, not need to take that reinsurance hit, not need to worry about the PML associated with that and still get their agents happy and stay active in these regions just in the other other lines of business. That plus we’re continually adding products to make sure that anybody that’s struggling to find coverage because of these catastrophe exposures, whatever it is that they want, we want to have a policy for them. You know, we’re working to get commercial property up and going this year. We already have home insurance and take it home in build address, but we’re working on an Airbnb product as well in commercial property. [00:24:11]Just a key [0.0s] just being able to build out a platform and build our distribution so anybody that’s in a potentially wildfire affected area that needs whatever type of policy they find their way to us and we have [00:24:26]approximately. [0.0s]

Pete Miller [00:24:27] Is there anything else you’re seeing that would be worth watching in the context of risk prevention? 

Kevin Stein [00:24:33] I would just encourage everybody in the insurance industry that’s looking at trying to understand how the landscape is changing, trying to understand how to approach catastrophes to be open to new types of solutions. We’ve have many large carriers that have a lot of exposure on the ports right now, and moving forward in a business as usual sense of, say, who’s had the best underwriting performance, who’s able to, you know, create portfolios that are diversified enough to be able to cover this. That’s good. And of course, we should continue that. But also we should be open to looking at groups that are taking a different approach to this type of thing, because moving forward, we’re going to need something that acknowledges that the perils of a different climate affected perils will be fundamentally different moving forward. And therefore, we will need a fundamentally different approach to be able to solve for this long term, not just in 2024-2025. 

Pete Miller [00:25:36] I couldn’t agree more, because I see this on many sort of aspects of the business. And as you say, as data comes up and there’s many more tools available and more effective, you know, to quantify and price risk. But it requires kind of a different way of thinking and a different tool. I think that is a key. I couldn’t agree more. I see it time and time again. I do think the industry and particularly the regulators are becoming more open to that. But it’s still it’s an organizational cultural change kind of across the industry that I think needs to happen. Kevin, thank you so much for your time. I appreciate it very much. Thank you for sharing your insights with us and certainly wish you best of luck. 

Kevin Stein [00:26:22] Appreciate it Pete. Thank you very much. 

Pete Miller [00:26:26] I hope you’re inspired by the work Kevin and a team at Delos are doing to better predict which properties are exposed to changing wildfire risk and helping to mitigate that risk. I know I was. Next, I’m sitting down with another inspiring technology leader who’s developed a small wearable solution to better predict unsafe repetitive motion. Sean Petterson founded StrongArm Technologies in 2011. The company started by creating a sophisticated full body exoskeleton to monitor workers on the job and capture data about potential unsafe movements that could lead to injuries. It became clear that clients valued the data most of all, so StrongArm pivoted. Using small wearables that provide real time feedback to workers and capture insightful data for employers, the company protects industrial athletes by providing information on unsafe situations and encouraging safer behavior. Sean had some interesting perspectives to share on how his technology can deliver a big ROI on workplace safety while also motivating and protecting employees. Here’s our conversation. Talk to me about or let me know what drove you to do something to create something, to prevent worker injury and to promote safety? 

Sean Petterson [00:27:56] It’s pretty personal, I think. I grew up working my father’s construction company and he actually passed away on the job. So instead of going and taking over the family business, I went off to school to create products and I went to create products that were meaningful. And, you know, solving injuries to me was one that was a good reason to wake up and get in the shop every morning and start building. 

Pete Miller [00:28:21] StrongArm just in a nutshell, if you the like elevator speech of, what’s the mission? 

Sean Petterson [00:28:25] The mission is to use data to create a better future for industrial workers. And we do that by deploying wearable devices that capture data in real time and deliver haptic feedback to eliminate the injury before it happens, and then take that data and inserted into every aspect of the value chain so that we can put safety in the center of all conversations at a business. 

Pete Miller [00:28:45] So tell us a little bit more about your background, I mean, and how you came to start the company. 

Sean Petterson [00:28:51] You know, as a kid, I always, always like to build and tinker and come from a family that does a lot of that. Spent a lot of time with tools and went to school for industrial design. And at that time it was a, you know, a deep dive and all things product and really learning everything from, you know, initial concept being down to the finite element engineering of of many of the manufacturing processes. I really enjoyed it. And I think what I enjoyed most was finding ways to create value in areas that had been overlooked. And one of those areas was human augmentation and what we can do to preserve the future in the health of blue collar workers, frankly, that we like to call industrial athletes, we actually started out making exoskeletons. So we were looking at ways that human centered design can have a real place in an impact in the safety space. So the exoskeletons were a really, you know, flashy name, right? You wear something that sounds like you’re going to be Ironman. So it got people interested, but it worked, right? So we spent a lot of time in the labs understanding not only how to build a contraption, but also how to make this device approachable, how to make it work for multiple body types, multiple industry types, but then got into the real validation of where’s the pain point that the the customer who is buying these devices to protect the industrial athletes. How could we show it to them? And when our customer base was looking for a case study, it was very difficult to show them in our way because when an injury happens that person is now in the world of work comp. And it was very difficult to not only show a preventative means, but also very difficult to get a large enough sample size to be able to show predictive data, predictive impacts before that injury happens to prove out what this, you know, this fledgling technology can really do. And that’s where we really stepped into the next phase of the business. So we looked at the industry and saw, you know, why is it why is it so hard to get the data? And typically it comes from claims that were settled a year ago and looking retroactively at this data and then trying to pinpoint that back to the person was very challenging because that person likely wasn’t coming back to work. Or you’re an environment where the turnover is just so high and in some cases that turnover rate was 50 percent. So we needed effectively to be way faster than the injuries that are happening. So what we did is we built an algorithm based on those that thesis of how we proved out the value of the exoskeletons in the lab. And then we put that algorithm on wearable device.  

Sean Petterson [00:31:42] And what we were able to do is effectively take what’s called a time series database. So calculating the data every single second along the way and marry that to the job type, marry that to the type of work of the nature of work the environment that the person is in. And we were able to see a baseline and that in that baseline in those variables would tell us an indicator of success for the exoskeletons or in failure where there was there needed to be a process improvement in that delta right there was we found through some more math was directly translatable to an injury rate and injury cost and effectively a real time ROI. So that’s where the business started to evolve. And this was about six or seven years ago now where we had the realization and this was through the excitement of our customers effectively that we could really predict these injuries in real time. And we can not only do that in the means of preventing the injury, but we can also do it by providing. New ways to assess other areas that can be improved. And, you know, the exoskeletons only became a small piece of that. So we actually pivoted out of those. We went fully into the data. And now we’re finding ways that we’re basically taking a whole bunch of complex information, making it really easy to collect, making it really simple to interpret. And then we’re guiding our customers into means that are changing the way that they’re managing their manual workforce. 

Pete Miller [00:33:10] So when you collect data, I mean, I looked at your website and I went through the, you know, the video, which was fascinating, but you collect data pretty comprehensive, as you’re saying, like a second on a second, you know, every second about what a person wearing a device is doing and ergonomically what they’re done. 

Sean Petterson [00:33:31] Yeah, that’s correct. You know, about a half a dozen data points we’re collecting every 12 seconds and we’re marrying that down so granularly because it’s in that signature where you find the nuance and it’s in that nuance where you find the areas that you can improve. 

Pete Miller [00:33:47] So I’m curious. So I put the device on and I’m a warehouse worker and I’m up and down taking packages up and down, off shelves. And then. But there’s I’m assuming and pardon my ignorance, there’s enough commonality whether I’m a five foot person or a six foot six person that like, you know, the fulcrum arm and everything, you can measure that I’m perhaps leaning over incorrectly, right? Regardless of my physical stature, if you will, is that, you know, do I have that correct? 

Sean Petterson [00:34:17] That’s a very good question. There’s enough commonality. But we did have to do a little bit of magic in developing the product in the sense that when you the way we deploy the sensors is they go on a smart dock which has 25 sensors on it, and you go through that in the morning and you select your name, and that now has a roster of what you’re doing, which helps us calibrate the level of haptic feedback we’re going to deliver. And then when you put that sensor on, it actually takes a few minutes to assess where it is in space. So it’ll know the difference between someone that is 5’2 and someone that is 6’5. So we actually adjust for that. And then and then in those movements we can adjust exactly how we’re capturing data based on that person’s physiology. 

Pete Miller [00:35:03] So just for clarity, for everybody who may be listening, haptic feedback, can you just define that? 

Sean Petterson [00:35:09] Sure. Haptic feedback is a fancy word for saying it vibrates when something bad is going to happen. It’s like a tap on the shoulder. It’s very similar to what you’ll see in your cell phone when it’s when it’s ringing on silent. 

Pete Miller [00:35:21] We talk about? Well, I mentioned warehouse workers, but what are the other types of workers specifically, like specific examples of folks that would use that and how they could benefit from its predictive and preventive capabilities? 

Sean Petterson [00:35:34] The reason for, you know, the truth is that this science applies to really all jobs where musculoskeletal and use are high. But as far as that’s the difference between an assessment tool and an ongoing part of the day to day. And where we benefit is in the ongoing part of the day to day. And the reason for that is a lot of the nuances happen when someone’s tired, when they’re fatigued, when they’ve been, you know, they’re at the end of a long week. And it’s in those behavioral changes where we want to be there, basically like a coach to say, hey, watch out, you’re about to do something dangerous. So in order for us to thrive in those environments at scale and keep it easy for our customers, we stay pretty narrowly focused in warehousing and distribution environments where the work is repetitive, where the folks are coming back to the same location every day. And we can be real experts in that little niche of our market. It’s a big market, but you know, we want to be experts in folks that are effectively moving heavy things very fast in a very repetitious environment, typically indoors. 

Pete Miller [00:36:41] The term I love is industrial athletes. And that you mentioned the word coach. How did you come up with that? I think that’s a great term. 

Sean Petterson [00:36:50] Just talking to the folks on the ground, you know, it’s it’s it’s just about, you know, we’re coming in with really fancy technology and we’re coming into an environment that doesn’t really see a ton of positive change. And, you know, we we want to be there for that person, right? This is a no brainer for companies. And we want to make sure that there’s no barriers for anyone to not be uncomfortable with the data we’re trying to collect because this is a win win solution. We’re stopping injuries, we’re saving costs for the employer, and we’re just creating a better life for the folks that are actually wearing it, using it every day. They’re not going home hurt. We have folks that are not taking pain pills and telling us these things, they’re sleeping better. And, you know, for us, it’s let’s just own that. Let’s figure out how we can make it every day. We strive to make it easier for them to use, easier to interact with, easier to use the data to to improve themselves. And when you really get into that granularity of taking this complex data, boiling it down to something simple and then finding actionable steps to improve, if you look at what they’ve done, you know, they just lifted boxes for 10 hours a day. That’s backbreaking work, quite frankly. Right. Pun intended. And the level of rigor is very much what an athlete would do. And so we think the industrial athlete moniker is one that has stuck around and it’s one that we want to continue to pursue. And that’s how we want folks to think about it. It’s just like you would coach someone on a field and it’s like having best in class data, you know, on the basketball court that you’re using to improve your jumpshot. You know, we’re just using that same methodology, but applying it to something that is critical for you to, you know, work the job that you’re working with health and in longevity. 

Pete Miller [00:38:32] You talked a little bit about change. And, you know, change is always difficult. And, you know, as we’ve talked to different folks that are trying to implement technologies to predict and prevent thing, it’s a new way of thinking. And, you know, people have to adjust to that. What advice would you give someone what have you learned in terms of saying, look, you know, implementing these kind of technologies? 

Sean Petterson [00:38:59] What we’ve learned is that it’s not a silver bullet. Right? And it all isn’t about process and it’s not all about the next gadget. It really starts with culture. So we have to go into a place where folks are leading with a positive safety culture. And even if things are perfect and injuries are high and the operation is still getting its feet underneath it. We need to be in an environment where everyone knows that this is a mutually beneficial exercise. And for us, we help build on that culture because that is that the safety culture, keeping safety top of mind is the best solution out there. But it’s very difficult to manage what works. So we want to help create the baseline where new initiatives are new incentive programs are all building off of our baseline and we’re all looking at the same goal, which for us is something binary. It’s we call it the safety score and that’s what we collect every day. And as your safety score improves, your injuries are reduced and it’s something that everyone can strive for. So I think that that’s been the biggest learning for us over the years is that we can’t solve everything. You know, there needs to be a cultural piece of this becoming part of the behavior and there needs to be a shared goal. And we just hope to facilitate that in environments that are open minded to it. 

Pete Miller [00:40:24] That’s very enlightened. There’s a famous saying by Peter Drucker, right, the management guru, that “Culture eats strategy for breakfast”, right? 

Sean Petterson [00:40:34] It’s exactly right. 

Pete Miller [00:40:35] We’re trying to talk a lot in insurance and risk management about predict and prevent. Right? And I would be interested in your perspective on how that impacts our industry, risk management and insurance, the ability to do this kind of thing going into the future. 

Sean Petterson [00:40:52] You know, the complicated part is introducing a whole new style. Right. It’s you know, insurers aren’t selling wearable sensors. They’re not selling typical technologies. Right. So I think understanding the business relationship and how they present this into, you know, as they’re going to bind becomes a complicated part. But the value proposition is a no brainer, right? Right now, the way that we are looking at the problem is you have to wait until the end of the year to really sharpen your pencils and understand what the impacts are, whether it be the impacts of something that you’ve invested into as an intervention to reduce injuries or it’s something that you’re just looking to create a baseline for a renewal. You’re still looking at old data and what we’re able to do is provide that data in real time so you can open up the door and see what’s the health of this company. And we can look at that with more granularity than that’s ever been offered in the history of the space. And I think that allows for continuity. It allows for conversations to be had around performance, around culture and around where we can free up more investments to further on other interventions, whether they be safety training programs, whether they be more wearables, whether they be redesigning the workflows. These are all things that we want to provide the tools to just have those discussions because it’s, again, a win win win now. Right. It’s saving the industrial athletes, saving the customer money on their bottom line and just providing a less risky business for the insurer, allowing them to look at this data and understand what they just took on, and then also how they can take this complex data, dissolve it down in and then be helpful and target exactly where they can be helpful, understanding where those pain points are. And then that’s all just in the predictive nature of it and understanding really the pulse. But the added benefit is we’re delivering haptic feedback and coaching that stop the injuries before they happen. So, you know, we stay right in that service layer. We’re a great conduit for these conversations to be had. We add another layer of validity effectively to what the projections were that either the underwriter or the actuary had developed. And then we help them correlate those and look forward into what a positive outcome could be in the future. Because you can use our tools as almost in our way, a calculator so we can say, here’s what the baseline is today. This is what it costs. If we invest here, we think it’ll drop this cost to Y and that delta is your ROI. And for us that is really the mission of the business, is to just get companies to look at problems differently, give them the data to make their own decisions for the benefit of the worker and for the benefit of their own bottom line. 

Pete Miller [00:43:42] Sean, that is so crystal clear. I mean, that just makes all the sense in the world, right? It’s here’s the baseline, here’s the delta, here’s what you made and the world a better place. Kind of neat. Pretty neat stuff you’re doing. So what is next for you and for StrongArm? 

Sean Petterson [00:43:59] So for us, it’s going to continue to find ways to improve the deployment of the programs, making it easier, getting into new areas such as remote work, but also get deeper into the culture. Because as I said before, it’s that reinforcement of the culture that is really the more nebulous, harder thing to solve. You know, I have a design background and I love to invent things that are mechanical, but engineering, behavioral sciences and human interactions is such a greater challenge. And now that we have over 25 million hours of data behind us, we can experiment with these things. We can benchmark these things against different injury industries, injuries, timelines, scopes, job types, and come up with new ways to further engage. And that’s what it’s shown us. The data has shown us the more engagement we can get from the industrial athletes and also from their managers, the more successful these programs are going to be. So effectively, we’re in a place to really digitize this culture, and that’s where we’re going to continue to lean into. We’re going to offer coaching. We’re going to offer specified coaching based on new hires, based on the certain job types that they’re entering, based on the environments that they sit in, and then offer that coaching in a learning platform where that everyone can continue with [00:45:17]a C [0.1s] improve. And that’s a big differentiator from what’s been offered today because that’s been the most common intervention has been someone coming in and doing a training or in the morning. Often companies will introduce a stretch and flex. What we’re saying is make that bite sized so that it can meet the needs of the operations. So it’s not everyone taking 15 minutes out of the day. It’s folks are learning along their own pathway. And it’s also folks learning along the pathway that is very specific to their job. So our data can inform specific issues that are happening in either a new hire coming on board or maybe someone just shifting into a new role or a new shift. We’ll find the areas where that person can be improved, where haptic feedback may be delivering some benefit. But we can always further that by just training the person themselves and giving them skills that are based on that job, that are not only beneficial to the job, but skills they can take home. And it’s that full cycle of really getting embedded in how safety is something that exists inside and outside of the warehouse is where we see the future for us. And the more data that we can collect around those successes, the more we’re going to refine those learning pathways. And that’s how we’re going to continue to progress effectively our little corner in the market of digitizing and defining safety culture. 

Pete Miller [00:46:42] Yeah, that’s really cool. They say it’s like mass customization of learning and intervention in line. 

Sean Petterson [00:46:49] Yeah, that’s great. I’m going to write that down. 

Pete Miller [00:46:54] I mean, that’s that’s like, you know, that’s like, you know, the Holy Grail, if you will, of, of education, right? I mean, that’s beautiful. Really? Seriously. Yeah. Sean, thank you. I really appreciate it. I’m fired up about this. I think the what you’re doing is very visionary and really are making people’s lives better and safer. So I congratulate you. Very grateful for your time. Thanks. 

Sean Petterson [00:47:18] Thanks so much. I really appreciate being here. And it was great talking to you. 

Pete Miller [00:47:23] Well, there you have it. Today’s episode helps shed some light on how insurtechs are transforming risk management. Kevin, Sean and their teams are making strides to use technology to capture better data on risks, making it possible to identify opportunities to predict and prevent losses. It was great learning more about their approach and what they’re working on at the moment. Big thanks to Kevin and Sean for coming on the show, and thank you for listening. I hope you learned something new and inspiring from this episode. Predict & Prevent is a podcast brought to you by The Institutes. Subscribe on your preferred listening platform and join us for future episodes where we continue to dig into this approach and the opportunities that come with it.