Drones for insurance claims on roofs?

"Knock, knock"

"Who's there?"

"I'm a drone"

"I'm a drone who?"

"I'm a drone here to inspect your roof"

In recent years there has been a lot of discussion on using drones for inspecting roofs following hail, wind, or other damage. I've had many conversations on drones with my insurance adjusting students, former colleagues, and friends in the industry. So, in this article I want to share my thoughts on this topic and explain the realities and concerns a homeowner, adjuster, or roofer should have using this technology. For the sake of this article, I am discussing drones for inspecting asphalt shingles on residential roofs (not commercial). Asphalt shingles make up the overwhelming majority of residential roofs in the U.S. at around 80%. This can vary regionally, but here in North Texas most homes are covered with asphalt shingles.

Before getting started, be it known that I inspected my first roof for hail damage in 1986 when I was a new property adjuster at State Farm at age 22. I'm old school. Since my rookie year, I've inspected thousands of roofs as a staff adjuster, independent adjuster, trainer, and contractor - all using the traditional method of propping up a ladder, walking to the top of a building, and going to work. This includes steep roofs too. In my younger days, there wasn't a roof I couldn't climb and I was proud of that.

Early in my career, I was taught by the best of the best that a thorough roof inspection means climbing to the highest peak (to take a picture and prove I was on the roof), then physically inspecting every slope (or facet) of the roof. I was taught to use my eyes and hands to check for damage. I want to emphasize 'eyes and hands'. Years of doing this has taught me much and I have gained valuable experience with both hail and wind damage.

Naturally as contractor and trainer, I have a vested interest in how drones are used to help adjusters and roofers inspect houses following a claim. Knowing what I know after decades in the field and after thousands of roof inspections, I have opinions. Valid opinions. And, I love it when software people try to reinvent the claim inspection process and work hard to convince others a new technology is better. Sometimes it is better. Sometimes it's not.

I am going to have an open mind about drones in 2018 and give it a chance, although a slight chance. But, in short, drones should only be thought of as another tool and used only with strict guidelines. Their use should also mean that any insurance company decision to deny or "repair" a roof should never be made using only a drone. There, I've said it.

First, lets walk through a scenario. Two homeowners residing side by side receive hail during a recent storm. They suspect hail damage because they see some small dents on their downspouts. Both submit an insurance claim on their roof. Let's say both houses were built 10 years ago, face the same direction, and are very similar otherwise. Both houses have the same asphalt shingle roofs too (original roofs). Both houses are insured by different insurance companies.

A few days pass and one homeowner's house is inspected by an adjuster using a drone. In this scenario, the adjuster did not get on the roof. Instead, this adjuster used a drone to fly over the roof taking photos. Perhaps, the adjuster propped up his or her ladder and physically checked some shingles nearest to the ladder but he or she did not leave their ladder. The adjuster completes the roof inspection in 10 minutes using this method and decides the roof is not totaled. That insurer offers a repair.

Our second homeowner's roof is inspected by an adjuster (with another insurer) who does not use a drone. This adjuster inspects using the traditional method of getting on top of the roof and inspecting all the slopes physically with their own eyes and hands. This adjuster is physically checking for hits, bruises, and looking at all suspicious areas with two eyes, fingers, and thumbs. A thorough inspection also requires (at times) for the adjuster to take a knee on the roof to gain a better view which this adjuster does. Perhaps this adjuster lifted a few shingles and checked for broken shingle mats. The adjuster marked several test square areas with chalk too. At the end of this 45 minute roof inspection, the adjuster's finger tips are dirty. This adjuster determines the roof cannot be repaired and so replaces the roof. Quite a contrast between the other adjuster who used a drone.

Oh, and one homeowner is happy and the other furious.

Both of these scenarios depict possible differences in outcomes with these two methods for inspecting a roof. There is a big difference in the quality of inspections too. Imagine being a 3rd neighbor watching it all. At one house, a drone is seen flying overhead while the adjuster is on the ground. Its cool but also over in 10 minutes. At the other house, an adjuster is on the roof working for 45 minutes (on the roof). One uses the eye of a camera flying overhead. The other puts the adjuster physically on the entire roof doing a thorough, methodic, and systematic inspection . In which scenario, would you trust a final decision? What does that 3rd neighbor think? I supposed these answers depend on the outcome, right? In other words, would a homeowner mind if we used pigeons with cameras so long as they were getting a new roof? Probably not.

In fairness, let me state that just because an adjuster gets on a roof it doesn't mean the roof will be a total loss or gets totaled. Disagreeing over roof damage has and will always will be a possible issue between adjusters and roofers. I learned that in 1986 and nothing has changed, except now I am on the other side of the table. This brings up another point. If an adjuster on the roof can miss damage, why wouldn't a drone? Hmm.

Knowing what I know about hail and wind damage, you cannot inspect every residential asphalt roof claim thoroughly with just your eyes alone or through a camera lens. Let me be more specific here and say this applies more on hail in fringe areas of a town and wind damage that appears marginal to less. What I mean by fringe areas are those areas in town where the hail was small but might cause some minor or light damage. Keyword: might. This is why its important to send veteran or experienced adjusters to the fringe areas to make the right decision. But, when hail is large as in tennis, baseball, or bigger, my 10 year old could identify hail damage.

On fringe hail claims, my fingers are brown and black at the end of each inspection because I am checking for suspicious areas on the shingles. Some of the time these areas turn out to be non-hail and other times it is hail. I'm also looking for collateral damage to other objects on the roof and tasks like rubbing chalk on metal to document dents or no dents can't be accomplished with a drone.

On wind claims, I cannot tell you how many times I have inspected and found where asphalt shingles lifted up through their fasteners and then laid down. Often those shingles were loose but resting in their original rows. Using the naked eye often shows nothing wrong and so I've seen inexperienced adjusters miss this type of damage. Its only until one lifts up rows of loose shingles will this type of damage be discovered. A drone certainly would not see this nor would a drone be able to lift 3 to 4 rows of wind damaged shingles for photo documentation. A drone helps someone on the ground see missing shingles but not other shingles damaged around it which have pulled through fasteners and laid back down.

So from my field experience, I would never tell a homeowner with an asphalt shingle they have no damage or offer a 'repair' to a roof if my inspection was done solely by a drone. As a roofer I couldn't stand behind that decision. I don't think any insurance company can in good faith deny a roof claim if the inspection was done solely by a drone. Insurers owe a thorough or at least reasonable inspection and drones currently don't offer this that I am aware of. The last time I checked, drones didn't have hands, fingers, and an ability to lift loose rows of shingles. Drones can't draw chalk marks. Drones can't check for multiple layers, felt, drip edge, metal under valleys, and more. People can do these tasks but drones are cutting edge, apparently.

Lets go back to our earlier scenario of the 2 neighbors who filed roof claims and paint another situation. Let's say, instead, that both houses had baseball size hail (or something large enough that leaves no doubt in photos). Let's say too that both houses had significant collateral damage to elevations including broken windows, broken exterior lights, dents to garage doors, air conditioners, and more. Often seeing extensive elevation damage suggests good roof damage is anticipated. In this situation, lets say both neighbors are going to receive new roofs from their insurance company so its just a matter of getting the adjusters on scene, take some photos, check other damage, explain policy coverage, etc.. In this situation, by all means use a drone if it helps.

If an adjuster knows they are going to total the roof, I see no issue with using a drone in this new example above... so long as all the accessories are correctly noted. An adjuster will still have to inspect for felt, drip edge, number of layers, valley metal, and other objects. But, if this helps keep adjusters off more roofs and safer, then that's a good decision. And in reality a lot of these drone decisions are safety driven and/or production driven.

But, using a drone to document a decision to total a roof requires insurers to change their inspection requirements. During my nearly 31+ years in this industry, insurers have historically required their adjusters on hail claims to mark off sample test squares on a roof using chalk. A test square is a 10' x 10' square on the roof. Usually, adjusters create at least 4 of these sample test areas to represent slopes facing north, east, south, and west. Some insurers want more test areas though and a few insurers may require less. This is one reason of why I wish the insurance industry had standards but that's another blog article.

Within each sample test square area, the adjuster does an analysis that involves marking hail damage, identifying mechanical damage, foot damage, brittleness, etc. Since its not feasible to inspect every single shingle on a whole roof, adjusters are supposed to inspect every single shingle within their sample test squares. The findings of these test areas (e.g. the number of hail damaged shingles ) are supposed to represent hail damage expected on the entire slope from where that sample test square was created. Each insurer has their own requirements for how many hail damaged shingles are required in a sample test square before they total a slope.

After completing their analysis on sample test squares on the various slopes, the next step is to determine if an entire roof is totaled. To total an entire roof, adjusters must determine either how many slopes are totaled and/or what % of the entire roof area is totaled. Usually, its both. Each insurer has their own formula which is part of the reason why sometimes neighboring houses are not all totaled.

Since there are no claim standards, insurers can create their own requirements for how many hail damaged shingles per sample test area are necessary before a slope is replaced. And, insurers create their own guidelines for what % of the roof is totaled before replacing an entire roof. In my experience, I have seen insurers require anywhere from 6 to 10 damaged shingles per test square before totaling a slope. I have seen insurers require anywhere from 51% to 75% of the roof to be totaled before replacing an entire roof. And, that's how it work, most of the time.

Essentially, what I'm saying here is there is a lot of work that has to be done by the adjuster to reach a decision to replace, repair, or deny. Doesn't this sound like a lot of work? This is why most roof inspections for hail should likely take 45 minutes or more, depending on the size and issues of the roof. This assumes the adjuster is creating their own drawings and measurements too. It could take less time if they already have that data. The same should apply to roofers as well but some only see 'totals'. Not all but some.

To use drones for obvious totals would be feasible but are insurers going to abandon the time tested physical inspection on roofs which are not obvious totals and go entirely to drones? I sure hope not. Own a drone as a tool? Yes. But, experience should guide us for when and how to use it. And, keep a ladder too.

There is a lot going on with drone technology though and drones will get smarter, no doubt. They will be used more for tasks such as getting roof drawings and measurements. In fact, its already being done for this and I think this purpose is fantastic. I plan to adopt drone technology for this particular data if it's instant, saves time, and accurate.

Drones are also being used currently for other tasks including insurance underwriting, inspecting fire damaged roofs (safety issues being the obvious), and much more. These are great examples of how drones can be used to benefit insurers, adjusters, and contractors.

There are worries with flying a drone aside from just the accuracy of evaluating all hail and wind damage. These include the worry over causing damage to other people's property and/or breaking it. Drone insurance is available because there is still a risk here. Invasion of privacy is another potential issue but one I won't go into within this article.

In conclusion, I plan to have an open mind and am anxious to see how the use of drones advance. But when I get a drone, please don't call me a "pilot". My nephew is a real pilot for an actual airline that flies passengers across the country at 30,000 feet. So, is it just me or does something seem wrong here. Every time someone who has a drone refers to himself as a 'pilot', I just laugh inside. I'm just not there yet. But, I'm old school.