From Unmanned Systems Magazine: How can insurers benefit from using drones?
Insurers fundamentally sell and must deliver on promises. So, we expect them to be rock-solid, reliable, and conservative. And yet we also hope that they will innovate, adopt new technologies and adapt accordingly. Each carrier strikes their own balance, which is reflected in their varied approaches to unmanned aircraft.
In 2016, the Federal Aviation Administration relaxed restrictions on commercial use of drones, opening the floodgates to opportunity. Insurers that had been dabbling under exemptions from prior rules, and those new to the party, started or expanded their efforts to integrate this emerging technology into their operations. In parallel, the set of UAS solutions has been expanding in breadth, functionality, and depth. Importantly, as costs have declined, the range of economically viable uses and deployments has also expanded.
FAA rules and the related processes for obtaining waivers or authorizations are changing, enabling additional use cases and expanding the value of existing uses.
An example of a technology-enabled process improvement is the FAA collaboration with industry to establish the Low Altitude Authorization and Notification Capability (LAANC). LAANC is "an industry developed application with the goal of providing drone operators near real time processing of airspace notifications and automatic approval of requests that are below approved altitudes in controlled airspace." In turn, the FAA's three listed approved service providers, AirMap, Project Wing and Skyward, are able to bring this capability to users. BetterView and AirMap announced an app-based solution to provide digital authorizations for drone flights in U.S. controlled airspace on Jan. 30 and Kespry reports that they will also have their own LAANC/AirMap integration live soon. Previously, this took up to 90 days, making drone inspections impractical in those restricted airspaces.
Drones can provide consistent, high quality aerial images, enabled by preprogrammed best practices while eliminating or reducing the need for physical inspections, ladder assists and their associated costs. Follow-on analytics can identify damage, dimensions, and through systems integration, create repair or replace estimates with streamlined workflow. Increased field and desk adjuster productivity improves cycle times and enables a remixing of skills and assignments. Watch for improved risk assessment, more accurate claim payments and lower loss adjustment expenses.
Decisions around scope, scale and operational integration will drive the ultimate value an insurer receives, but the cost to explore can be minimal. We should expect multi-phased plans beginning with research and development.
To understand how much value drones can and will create, let's explore the drivers as if we were outlining a business case.
The scope: The inspection volume will be driven by the scope definition — that is, the lines of business, the inspection type and its intended use — and the assignment criteria. Some obvious candidates include tall and steep roofs; hard-to-access or dangerous areas; geographically disperse locations (fields, large buildings and complexes); or when speed and efficiency are paramount, such as catastrophe response or assessing hail damage. The return on investment for claims is the most straightforward, but underwriting and loss control shouldn't be overlooked.
Loss costs: in general, insurers can impact loss costs through a combination of accepting good risks, pricing appropriately, identifying opportunities for the insured to remediate issues (overhanging branches, pooling water) and paying claims accurately. By applying drone-enabled advanced analytics, adjusters can also receive pre-identified damage assessments (pre-existing or claimed damage), wire-frames with measurements and more. This can enable both better decisions (underwriting and claims) but also free staff up to focus on the exceptions and complex judgement calls. Key to this is the quality of the underlying data and analysis — such as images, wireframes and reports.
Loss adjustment expenses: Several carriers and service providers report an increase in field inspector/adjuster productivity from three to four inspections per day to as many as a dozen per day, reducing the need for ladder assist (at a cost of around $150 to $250 per inspection) and cutting fall exposures. Drone-enabled inspections and analytics also enable increased desk adjuster productivity.
In addition, drone-enabled productivity speeds up claim investigation and resolution, leading to a more positive claims experience and ultimately an improved brand image for the insurance companies.
Simple, or complex
While the above may sound complex, it need not be. For carriers that want a simple point of entry, they can order drone-based images through services such as Geomni or WeGoLook. This can serve as either a replacement for or complement to a traditional or ladder-assist inspection. The insurer can request just the raw photos or a report with photos. Other carriers may want to take advantage of the enhanced functionality provided by drone solution and service providers including: Orthomosiacs, 3-D and interactive images, damage identification, comparison to prior states, measurements and wire frames, heat maps and alternative image capture. And, through integration with other systems such as Xactimate and Guidewire, they can produce repair/replace estimates and deliver content into the claims workflow.
Drone service providers such as BetterView, EagleView, and DataWing offer a more robust suite of field operations, flight operations and data and analytical services, minimizing the burden on insurers while delivering enhanced value. Some insurers will elect to develop in-house drone capabilities, possibly in concert with complementary third-party service providers. We have seen these hybrid models in the IT world, so this isn't entirely new ground.
There are plenty of articles in the media about large carriers deploying drones: Allstate, Farmers, Liberty Mutual, State Farm, Travelers, USAA and others. Discussions with drone service providers identify a number of small and mid-sized carriers that are using drones as well, but they aren't authorized to share names. Many carriers called their drone programs into action in the aftermath of the hurricanes, floods and fires of 2017. Drones enabled increased adjuster efficiency and were able to take photos where people simply couldn't reach.
With some consistency, service providers report insurers are currently focused on using the high-quality photographs a drone can provide, but are not taking advantage of available higher-end analytics. By testing the incremental value enabled by drone-enabled aerial imaging, carriers report that they are gaining familiarity with the technologies and solution providers, laying the groundwork to make informed decisions about how to integrate aerial imaging and AI into their operations. From there, we can expect carriers to focus on integrating more functionality and use cases, to optimize transactions (e.g., pre-identify damage, generate estimates, automate processes) and ultimately to develop more strategic actionable insights (e.g., patterns, trends, co-variances).
On this point, Landon Phillips, cofounder of DataWing, shares a bit of a reality check regarding current uses. Frequently, DataWing works with insurers' "innovation teams" — mostly centered around hail and wind claims — a use case that was proven across the industry during 2017's hurricanes Harvey, Irma and Maria. DataWing also works with carriers to facilitate desk adjusting, "to capture the data and bring it back to the adjuster who operates at a higher price point than the drone operator. This increases adjuster productivity to 10 claims per day, from maybe four claims per day [plus or minus]."
However, while some carriers request orthomosaics and other advanced analytics, the vast majority desk adjust based on raw photos. Undaunted, DataWing is preparing for the future by working to solve issues around automated claim adjusting in collaboration with IBM Watson.
Similarly, Nathan Stump of Kespry reports that while Kespry offers higher-end analytics, including a hail detection algorithm, "far and away the most valuable thing that we deliver is the high resolution imagery, and that the drone captures the entire surface of the roof."
Highlighting the value of having a complete image of a roof, Stump points out that "with a manual inspection, there are frequently areas that are inaccessible [roof edges, very steep portions]. Therefore they use a Test Square and extrapolate. The Kespry drone provides a more accurate, complete inspection."
WeGoLook's Devin Loehrs points out that "now that drones have more sensors, they can fly more closely to the roof and can capture more detail," such as a crease in an asphalt shingle.
While this may be the case with some carriers, one commercial insurer that prefers not to be named is using a scoring algorithm to identify recent hail damage. Their senior vice president of operations, claims and loss control was paying $200-250 for a physical roof inspection. Now, working with BetterView, the insurer is seeing savings up to 75 percent, plus much improved turnaround times. The SVP also reports fewer disputes with those being insured. There is "such high quality of images that there really is no contesting of findings," he says.
The scope, scale and value of use cases will likely expand as carriers become more familiar and comfortable with the technologies and confident in their capabilities. In parallel, regulatory constraints will become less constraining and solutions will become increasingly robust.
As one potential outcome, consider what would be required to process a low value hail claim automatically.
Make assignments: For instance, DataWing's SmartSky application checks flight conditions, pilot availability and automate assignments, including route optimization.
Perform inspection: Kespry produces fully automated drones, and Airware software automates the use of DJI and other drones. A Part 107 certified pilot is still required, but the flight plan is automatically generated and run. The images are uploaded to the cloud, analyzed within an hour of flight and available to a desk adjuster. By embedding best-practices into flight operations software and flight plan, consistency and quality are automated.
Automate decision-making. IBM is working with several large clients to provide image analysis using Watson Visual Recognition Insights, drawing on its in-house Watson artificial intelligence. Kespry offers image analysis that identifies hail damage and has the ability to integrate with estimatics software. Dolphin AI has similar drone-agnostic capabilities as well. Add in additional context (hail history, fraud risk assessment) and artificial intelligence or a rules engine, and claims that qualify will be ready for a settlement offer.
Integrating the workflow: The workflow is integrated with an insurer's core systems, such as policy and claims administration systems. This enables the full claims workflow, including document management, reserving, audit and exception processing.
Turning this into reality may not be easy, but the elements all exist. Who will have the vision, will and skill to implement the next step in value creation? What other opportunities exist?
The SVP at the previously mentioned commercial insurer says he resisted pressure to adopt drone technology faster than he was prepared for. He recommends that insurers first think about their business operations, then consider how drone-related technologies can enable what they do. It sounds straightforward, but it is a discipline that you don't always see.
Jim Rosen is the president of Ridgeline Consulting, and specializes in creating value for InsurTech companies, Insurers and their clients. He can be reached at firstname.lastname@example.org or +1 917-754-9938.