Insurance drones, boots on the ground, and big data

Drones are often touted as being the answer to a great number of modern day challenges. Soon, we are told, they will be making shopping deliveries for us, dropping off pizzas, and even taxiing us around. 

A similar expectation about drone capability has been seen in the world of insurance. It was thought that insurers would have large drone divisions, allowing them to easily assess large or hazardous properties, or make claims assessments on otherwise difficult to view property. But this has not happened for a few reasons. 

There are not many commercial drone pilots who also happen to be qualified assessors, or vice-versa. As a result, when insurers have made use of drone technology, they have found themselves sending two people to do the job (a pilot and an assessor), doubling their costs. 

But the primary issue is the strict regulation, certainly in the UK, that governs the flying of commercial drones. Air traffic laws state that drones can only be flown at certain heights and must remain within sight of their pilots, with all manner of permissions and licenses being needed, especially in densely populated areas. These controls and associated costs have ensured that commercial drone use for insurance has remained grounded. 

However, the destructive effects of nature have seen this change. 2017 was a year of extreme weather, resulting in record levels of property damage and  costs for American insurers and global markets like Lloyd’s of London. In the event of extreme and extensive property damage it is important for people, businesses and whole economies to begin the process of rebuilding as quickly as possible. This includes a responsibility on insurers to respond quickly and effectively. Obviously drones are a rapid and effective solution to assess such large-scale damage. A clear indicator of them being employed was evidenced in the fact that US registrations for commercial drones to carry out appraisals shot up following the 2017 natural disasters. 

If drones are becoming part of the insurer’s toolbox, what can we expect this to look like? Are we finally to see insurers acquire and fly fleets of drones? 

Probably not. With the appearance of drone service models, it seems that insurers are less likely to invest in their own drone infrastructure. And the need for both a pilot and an assessor, and the cost that entails, remain. Realistically, the requirement for drone services will be limited to specific situations. 

There are companies that are looking to alleviate these challenges however. At our main customer conference last year we met with Kenneth Knoll, COO of US insurtech WeGoLook, who have developed a novel solution using a gig economy model and intelligent software. 

WeGoLook’s insurance proposition works through a crowdsourced platform of “Lookers” who are available to complete on-demand field services, such as claims inspections. Their agents collect field data, photos and videos via the WeGoLook app. The app works by taking and processing images captured from a smartphone, using the data to make assessments of the damage, and feeding this back to the insurer. In this instance, the application does most of the assessing, utilizing artificial intelligence to make judgements on the severity of the damage and the likely cost of repair. 

This solves the cost issue of having to send a pilot and an assessor to the same job. It also solves the problems of scalability and shortening incident response times. “We live in a world of immediate fulfilment,” Knoll explained. “In a crisis situation, using workers contracted through an on-demand provider allows carriers to have boots on the ground within a matter of hours, instead of days. This means that claims are processed faster, and policyholders receive payments sooner.” 

WeGoLook’s approach to assessments highlights another key factor that will be instrumental in whether drones become an effective tool for insurers, the use of artificial intelligence. Drones undoubtedly have the potential to make the collection of vast amounts of data much easier, but the interpretation and practical use of that data by insurers is what counts. Here the use of new deep learning techniques, capable of combing through large volumes of visual data, and offering insight to insurers, must come into play. 

Whether drones will become a critical tool in the insurers’ arsenal is to be seen. There have been some limited successful use cases, where circumstance or geography make them an attractive option, but mainstream adoption is some way off yet. What is certain, for drones and in insurance more generally, is that insurers who do not make proper use of the data they collect will fail. The adoption of intelligent machine learning techniques to process policyholder and claims data is gaining momentum and those that do not take action now will be left dead on the ground as the rest of the industry takes flight.

Follow me on Twitter here.