Computer Vision Means Satisfied Customers

Insurers need to find a way to speed up claims — and fortunately, advanced computer vision can provide insurers with the means to do just that.

A man sitting at a desk on a computer

Giving the customer what they want — or need, depending on whom you're quoting — is a basic tenet of business. And perhaps the one thing all customers want — and need — is a good experience with the company they are working with.

It's a truth that the insurance industry needs to understand. One of the biggest complaints consumers have against their insurers is the often lengthy time it takes to settle claims. And dissatisfied customers often turn into former customers; studies show that over 60% of customers are likely to switch to a competitor after just one bad experience — while more than three-quarters would jump ship after just two such experiences.

If customers equate slow claims processing with poor service, insurers need to find a way to speed up claims — and fortunately, modern technology, in the form of advanced computer vision, can provide insurers with the means to do just that. These advanced computer vision systems use cameras and visual detection systems in vehicles, drones equipped with high-resolution cameras and images taken by devices such as phones and tablets. The images are then automatically uploaded to machine learning-based AI systems that analyze thousands of data points — type of damage, estimated repair costs, structural integrity (of buildings or vehicles) and much more.

These AI-based systems — relying on advanced data analysis and using machine learning — are able to develop a clear basis for claims far more quickly and accurately than with the traditional method, where human adjusters must visit the scene of the claim and human actuaries must determine the amount to be paid. Companies can thus cut down the investigation phase of a claim from weeks to minutes — rapidly approving settlements and ensuring high customer satisfaction and retention.

Investigations of structural damage from natural disasters, fires, hurricanes and similar events — where conditions make it difficult, if not impossible, to send out adjusters — become much more efficient with computer vision technologies. Drones can much more easily navigate these disaster scenes, flying into corners and crevices and recording images over, under or inside damaged structures. AI-based systems receive and examine the images, determining what damage was caused by the event and what damage may have already been in place. The analysis is based on damage models — patterns of damage based on specific events associated with disasters, such as how a roof would appear if it were blown off by a gust of wind measured at 75 mph — as well as with data on other similar incidents. With all the data taken into account, the system can quickly provide a full assessment of the damage and determine how much the insurer needs to pay to satisfy claims.

Computer vision can also streamline the claims process for incidents involving vehicles — often a sore point for customers, and the source of the largest number of complaints by policyholders. Here, too, the traditional process of claim evaluation — with adjusters sent out to evaluate the condition of a vehicle, the circumstances of an incident (traffic, weather, road conditions, etc.), along with statements by those involved in the incident, as well as police — becomes a time-consuming project that taxes the resources of insurers and leads to undue delays in claim settlement.

See also: Customer Segmentation Is Key

Images taken right away with a customer's mobile phone camera, along with data collected from sensors, road cameras and images recorded by in-vehicle cameras, are collected and analyzed by AI models, which provide an accurate picture of the circumstances of an incident, along with the liability of drivers and the amount of money the company needs to pay. Neural networks running on the customer's mobile device can provide real-time guidance to ensure suitable images are captured that can be used as evidence as part of the claim process.

Computer vision systems can also make sure that insurers get a clear picture of the condition of a vehicle before they even issue a policy — with previous damage noted and excluded from the policy, ensuring that they, too, are treated fairly in their relationship with their customers, eliminating potential fraud on the part of customers.

In addition to increased efficiency and speedier claim settlements, computer vision-based systems ensure that insurance processes are more likely to be perceived as fair by customers — increasing the level of customer satisfaction. Unlike human adjusters, computer vision systems are unlikely to miss small details that could end up being crucial to ensure an accurate evaluation. Customers will appreciate that insurers are doing everything possible to make sure that they get all the money they are entitled to as quickly as possible — and will reward their insurers with continued loyalty.

The technology to implement these systems is available right now, and as the technology proliferates, the price of systems continues to fall. That insurance companies have not yet widely adopted these systems yet is understandable  insurance is a very conservative business, and companies need to ensure that stockholders, stakeholders, and regulators are on board with major changes to the claims process. But implementing computer vision systems for damage assessment is worth the effort; companies will avoid wasting time and money, and ensure that they use their resources as efficiently as possible — while customers will get their money faster, increasing their satisfaction with their insurers and ensuring that they remain customers for many years to come.


Neil Alliston

Profile picture for user NeilAlliston

Neil Alliston

Neil Alliston is executive vice president of product & strategy at Ravin AI, a startup offering computer-vision and AI-based solutions for vehicle inspections. He has a wide range of experience working on machine learning in the transportation and logistics sectors.

Read More