DETECTING HIDDEN, PREEXISTING AND FRAUDULENT DAMAGES
At our last global webinar, What’s Next: A Deeper Dive into Digital Transformation, attendees submitted questions to Solera’s leadership team. Today we’re answering some of those raised on how Solera’s Modern Workflow approaches detecting vehicle damage with digital solutions.
Will making the claims process “too easy” through guided steps that explain the approval process encourage/facilitate fraud? How will image capture differentiate between preexisting and accident-related damage?
In our previous blog, we touched on our future use cases. One of the first use cases we are bringing to the market is Claims Check, which gives the insurer the ability to collect images of the vehicle’s condition at policy inception. Tackling one of the first instances of fraud for many countries, to confirm the vehicle in question does actually exist. If a claim is then made at a later date for any accident damage, the driver would use that same image capture technology to once again take images of the vehicle and damage at FNOL.
Our AI technology then uses the vehicle’s pre and post images to instantly detect what damage was preexisting and what damage was caused as part of the accident, resulting in reduced fraudulent claims and ensuring insurers only pay for accident-related damage on the vehicle.
What is Solera’s AI and machine learning solution’s capacity to identify and detect hidden damages?
If you take images at face value for the damage you can see, you will never have an accurate assessment. But combining the images with Solera’s data science and repair science allows our solutions to better understand the potential damage that cannot be seen by the camera.
By leveraging Solera’s wealth of historical repair data, our solutions are trained to predict—based on the type, level and size of the visible damage—what hidden parts of a vehicle may be damaged and/or need replacing.
We can create a cluster of claims that previously existed for a specific vehicle with a specific type of damage. Using those analytics, we’re able to preload an estimate based on the estimates of previous repairs to see down to the nuts and bolts level what may be damaged.
There is no replacement for human interaction, and this technology is not meant to replace experienced employees. Someone will always need to look under the hood to check and verify the extent of damage, but our intelligent AI and repair science can support all parties involved by saving time and boosting efficiencies.