How GoMechanic Is Streamlining Car Insurance Claim Using A.I

Bharat chandra sahu
GoMechanic
Published in
4 min readJul 20, 2021
Artificial Intelligence X Car Insurance

We have always believed in backing consumer-facing services with technology. It just makes the whole process much straightforward. The Tech Team at GoMechanic is upto all sorts of all good, this time tweaking a bit of top-notch AI tech in the field of motor insurance. The objective? To streamline the insurance claim and damage assessment process. Here’s what it is all about:

First, The Objective

So, why are we doing this? What is the driving force behind merging AI into insurance?

Motor insurance forms the largest sector of the general insurance industry in India. The car insurance sector is valued at ₹70,000 crores with insurance claims peaking at a staggering 1.2 lakh crore nationally [1].

And claiming car insurance is not an easy process. There are multiple processes involved right from the claim intimation to the settlement. We at GoMechanic, believe this shouldn’t be this way.

Insurance Claim + AI = 2X Faster Settlement

The currently under development project is expected to be capable of assessing the magnitude of damage and eventually calculate the estimated cost of the damage caused.

With this too, the car owner can upload images of their car post-accident from all the angles i.e. (Left, Right, Front, Rear, Front-Left, Front-Right, Rear-Left, Rear-Right). It is advanced enough to predict the severity of damage and also conclude whether that part is required to be replaced or repaired along with a calculated estimated cost. Estimated costs also include labour cost and paint cost.

How Does It Work?

  • Upon receiving the claim survey request, The AI architecture extracts data like Registration number, Policy number, Insurance expiration date, insurance type, etc. of the affected vehicle.
  • If the claim number is already provided, then our system stores this data with the claim number. If a claim number is not provided, then the system generates a claim number and stores the above-mentioned data with the generated claim number.
  • To select the area for which the claim is demanded, Workshop\Insurance Company can access our dashboard and enter the claim number to select the claimed area of the car to the corresponding claim number.
  • After this, the system will ask some queries related to the workshop and surveyor to take note of all the necessary credentials of the parties involved.
  • Once done, one shareable link will be generated. This link is for uploading the images of the affected car to be shared with customers, workshops and surveyors. This link will expire in 24 hours.
Link Generating Dashboard
  • A person with the shared link will open that link on his/her mobile phone and will upload the images from all angles. Now here’s a catch, the product is trained to identify a distorted or false (Images of some other car). Finding any image belonging to these types will prompt the system to reject and delete the image.
The images are uploaded here
  • After successfully uploading all the images the product starts assessing the orientation of images to deduce the intensity of damage caused to the vehicle.
  • After completion of damage detection, The system will provide you with the estimated repair or replacement cost. Users can select OEM or OES spare parts in case of replacement.
How A.I generate the repair estimate
  • If in case the product didn’t find the damage significant enough or if there is a need to edit the uploaded images, then the surveyor or workshop technician can draw/edit the damaged area (image) in our dashboard. The estimate for the same will change accordingly.

It’s A WIN-WIN Situation

GoMechanic’s AI-based Damage Detection is a revolutionary tool that will help both the insurance company and the workshop to calculate the damage estimate. We have around 60L spares parts (OEM + OES) and their prices as backing data. Data for labour and paint costs for all Indian cars and their spares parts has also been integrated into the product enabling the product to provide accurate estimates for the damage caused to the vehicle.

--

--