Methods and systems for automatically predicting the repair costs of a damaged vehicle from images

Patent metadata

Field	Value
Publication number	US-11106926-B2
Application number	US-202017081656-A
Country	US
Kind code	B2
Filing date	Oct 27, 2020
Priority date	Jun 15, 2018
Publication date	Aug 31, 2021
Grant date	Aug 31, 2021

Abstract

A system and computer-implemented method for automatically predicting the labor, hours, and parts costs for repair of a vehicle includes receiving one or more images of the vehicle from a policyholder. A damage assessment model is accessed. The damage assessment model corresponds to features of vehicle damage based on a plurality of damaged vehicle images contained in an image training database. The damage assessment model is compared to the images of the vehicle and vehicle damage is identified based on the images. In addition, in response to identifying the vehicle damage, total labor costs, total parts costs, and total hours for repair of the vehicle are predicted based on the associated total labor costs, total parts costs, and total hours for repair data contained in the historical claims database.

First claim

We claim:

1. A computer-implemented method for automatically predicting the labor, hours, and parts costs for repair of a vehicle, said method comprising:
receiving one or more images of the vehicle from a policyholder;
accessing a damage assessment model corresponding to features of vehicle damage based on a plurality of damaged vehicle images contained in a historical claims database;
comparing the damage assessment model to the received one or more images of the vehicle;
identifying vehicle damage to the vehicle based on the received one or more images of the vehicle;
in response to identifying the vehicle damage, predicting total labor costs, total parts costs, and total hours for repair of the vehicle based on associated total labor costs, total parts costs, and total hours for repair data contained in the historical claims database; and
determining whether the vehicle is repairable based on comparing the predicted total labor costs, total parts costs, and total hours for repair of the vehicle to a threshold and wherein if the vehicle is determined to not be repairable, processing a damage claim indicating the vehicle is a total loss.

2. The computer-implemented method in accordance with claim 1 , further comprising presenting to the policyholder an estimated cost for repair of the vehicle, including the predicted total labor costs, total parts costs, and total hours for repair of the vehicle.

3. The computer-implemented method in accordance with claim 2 , wherein the one or more images of the vehicle received from the policyholder correspond to the damage claim.

4. The computer-implemented method in accordance with claim 2 , said method further comprising processing the damage claim based upon the predicted total labor costs, total parts costs, and total hours for repair of the vehicle.

5. The computer-implemented method in accordance with claim 1 , wherein the damage assessment model includes a machine learning program trained to identify damage to the vehicle.

6. The computer-implemented method in accordance with claim 5 , further comprising training the machine learning program utilizing the plurality of damaged vehicle images contained in the historical claims database and corresponding metadata.

7. The computer-implemented method in accordance with claim 5 , wherein the machine learning program uses a plurality of regression operations.

8. The computer-implemented method in accordance with claim 7 , wherein each of the regression operations is a linear regression operation.

9. The computer-implemented method in accordance with claim 1 , further comprising receiving policyholder data corresponding to the policyholder from a policyholder database.

10. The computer-implemented method in accordance with claim 9 , wherein the one or more images of the vehicle received from the policyholder correspond to the damage claim, the method further comprising initiating the damage claim for the policyholder using the received policyholder data and the predicted total labor costs, total parts costs, and total hours for repair of the vehicle.

11. The computer-implemented method in accordance with claim 1 , wherein the damage assessment model is configured to infer internal damage to a set of parts of the vehicle that are not visible in a received one or more images of the vehicle.

12. A system for automatically predicting the labor, hours, and parts costs for repair of a vehicle, said system comprising:
an image training database including a plurality of damaged vehicle images and corresponding metadata;
a historical claims database including a plurality of claims, each claim of the plurality of claims associated with one or more of the plurality of damaged vehicle images and the corresponding metadata, each claim including total labor costs, total parts costs, and total hours for repair data;
a damage assessment model corresponding to features of vehicle damage based on the plurality of damaged vehicle images in the image training database; and
a processor coupled to said image training database and said historical claims database, said processor programmed to:
receive one or more images of the vehicle from a policyholder;
access the damage assessment model;
compare the damage assessment model to the received one or more images of the vehicle to identify vehicle damage to the vehicle based on the received one or more images of the vehicle;
in response to identifying the vehicle damage, predict total labor costs, total parts costs, and total hours for repair of the vehicle based on associated total labor costs, total parts costs, and total hours for repair data contained in the historical claims database; and
determining whether the vehicle is repairable based on comparing the predicted total labor costs, total parts costs, and total hours for repair of the vehicle to a threshold and wherein if the vehicle is determined to not be repairable, processing a damage claim indicating the vehicle is a total loss.

13. The system in accordance with claim 12 , said processor further programmed to present to the policyholder an estimated cost for repair of the vehicle, including the predicted total labor costs, total parts costs, and total hours for repair of the vehicle.

14. The system in accordance with claim 13 , wherein the one or more images of the vehicle received from the policyholder correspond to the damage claim.

15. The system in accordance with claim 14 , said processor further programmed to process the damage claim based upon the estimated cost for repair of the vehicle.

16. The system in accordance with claim 12 , wherein the damage assessment model includes a machine learning program trained to identify damage to the vehicle.

17. The system in accordance with claim 16 , said processor programmed to train the machine learning program utilizing the plurality of damaged vehicle images and the corresponding metadata contained in the image training database.

18. The system in accordance with claim 16 , wherein the machine learning program uses a plurality of regression operations.

19. The system in accordance with claim 18 , wherein each regression operation is a linear regression operation.

20. The system in accordance with claim 12 , said processor programmed to receive policyholder data corresponding to the policyholder from a policyholder database.

21. The system in accordance with claim 20 , wherein the one or more images of the vehicle received from the policyholder correspond to the damage claim, said processor programmed to initiate the damage claim for the policyholder using the received policyholder data and the predicted total labor costs, total parts costs, and total hours for repair of the vehicle.

22. The system in accordance with claim 12 , wherein the damage assessment model is configured to infer internal damage to a set of parts of the vehicle that are not visible in a received one or more images of the vehicle.

Assignees

• State Farm Mutual Automobile Insurance Co

Inventors

• Lambert Holly
• Andrus Jennifer Malia
• Bokshi-Drotar Marigona
• Tomlinson Shane
• Green Daniel J
• Bernico Michael
• Sliz Bradley A
• Yang He

Classifications

• G06Q30/0283 Primary

  Price estimation or determination · CPC title

• G06V20/584

  of vehicle lights or traffic lights · CPC title

• H04N23/64

  Computer-aided capture of images, e.g. transfer from script file into camera, check of taken image quality, advice or proposal for image composition or decision on when to take image · CPC title

• H04N23/633

  for displaying additional information relating to control or operation of the camera · CPC title

• G06F18/214

  Generating training patterns; Bootstrap methods, e.g. bagging or boosting · CPC title