Granted Patent B2

US 12,026,679 · App. 16/585,468 · Granted Jul 2, 2024

Inventors: Abhijeet Gulati (San Diego, CA); Ravi Nemani (San Diego, CA); Joseph Hyland (San Diego, CA); Prarit Lamba (San Diego, CA)

Assignee: Mitchell International, Inc.

G06Q10/20G06N3/08G06Q40/08G06V20/10G07C5/0808

Abstract

Claims (92)

1. A method for automated estimation of repair data, the method comprising:

receiving, by a computer apparatus, a first vehicle damage image associated with an electronic claim for a damaged vehicle;

determining, by the computer apparatus, one or more damaged components on the damaged vehicle by:

providing the first vehicle damage image to a first generated artificial intelligence model, wherein the first generated artificial intelligence model has been trained using prior vehicle damage images;

determining, by the computer apparatus, a damage severity value associated with the one or more damaged components on the damaged vehicle by performing a heat map analysis on the first vehicle damage image, wherein the damage severity value corresponds with a shade illustrated in a heat map;

determining, by the computer apparatus, a first repair-or-replace designation of the one or more damaged components on the damaged vehicle by:

providing the first vehicle damage image and the damage severity value to a second generated artificial intelligence model,

wherein the first generated artificial intelligence model and the second generated artificial intelligence model correspond with different aspects of the automated estimation of repair data, and

wherein the first repair-or-replace designation indicates whether the one or more damaged components should be repaired or replaced;

providing, by the computer apparatus, the first repair-or-replace designation to a graphical user interface of a claims management device;

upon providing the first repair-or-replace designation to the claims management device, receiving, by the computer apparatus and from the claims management device, an adjustment request to change the first repair-or-replace designation, wherein the adjustment request comprises a second vehicle damage image that is different than the first vehicle damage image and the second vehicle damage image is associated with the electronic claim;

based on the adjustment request, determining, by the computer apparatus, a second repair-or-replace designation of the one or more damaged components on the vehicle by:

providing the second vehicle damage image and the damage severity value to the second generated artificial intelligence model; and

providing, by the computer apparatus, the second repair-or-replace designation to the graphical user interface of the claims management device.

2. The method as set forth in claim 1 further comprising: refining, by the computer apparatus, the second generated artificial intelligence model, wherein the refining comprises:

applying learning from the prior vehicle damage images to the second generated artificial intelligence model;

receiving training images at different angles for the first vehicle;

orienting the training images irrespective of underlying damage or no damage;

applying one or more transformation functions to the training images, wherein the second generated artificial intelligence model presents an understanding of how much orientation of the training images occurred irrespective of the underlying damage of no damage;

obtaining, by the computer apparatus, repair data associated with the training images for the damaged vehicle; and

training, by the computer apparatus, the second generated artificial intelligence model by correlating the training images to the repair data.

3. The method as set forth in claim 1 wherein the second generated artificial intelligence model is trained and refined by using a deep neural network architecture.

4. The method as set forth in claim 1 further comprising: determining, by the computer apparatus, when an adjustment to repair data is required based on an input received from the claims management device.

5. The method as set forth in claim 4 further comprising:

receiving, by the computer apparatus, one or more changes to the repair data when the adjustment is determined to be required;

revising, by the computer apparatus, the repair data based on the received one or more changes; and

providing, by the computer apparatus, the revised repair data to the claims management device.

6. The method as set forth in claim 1 wherein a heat map is generated as a graphical representation of a localized area of the one or more damaged components used to identify the damage severity value and a corresponding operation code.

7. The method as set forth in claim 1 , wherein the heat map of the heat map analysis is determined after training of the first generated artificial intelligence model and the second generated artificial intelligence model completes.

8. The method as set forth in claim 1 , wherein the adjustment request is associated with estimated labor hours and the estimated labor hours are provided to the graphical user interface of the claims management device.

9. The method as set forth in claim 1 , wherein the adjustment request is associated with an operation code identifying a panel of the vehicle and the operation code is provided to the graphical user interface of the claims management device.

10. A non-transitory computer readable medium having stored thereon instructions for automated estimating of repair data comprising executable code, which when executed by a processor, cause the processor to:

receive a first vehicle damage image associated with an electronic claim for a damaged vehicle;

determine one or more damaged components on the damaged vehicle by:

providing the first vehicle damage image to a first generated artificial intelligence model, wherein the first generated artificial intelligence model has been trained using prior vehicle damage images;

determine a damage severity value associated with the one or more damaged components on the damaged vehicle by performing a heat map analysis on the first vehicle damage image, wherein the damage severity value corresponds with a shade illustrated in a heat map;

determine a first repair-or-replace designation of the one or more damaged components on the damaged vehicle by:

providing the first vehicle damage image and the damage severity value to a second generated artificial intelligence model,

wherein the first generated artificial intelligence model and the second generated artificial intelligence model correspond with different aspects of the automated estimation of repair data,

wherein the first generated artificial intelligence model is used to train the second generated artificial intelligence model, and

wherein the first repair-or-replace designation indicates whether the one or more damaged components should be repaired or replaced;

provide the first repair-or-replace designation to a graphical user interface of a claims management device;

upon providing the first repair-or-replace designation to a graphical user interface of the claims management device, receiving, from the claims management device, an adjustment request to change the first repair-or-replace designation, wherein the adjustment request comprises a second vehicle damage image that is different than the first vehicle damage image and the second vehicle damage image is associated with the electronic claim;

based on the adjustment request, determining a second repair-or-replace designation of the one or more damaged components on the vehicle by:

providing the second vehicle damage image and the damage severity value to the second generated artificial intelligence model; and

providing the second repair-or-replace designation to the graphical user interface of the claims management device.

11. The non-transitory computer readable medium as set forth in claim 10 further causing the processor to:

refine the second generated artificial intelligence model, wherein the refining comprises:

applying learning from the prior vehicle damage images to the second generated artificial intelligence model;

receiving training images at different angles for the first vehicle;

orienting the training images irrespective of underlying damage or no damage;

applying one or more transformation functions to the training images, wherein the second generated artificial intelligence model presents an understanding of how much orientation of the training images occurred irrespective of the underlying damage of no damage;

obtaining repair data associated with the training images for the first damaged vehicle; and

training the second generated artificial intelligence model by correlating the training images to the repair data.

12. The non-transitory computer readable medium as set forth in claim 10 wherein the second generated artificial intelligence model is iteratively trained and refined by using a deep neural network architecture.

13. The non-transitory computer readable medium as set forth in claim 10 further comprising:

determining when an adjustment to the determined repair data is required based on an input received from the claims management device.

14. The non-transitory computer readable medium as set forth in claim 13 further comprising:

receiving one or more changes to the repair data when the adjustment is determined to be required;

revising the repair data based on the received one or more changes; and

providing the revised repair data to the claims management device.

15. The non-transitory computer readable medium as set forth in claim 10 wherein the heat map analysis is generated from a heat map as a graphical representation of a localized area of the one or more damaged components used to identify the damage severity value and a corresponding operation code.

16. A repair management computer apparatus for automated estimation of repair data comprising:

a processor; and

a memory coupled to the processor which is configured to be capable of executing programmed instructions comprising and stored in the memory to:

receive a first vehicle damage image associated with an electronic claim for a damaged vehicle;

determine one or more damaged components on the damaged vehicle by:

providing the first vehicle damage image to a first generated artificial intelligence model, wherein the first generated artificial intelligence model has been trained using prior vehicle damage images;

determine a damage severity value associated with the one or more damaged components on the damaged vehicle by performing a heat map analysis on the first vehicle damage image, wherein the damage severity value corresponds with a shade illustrated in a heat map;

determine a first repair-or-replace designation of the one or more damaged components on the damaged vehicle by:

providing the first vehicle damage image and the damage severity value to a second generated artificial intelligence model,

wherein the first generated artificial intelligence model and the second generated artificial intelligence model correspond with different aspects of the automated estimation of repair data, and

wherein the first repair-or-replace designation indicates whether the one or more damaged components should be repaired or replaced;

provide the first repair-or-replace designation to a graphical user interface of a claims management device;

upon providing the first repair-or-replace designation to a graphical user interface of the claims management device, receive, from the claims management device, an adjustment request to change the first repair-or-replace designation, wherein the adjustment request comprises a second vehicle damage image that is different than the first vehicle damage image and the second vehicle damage image is associated with the electronic claim;

based on the adjustment request, determine a second repair-or-replace designation of the one or more damaged components on the vehicle by:

provide the second vehicle damage image and the damage severity value to the second generated artificial intelligence model; and

provide the second repair-or-replace designation to the graphical user interface of the claims management device.

17. The repair management computer apparatus as set forth in claim 16 wherein the processor is further configured to be capable of executing the stored programmed instructions to refine the second generated artificial intelligence model, wherein the refining comprises:

applying learning from the prior vehicle damage images to the second generated artificial intelligence model;

receiving training images at different angles for the first vehicle;

orienting the training images irrespective of underlying damage or no damage;

applying one or more transformation functions to the training images, wherein the second generated artificial intelligence model presents an understanding of how much orientation of the training images occurred irrespective of the underlying damage of no damage;

obtaining repair data associated with the training images for the damaged vehicle; and

training the second generated artificial intelligence model by correlating the training images to the repair data.

18. The repair management computer apparatus as set forth in claim 16 wherein the second generated artificial intelligence model is iteratively trained and refined by using a deep neural network architecture.

19. The repair management computer apparatus as set forth in claim 16 wherein the processor is further configured to be capable of executing the stored programmed instructions to determine when an adjustment to repair data is required based on an input received from the claims management device.

20. The repair management computer apparatus as set forth in claim 19 wherein the processor is further configured to be capable of executing the stored programmed instructions to:

receive one or more changes to the repair data when the adjustment is determined to be required;

revise the repair data based on the received one or more changes; and

provide the revised repair data to the claims management device.

21. The repair management computer apparatus as set forth in claim 16 wherein the heat map analysis is generated from a heat map as a graphical representation of a localized area of the one or more damaged components used to identify the damage severity value and a corresponding operation code.

Assignments (8)

Continuity (2)