Method and system for comparing automatically determined crash information to historical collision data to detect fraud

Patent metadata

Field	Value
Publication number	US-9904928-B1
Application number	US-201514627104-A
Country	US
Kind code	B1
Filing date	Feb 20, 2015
Priority date	Jul 11, 2014
Publication date	Feb 27, 2018
Grant date	Feb 27, 2018

Abstract

To detect a risk of insurance fraud in a vehicle insurance claim for physical injuries, injury data may be retrieved from previous crashes involving similar type vehicles as the damaged vehicle, and having similar crash characteristics. A likelihood that the claimant suffered physical injuries from the crash may be determined based on the percentage of people who suffered similar injuries according to the injury data. If the likelihood is very small, the claim for physical injuries may be flagged as a risk of fraud.

First claim

What is claimed:

1. A method for detecting potential fraud in vehicle insurance claims, the method executed by one or more processors programmed to perform the method, the method comprising:
receiving, at one or more processors, identification information for a damaged vehicle and sensor data captured during a vehicle crash involving the damaged vehicle including at least one of: vehicle part position data from spatial sensors disposed within the damaged vehicle, vehicle acceleration data from an accelerometer disposed within the damaged vehicle, vehicle velocity data from a positioning device disposed within the damaged vehicle, or vehicle direction data from the positioning device;
receiving, at the one or more processors, an indication of an injury for a person involved in the vehicle crash with the damaged vehicle, including a location of the person with respect to the damaged vehicle when the injury occurred and at least one of: an indication of a type of the injury or an indication of a severity of the injury;
analyzing, by the one or more processors, the identification information and the sensor data to identify a matching set of historical collision data for a same type of vehicle as the damaged vehicle, wherein the matching set of historical collision data includes historical injury data;
determining, by the one or more processors, one or more likelihoods that at least one of: the received type of the injury occurred or the received severity of the injury occurred from the location of the person with respect to the damaged vehicle based on a subset of the historical injury data corresponding to the matching set of historical collision data for injuries occurring at the location of the person with respect to the damaged vehicle;
comparing, by the one or more processors, each of the one or more determined likelihoods to one of a plurality of predetermined threshold likelihoods;
when each determined likelihood is less than the respective predetermined threshold likelihood, generating, by the one or more processors, a suspicious loss indicator for the injury based on the likelihood that the received injury occurred; and
transmitting, by the one or more processors via a communication network, the suspicious loss indicator to a mobile computing device for display to flag the injury as a risk of fraud.

2. The method of claim 1 , wherein historical injury data includes at least one of:
a number of people injured in past collisions involving the same type of vehicle having matching crash information with the damaged vehicle,
a number of people having a particular type of injury in past collisions involving the same type of vehicle having matching crash information with the damaged vehicle,
a number of people having a particular injury severity from a set of injury severities in past collisions involving the same type of vehicle having matching crash information with the damaged vehicle, wherein the set of injury severities includes minor, moderate, and severe, or
a number of people injured in a particular location from a set of locations in past collisions involving the same type of vehicle having matching crash information with the damaged vehicle, wherein the set of locations includes a driver seat in the same type of vehicle, a passenger seat in the same type vehicle, a driver seat in another vehicle which collided with the same type of vehicle, and a passenger seat in the other vehicle which collided with the same type of vehicle.

3. The method of claim 2 , wherein determining a likelihood that the received type or severity of the injury occurred based on the historical injury data comprises determining, by the one or more processors, a percentage of people who were injured in past collisions involving the same type of vehicle having matching crash information with the damaged vehicle based on the historical injury data.

4. The method of claim 1 , wherein receiving an indication of an injury further comprises receiving, at the one or more processors,
an indication of medical expenses associated with the injury.

5. The method of claim 1 , wherein the location of the person involved in the vehicle crash with the damaged vehicle includes at least one of:
a driver seat in the damaged vehicle,
a passenger seat in the damaged vehicle,
a driver seat in another vehicle which collided with the damaged vehicle, or
a passenger seat in the other vehicle which collided with the damaged vehicle.

6. The method of claim 1 , wherein generating the suspicious loss indicator further comprises determining, by the one or more processors, a suspicious loss category from a set of suspicious loss categories based on the likelihood that the received injury occurred including high risk, medium risk, and low risk.

7. A system for detecting potential fraud in vehicle insurance claims, the system comprising:
a communication network; and
one or more computing devices communicatively coupled to the communication network, each of the one or more computing devices having a memory and one or more processors and at least one of the computing device configured to:
receive identification information for a damaged vehicle and sensor data captured during a vehicle crash involving the damaged vehicle including at least one of: vehicle part position data from spatial sensors disposed within the damaged vehicle, vehicle acceleration data from an accelerometer disposed within the damaged vehicle, vehicle velocity data from a positioning device disposed within the damaged vehicle, or vehicle direction data from the positioning device via the communication network,
receive an indication of an injury for a person involved in the vehicle crash with the damaged vehicle via the communication network, including a location of the person with respect to the damaged vehicle when the injury occurred and at least one of: an indication of a type of the injury or an indication of a severity of the injury,
analyze the identification information and the sensor data to identify a matching set of historical collision data for a same type of vehicle as the damaged vehicle, wherein the matching set of historical collision data includes historical injury data,
determine one or more likelihoods that at least one of: the received type of injury occurred or the received severity of the injury occurred from the location of the person with respect to the damaged vehicle based on a subset of the historical injury data corresponding to the matching set of historical collision data for injuries occurring at the location of the person with respect to the damaged vehicle,
compare each of the one or more determined likelihoods to one of a plurality of predetermined threshold likelihoods,
when each determined likelihood is less than the respective predetermined threshold likelihood, generate a suspicious loss indicator for the injury based on the likelihood that the received injury occurred; and
transmit, via the communication network, the suspicious loss indicator to a mobile computing device for display to flag the injury as a risk of fraud.

8. The system of claim 7 , wherein historical injury data includes at least one of:
a number of people injured in past collisions involving the same type of vehicle having matching crash information with the damaged vehicle,
a number of people having a particular type of injury in past collisions involving the same type of vehicle having matching crash information with the damaged vehicle,
a number of people having a particular injury severity from a set of injury severities in past collisions involving the same type of vehicle having matching crash information with the damaged vehicle, wherein the set of injury severities includes minor, moderate, and severe, or

Assignees

• State Farm Mutual Automobile Insurance Co

Inventors

• Leise William J

Classifications

• G06Q10/20

  Administration of product repair or maintenance · CPC title

• G06Q10/08355

  Routing methods · CPC title

• G06Q40/04

  Trading; Exchange, e.g. stocks, commodities, derivatives or currency exchange · CPC title

• G06Q30/0206

  Price or cost determination based on market factors · CPC title

• G06Q30/0185 Primary

  Product, service or business identity fraud · CPC title