Patent Issued for Driver Identification For Trips Associated With Anonymous Vehicle Telematics Data (USPTO 10,699,498)

2020 JUL 10 (NewsRx) -- By a News Reporter-Staff News Editor at Insurance Daily News -- A patent by the inventors Goldfarb, Meghan Sims (Bloomington, IL); Christensen, Scott Thomas (Salem, OR); Thiel, Taylor Michael (Urbana, IL); Dahiya, Anuj (Naperville, IL), filed on September 28, 2018, was published online on July 13, 2020, according to news reporting originating from Alexandria, Virginia, by NewsRx correspondents.

Patent number 10,699,498 is assigned to State Farm Mutual Automobile Insurance Company (Bloomington, Illinois, United States).

The following quote was obtained by the news editors from the background information supplied by the inventors: “Current technologies make use of vehicle telematics data to assess driving behavior. For example, the telematics data may be collected and analyzed to determine the acceleration, braking and/or cornering habits of a driver of a vehicle, and the results of the analysis may be used to measure the performance of the driver over time. The telematics data may be generated by sensors on the vehicle, or by a mobile device (e.g., smart phone) carried by the driver, for example. The measured performance may then be used for various purposes, such as modifying an insurance rating of the driver.

“Typically, the insurance policy with which a particular set of vehicle telematics data is associated is known. If the vehicle telematics data is generated by a data collection device installed in a vehicle, for example, the data may include a unique identifier of the vehicle and/or data collection device, which can be matched to a particular auto insurance policy. As another example, if the vehicle telematics data is generated by a personal data collection device (e.g., a smart phone), the data may include a unique identifier of the driver and/or personal data collection device, which can also be matched to a particular auto insurance policy. In some scenarios, however, there can be ambiguity concerning who was driving a vehicle when a particular set of vehicle telematics data was generated. Even if a mobile device is known to be owned by a particular person, for example, vehicle telematics data generated by that mobile device may have been generated while the person was sitting in a passenger seat. As another example, a data collection device installed on a vehicle may generate vehicle telematics data without providing any indication of who was driving while that data was generated. Thus, an insurance provider may not know whether a particular portion of the vehicle telematics data (e.g., a portion corresponding to a particular driving trip) reflects the driving performance of a primary insured, the driving performance of another individual disclosed on the insurance policy (e.g., the spouse of the primary insured), or even the driving performance of a driver not disclosed on the insurance policy (e.g., a friend of the primary insured, or a friend of another disclosed driver). If driving performance is not correctly attributed, the insurance provider may be unable to accurately assess the risk associated with particular drivers. For these and other reasons, techniques for accurately attributing sets of vehicle telematics data to individuals would be beneficial. Further, because past driving performance or behaviors for individuals listed on an insurance policy may not be known (e.g., due to an absence of any vehicle telematics data that is known to correspond to each individual), attribution techniques that do not require such historical information may be particularly beneficial.”

In addition to the background information obtained for this patent, NewsRx journalists also obtained the inventors’ summary information for this patent: “The present embodiments may, inter alia, utilize insurance policy information to more easily, efficiently and/or accurately determine which sets of vehicle telematics data should be attributed to various individuals, even in the absence of historical vehicle telematics data and/or driving performance information for those individuals. By increasing the accuracy of the data attribution in these situations, an insurance provider or other entity may more accurately assess driving risk and/or performance for one or more individuals.

“In one aspect, method for attributing vehicle telematics data to individuals includes receiving, at a server, vehicle telematics data collected by a data collection device during a plurality of trips. The vehicle telematics data includes a plurality of subsets of vehicle telematics data each corresponding to a different one of the plurality of trips. The method also includes, for each subset of the plurality of subsets of vehicle telematics data, using the subset of vehicle telematics data to generate a respective one of a plurality of metric sets. Each of the plurality of metric sets includes a plurality of metrics each indicative of a different driving behavior or a different feature of a driving environment. The method also includes retrieving, by the server and from a policy database, policy information pertaining to an insurance policy associated with the data collection device, and determining, by the server and based upon the retrieved policy information, a number N of disclosed drivers associated with the insurance policy. N is greater than or equal to one. The method also includes performing, by the server, a statistical analysis on the plurality of metric sets at least by executing a clustering algorithm on the plurality of metric sets. Executing the clustering algorithm includes generating indications of which of the plurality of metric sets belong to which of a plurality of clusters. The method also includes, for each metric set of at least some of the plurality of metric sets, and based upon the generated indications, assigning, by the server, one or both of (i) the metric set, and (ii) the corresponding one of the plurality of subsets of vehicle telematics data, to one of the N disclosed drivers. The method also includes causing, by the server, an insurance rating associated with the insurance policy to be adjusted based at least in part on (i) the vehicle telematics data and (ii) the manner in which the corresponding ones of the plurality of metric sets or the plurality of subsets of vehicle telematics data are assigned to the N disclosed drivers.

“In another aspect, a server includes one or more processors and a memory. The memory stores instructions that, when executed by the server, cause the server to receive vehicle telematics data collected by a data collection device during a plurality of trips. The vehicle telematics data includes a plurality of subsets of vehicle telematics data each corresponding to a different one of the plurality of trips. The instructions also cause the server to, for each subset of the plurality of subsets of vehicle telematics data, use the subset of vehicle telematics data to generate a respective one of a plurality of metric sets. Each of the plurality of metric sets includes a plurality of metrics each indicative of a different driving behavior or a different feature of a driving environment. The instructions also cause the server to retrieve, from a policy database, policy information pertaining to an insurance policy associated with the data collection device, and determine, based upon the retrieved policy information, a number N of disclosed drivers associated with the insurance policy. N is greater than or equal to one. The instructions also cause the server to perform a statistical analysis on the plurality of metric sets at least by executing a clustering algorithm on the plurality of metric sets. Executing the clustering algorithm includes generating indications of which of the plurality of metric sets belong to which of a plurality of clusters. The instructions also cause the server to, for each metric set of at least one of the plurality of metric sets, and based upon the generated indications, assign one or both of (i) the metric set, and (ii) the corresponding one of the plurality of subsets of vehicle telematics data, to one of the N disclosed drivers.

“In another aspect, a non-transitory computer-readable medium stores instructions that, when executed by one or more processors, cause the one or more processors to receive vehicle telematics data collected by a data collection device during a plurality of trips. The vehicle telematics data includes a plurality of subsets of vehicle telematics data each corresponding to a different one of the plurality of trips. The instructions also cause the one or more processors to, for each subset of the plurality of subsets of vehicle telematics data, use the subset of vehicle telematics data to generate a respective one of a plurality of metric sets. Each of the plurality of metric sets includes a plurality of metrics each indicative of a different driving behavior or a different feature of a driving environment. The instructions also cause the one or more processors to retrieve, from a policy database, policy information pertaining to an insurance policy associated with the data collection device, and determine, based upon the retrieved policy information, a number N of disclosed drivers associated with the insurance policy. N is greater than or equal to one. The instructions also cause the one or more processors to perform a statistical analysis on the plurality of metric sets at least by executing a clustering algorithm on the plurality of metric sets. Executing the clustering algorithm includes generating indications of which of the plurality of metric sets belong to which of a plurality of clusters. The instructions also cause the one or more processors to, for each metric set of at least one of the plurality of metric sets, and based upon the generated indications, assign one or both of (i) the metric set, and (ii) the corresponding one of the plurality of subsets of vehicle telematics data, to one of the N disclosed drivers. The instructions also cause the one or more processors to cause an insurance rating associated with the insurance policy to be adjusted based at least in part on (i) the vehicle telematics data and (ii) the manner in which the corresponding ones of the plurality of metric sets or the plurality of subsets of vehicle telematics data are assigned to the N disclosed drivers.”

The claims supplied by the inventors are:

“What is claimed:

“1. A method for attributing vehicle telematics data to individuals, the method comprising: receiving, at a server, vehicle telematics data collected by a data collection device during a plurality of trips, wherein the vehicle telematics data includes a plurality of subsets of vehicle telematics data each corresponding to a different one of the plurality of trips; for each subset of the plurality of subsets of vehicle telematics data, using the subset of vehicle telematics data to generate a respective one of a plurality of metric sets, wherein each of the plurality of metric sets includes a plurality of metrics each indicative of a different driving behavior or a different feature of a driving environment; retrieving, by the server and from a policy database, policy information pertaining to an insurance policy associated with the data collection device; determining, by the server and based upon the retrieved policy information, a number N of disclosed drivers associated with the insurance policy, N being greater than or equal to one; performing, by the server, a statistical analysis on the plurality of metric sets at least by executing a clustering algorithm on the plurality of metric sets, wherein executing the clustering algorithm includes generating indications of which of the plurality of metric sets belong to which of a plurality of clusters; for each metric set of at least some of the plurality of metric sets, and based upon the generated indications, assigning, by the server, one or both of (i) the metric set, and (ii) the corresponding one of the plurality of subsets of vehicle telematics data, to one of the N disclosed drivers; and causing, by the server, an insurance rating associated with the insurance policy to be adjusted based at least in part on (i) the vehicle telematics data and (ii) the manner in which the corresponding ones of the plurality of metric sets or the plurality of subsets of vehicle telematics data are assigned to the N disclosed drivers.

“2. The method of claim 1, wherein using the subset of vehicle telematics data to generate the respective one of the plurality of metric sets includes using at least two of (i) global positioning satellite (GPS) data, (ii) accelerometer data, and (iii) gyroscope data, to calculate a respective first metric.

“3. The method of claim 1, wherein determining the number N of disclosed drivers includes counting all drivers specified on the insurance policy.

“4. The method of claim 1, wherein executing the clustering algorithm on the plurality of metric sets includes: prior to executing the clustering algorithm, setting a parameter of the clustering algorithm equal to N.

“5. The method of claim 4, wherein the clustering algorithm is a k-means clustering algorithm, and wherein setting the parameter of the clustering algorithm equal to N includes setting k equal to N.

“6. The method of claim 1, wherein the clustering algorithm is an expectation-maximization (EM) clustering algorithm.

“7. The method of claim 1, wherein the clustering algorithm is a Ward clustering algorithm.

“8. The method of claim 1, wherein the clustering algorithm is a flexible clustering algorithm.

“9. The method of claim 1, wherein: executing the clustering algorithm on the plurality of metric sets includes identifying m clusters, m being an integer greater than N; and the method further comprises determining, by the server and in response to identifying the m clusters, that there are m-N undisclosed drivers.

“10. The method of claim 1, wherein: executing the clustering algorithm on the plurality of metric sets includes determining that a first metric set of the plurality of metric sets is an outlier not associated with any cluster; and the method further comprises, not assigning to any driver both (i) the first metric set and (ii) the one of the plurality of subsets of vehicle telematics data that corresponds to the first metric set.

“11. The method of claim 10, wherein determining that the first metric set is an outlier includes determining that the first metric set is at least a threshold distance from any cluster.

“12. The method of claim 1, wherein receiving vehicle telematics data collected by a data collection device during a plurality of trips includes receiving vehicle telematics data collected by a mobile telephone device during the plurality of trips.

“13. A server comprising: one or more processors; and a memory storing instructions that, when executed by the server, cause the server to receive vehicle telematics data collected by a data collection device during a plurality of trips, wherein the vehicle telematics data includes a plurality of subsets of vehicle telematics data each corresponding to a different one of the plurality of trips, for each subset of the plurality of subsets of vehicle telematics data, use the subset of vehicle telematics data to generate a respective one of a plurality of metric sets, wherein each of the plurality of metric sets includes a plurality of metrics each indicative of a different driving behavior or a different feature of a driving environment, retrieve, from a policy database, policy information pertaining to an insurance policy associated with the data collection device, determine, based upon the retrieved policy information, a number N of disclosed drivers associated with the insurance policy, N being greater than or equal to one, perform a statistical analysis on the plurality of metric sets at least by executing a clustering algorithm on the plurality of metric sets, wherein executing the clustering algorithm includes generating indications of which of the plurality of metric sets belong to which of a plurality of clusters, for each metric set of at least one of the plurality of metric sets, and based upon the generated indications, assign one or both of (i) the metric set, and (ii) the corresponding one of the plurality of subsets of vehicle telematics data, to one of the N disclosed drivers, and cause an insurance rating associated with the insurance policy to be adjusted based at least in part on (i) the vehicle telematics data and (ii) the manner in which the corresponding ones of the plurality of metric sets or the plurality of subsets of vehicle telematics data are assigned to the N disclosed drivers.

“14. The server of claim 13, wherein the plurality of metrics include a first metric calculated using at least two of (i) global positioning satellite (GPS) data, (ii) accelerometer data, and (iii) gyroscope data.

“15. The server of claim 13, wherein the clustering algorithm is either a k-means clustering algorithm or an expectation-maximization (EM) clustering algorithm, and wherein the instructions cause the server to execute the clustering algorithm on the plurality of metric sets at least in part by, prior to executing the clustering algorithm, setting a parameter of the clustering algorithm equal to N.

“16. The server of claim 13, wherein the instructions cause the server to: execute the clustering algorithm at least in part by determining a number m of clusters; and when m is greater than N, determine that there are m-N undisclosed drivers.

“17. A non-transitory computer-readable medium storing instructions that, when executed by one or more processors, cause the one or more processors to: receive vehicle telematics data collected by a data collection device during a plurality of trips, wherein the vehicle telematics data includes a plurality of subsets of vehicle telematics data each corresponding to a different one of the plurality of trips; for each subset of the plurality of subsets of vehicle telematics data, use the subset of vehicle telematics data to generate a respective one of a plurality of metric sets, wherein each of the plurality of metric sets includes a plurality of metrics each indicative of a different driving behavior or a different feature of a driving environment; retrieve, from a policy database, policy information pertaining to an insurance policy associated with the data collection device; determine, based upon the retrieved policy information, a number N of disclosed drivers associated with the insurance policy, N being greater than or equal to one; perform a statistical analysis on the plurality of metric sets at least by executing a clustering algorithm on the plurality of metric sets, wherein executing the clustering algorithm includes generating indications of which of the plurality of metric sets belong to which of a plurality of clusters; for each metric set of at least one of the plurality of metric sets, and based upon the generated indications, assign one or both of (i) the metric set, and (ii) the corresponding one of the plurality of subsets of vehicle telematics data, to one of the N disclosed drivers; and cause an insurance rating associated with the insurance policy to be adjusted based at least in part on (i) the vehicle telematics data and (ii) the manner in which the corresponding ones of the plurality of metric sets or the plurality of subsets of vehicle telematics data are assigned to the N disclosed drivers.

“18. The non-transitory computer-readable medium of claim 17, wherein the plurality of metrics include a first metric calculated using at least two of (i) global positioning satellite (GPS) data, (ii) accelerometer data, and (iii) gyroscope data.

“19. The non-transitory computer-readable medium of claim 17, wherein the clustering algorithm is either a k-means clustering algorithm or an expectation-maximization (EM) clustering algorithm, and wherein the instructions cause the one or more processors to execute the clustering algorithm on the plurality of metric sets at least in part by, prior to executing the clustering algorithm, setting a parameter of the clustering algorithm equal to N.

“20. The non-transitory computer-readable medium of claim 17, wherein the instructions cause the one or more processors to: execute the clustering algorithm at least in part by determining a number m of clusters; and when m is greater than N, determine that there are m-N undisclosed drivers.”

URL and more information on this patent, see: Goldfarb, Meghan Sims; Christensen, Scott Thomas; Thiel, Taylor Michael; Dahiya, Anuj. Driver Identification For Trips Associated With Anonymous Vehicle Telematics Data. U.S. Patent Number 10,699,498, filed September 28, 2018, and published online on July 13, 2020. Patent URL: http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PALL&p=1&u=%2Fnetahtml%2FPTO%2Fsrchnum.htm&r=1&f=G&l=50&s1=10,699,498.PN.&OS=PN/10,699,498RS=PN/10,699,498

(Our reports deliver fact-based news of research and discoveries from around the world.)