Patent Issued for Shared mobility service passenger matching based on passenger attributes (USPTO 11590981): Allstate Insurance Company

2023 MAR 20 (NewsRx) -- By a News Reporter-Staff News Editor at Insurance Daily News -- A patent by the inventors Allen, Rachel Edelshteyn (Chicago, IL, US), Coleman, II, Neal Edgar (Evanston, IL, US), Irey, Grady (Des Plaines, IL, US), filed on July 31, 2018, was published online on February 28, 2023, according to news reporting originating from Alexandria, Virginia, by NewsRx correspondents.

Patent number 11590981 is assigned to Allstate Insurance Company (Northbrook, Illinois, United States).

The following quote was obtained by the news editors from the background information supplied by the inventors: “Today, shared mobility programs differentiate on service (e.g., some highlight safety, some highlight speed of pickup, etc.). These platforms are introducing pooling services to pick up more than one individual going in similar directions. This is the vision for increased efficiency on the road, and shared mobility platforms that plan to transition into autonomous platforms may use pooling as a mechanism to develop future autonomous routing.”

In addition to the background information obtained for this patent, NewsRx journalists also obtained the inventors’ summary information for this patent: “When riders, presumably strangers, get into a pool, the experience becomes uncomfortable and the shared mobility platform loses control of the experience. Riders also lose trust in the platform. Techniques disclosed herein address this problem by profiling riders and providing insights to shared mobility platforms that help them match riders in a pooling situation based on likelihood of those riders to enjoy each other’s company. Riders may opt in to personal data sharing in order to allow the shared mobility platform to match them with other riders having similar and/or compatible attributes. For example, two riders that are moms may be paired together or two riders that like soccer may be paired together. They may even be paired together based on a desire to listen to the same song or have the same temperature in the vehicle. This enhances the in-vehicle experience and helps build trust for the platform.

“Techniques described herein provide a computing platform that receives, by at least one processor, via a communication interface, information about a first user of the shared mobility service, generates, based on the information about the first user, a first user profile comprising one or more first user attributes, compares the one or more first user attributes to one or more second user attributes of a second user profile associated with a second user of the shared mobility service, and based on the comparison, causes a vehicle carrying the second user to pick up the first user. In some cases, the comparing comprises generating, using a trained model, a propensity score indicating a compatibility of the first user and the second user. In some cases, the trained model, when executed by the computing platform, outputs the propensity score based on inputs associated with the one or more first attributes of the first user and the one or more second attributes of the second user. In some cases, the propensity score further indicates a compatibility of the first user and a driver of the vehicle.

“In some cases, prior to causing the vehicle carrying the second user to pick up the first user, the computing platform determines a plurality of vehicles available to pick up the first user, the plurality of vehicles including the vehicle carrying the second user, and ranks, based on the propensity score, the vehicle carrying the second user in relation to the plurality of vehicles.

“In some cases, causing the vehicle carrying the second user to pick up the first user comprises transmitting an instruction to an autonomous vehicle control system controlling the vehicle. In some cases, causing the vehicle carrying the second user to pick up the first user comprises transmitting an instruction to an application running on a device associated with a driver of the vehicle.

“In some cases, the information about the user comprises audio or video information.”

The claims supplied by the inventors are:

“1. A method comprising: at a shared mobility service management computing platform comprising at least one processor, memory, and a communication interface: training an initial score machine-learning model using training data that correlates behavior data with indicators of particular personality traits to output a continuous value indicating an estimated strength of a personality trait for an occupant of a vehicle, wherein the training data includes training driving data; receiving, by the at least one processor, via the communication interface, information about a user of a shared mobility service, wherein the user is a driver; generating, based on the information about the user, a user profile comprising user driving data and one or more user attributes based on appended data from third-party sources including at least one of surveys, historic purchase habits, observed travel patterns, or online browsing activity, wherein the one or more user attributes include at least one of a personality factor, an interest, a demographic factor, a preferred temperature, a preferred vehicle condition, or preference of genre of media, and wherein the one or more user attributes are numerical indicating a degree to which the user has a corresponding attribute and are determined based on rules that indicate a threshold number of instances determined from the appended data; extracting features from the user attributes and using the extracted features as inputs to the initial score machine-learning model, wherein the initial score machine-learning model outputs a continuous value indicating an estimated strength of a personality trait for the user; determining a plurality of rides associated with the shared mobility service that are available to a passenger including a ride with the user; running a trained machine-learning model to determine a propensity score between the user and each of the plurality of rides, wherein a lookup is performed via accessing an application programming interface (API) and the trained machine-learning model matches attributes using a multi-dimensional propensity matrix indicating propensities between groups of attributes to determine propensity sub-scores for each pair of attributes corresponding to the user and each attribute of each ride, and wherein each propensity score is a combination of respective propensity sub-scores between the user and each ride; ranking the plurality of rides based on the respective plurality of propensity scores; sending a ride opportunity to pick up the user based on a highest rated ride; and training the initial score machine-learning model with the extracted features.

“2. The method of claim 1, wherein the trained machine-learning model, when executed by the shared mobility service management computing platform, outputs the propensity score for the vehicle associated with the driver based on inputs associated with the one or more user attributes and one or more attributes associated with the passenger.

“3. The method of claim 1, further comprising: responsive to receiving a selection of a vehicle of the user, transmitting an instruction to an autonomous vehicle control system controlling the vehicle.

“4. The method of claim 1, further comprising: responsive to receiving a selection of the user, transmitting an instruction to an application running on a device associated with the user.

“5. The method of claim 1, wherein the shared user preference is generated based on the user profile and a profile associated with the passenger.

“6. The method of claim 1, further comprising: sending instruction to comply with a shared user preference for the user and the passenger that comprises one or more of an instruction to set a temperature and an instruction to set a type of music.

“7. The method of claim 1, further comprising: capturing audio or video information associated with the user; and updating the user profile based on the audio or video information.

“8. A computing platform comprising at least one processor, memory, and a communication interface, wherein the memory stores instructions that, when executed by the at least one processor, cause the computing platform to: train an initial score machine-learning model using training data that correlates behavior data with indicators of particular personality traits to output a continuous value indicating an estimated strength of a personality trait for an occupant of a vehicle , wherein the training data includes training driving data; receive, by the at least one processor, via the communication interface, information about a user of a shared mobility service, wherein the user is a driver; generate, based on the information about the user, a user profile comprising user driving data and one or more user attributes based on appended data from third-party sources including at least one of surveys, historic purchase habits, observed travel patterns, or online browsing activity, wherein the one or more user attributes include at least one of a personality factor, an interest, a demographic factor, a preferred temperature, a preferred vehicle condition, or preference of genre of media, and wherein the one or more user attributes are numerical indicating a degree to which the user has a corresponding attribute and are determined based on rules that indicate a threshold number of instances determined from the appended data; extract features from the user attributes and using the extracted features as inputs to the initial score machine-learning model, wherein the initial score machine-learning model outputs a continuous value indicating an estimated strength of a personality trait for the user: determine a plurality of rides associated with the shared mobility service that are available to a passenger including a ride with the user; run a trained machine-learning model to determine a propensity score between the user and each of the plurality of rides, wherein a lookup is performed via accessing an application programming interface (API) and the trained machine-learning model matches attributes using a multi-dimensional propensity matrix indicating propensities between groups of attributes to determine propensity sub-scores for each pair of attributes corresponding to the user and each attribute of each ride, and wherein each propensity score is a combination of respective propensity sub-scores between the user and each ride; rank the plurality of rides based on the respective plurality of propensity scores; send a ride opportunity to pick up the user based on a highest rated ride; and train the initial score machine-learning model with the extracted features.

“9. The computing platform of claim 8, wherein to compare the one or more user attributes to the one or more attributes of the passenger, the instructions cause the computing platform to generate, using a trained model, a propensity score indicating a compatibility of the user and the passenger.

“10. The computing platform of claim 9, wherein the trained model, when executed by the computing platform, outputs the propensity score based on inputs associated with the one or more user attributes and the one or more attributes of the passenger.

“11. The computing platform of claim 9, wherein the propensity score further indicates a compatibility of the passenger and the user.

“12. The computing platform of claim 8, wherein, responsive to receiving a selection of a vehicle associated with the user to pick up the passenger, the instructions cause the computing platform to transmit an instruction to an autonomous vehicle control system controlling the vehicle.

“13. The computing platform of claim 8, wherein, the instructions cause the computing platform to transmit an instruction to an application running on a device associated with the user.”

There are additional claims. Please visit full patent to read further.

URL and more information on this patent, see: Allen, Rachel Edelshteyn. Shared mobility service passenger matching based on passenger attributes. U.S. Patent Number 11590981, filed July 31, 2018, and published online on February 28, 2023. Patent URL (for desktop use only): https://ppubs.uspto.gov/pubwebapp/external.html?q=(11590981)&db=USPAT&type=ids

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