Trip anomaly detection system

Patent metadata

Field	Value
Publication number	US-9723469-B2
Application number	US-201615366387-A
Country	US
Kind code	B2
Filing date	Dec 1, 2016
Priority date	Jun 17, 2015
Publication date	Aug 1, 2017
Grant date	Aug 1, 2017

Abstract

An anomaly detection system is provided in connection with a transport service. The anomaly detection system can construct routine route profiles for individual users of the transport service using historical route data. The anomaly detection system can monitor a current route traveled by a user. The anomaly detection system can further identify a matching routine route profile of the respective user. The anomaly detection system can utilize the matching routine route profile to identify a probable anomaly in the current route. In response to detecting the probable anomaly, the anomaly detection system can enable a safety protocol to perform a number of actions.

First claim

What is claimed is:

1. An anomaly detection system for a service comprising:
one or more processors; and
one or more memory resources storing instructions that, when executed by the one or more processors, cause the anomaly detection system to:
construct routine route profiles for users of a network service by, for each of the users, (i) collecting historical route data, and (ii) identifying correlated route shapes in the historical route data, wherein the correlated route shapes comprise a plurality of distinct routes between a common start point or region and a common end point or region;
monitor a current route traveled by a respective one of the users;
identify a matching routine route profile of the respective user;
determine a probable anomaly in the current route based on the matching routine route profile; and
in response to determining the probable anomaly, perform one or more actions in accordance with a safety protocol, wherein an initial action of the safety protocol comprises transmitting a status query to a mobile device of the respective user, and wherein a subsequent action in the safety protocol comprises contacting a predetermined user contact to report the probable anomaly.

2. The anomaly detection system of claim 1 , wherein the executed instructions further cause the anomaly detection system to, prior to monitoring the current route:
receive a request for service from the respective user; and
select a driver to provide the requested service.

3. The anomaly detection system of claim 2 , wherein the request includes a destination, and wherein the executed instructions cause the anomaly detection system to identify the matching routine route profile prior to monitoring the current route, and based on an initial location of the respective user and the destination.

4. The anomaly detection system of claim 2 , wherein the executed instructions cause the anomaly detection system to select the driver based on (i) a proximity between the respective user and the selected driver, and (ii) a comparison between a user profile of the respective user and a driver profile of the selected driver, the driver profile comprising reputation data.

5. The anomaly detection system of claim 4 , wherein the executed instructions cause the anomaly detection system to perform the comparison by (i) identifying one or more factors for the respective user comprising one or more of an age, a gender, or user preferences in the user profile, and (ii) determining a reputation score for the selected driver based on the reputation data.

6. The anomaly detection system of claim 1 , wherein the executed instructions cause the anomaly detection system to identify the matching routine route profile during the current route traveled by the user.

7. The anomaly detection system of claim 1 , wherein the executed instructions cause the anomaly detection system to determine the probable anomaly by performing a real-time probabilistic calculation comprising a number of probability factors collectively exceeding an anomaly threshold.

8. The anomaly detection system of claim 7 , wherein the probability factors comprise the current route entering a flagged region.

9. The anomaly detection system of claim 8 , wherein the executed instructions cause the anomaly detection system to determine the flagged region based on one or more of third-party crime data, or the flagged region being identified in a user profile of the respective user.

10. The anomaly detection system of claim 7 , wherein the probability factors comprise (i) a divergence factor between the matching routine route profile and the current route, and (ii) a reputation factor of a selected driver.

11. The anomaly detection system of claim 10 , wherein the reputation factor of the selected driver comprises (i) a complaint history of the selected driver, and (ii) reputation indicators of the selected driver pulled from third-party resources.

12. The anomaly detection system of claim 1 , wherein the executed instructions further cause the anomaly detection system to:
in response to the status query, receiving an indication of a negative condition from the mobile device of the respective user; and
in response to receiving the indication, transmit a driver directive to a mobile device of a driver providing service to the respective user indicating an emergency condition.

13. The anomaly detection system of claim 1 , wherein the executed instructions cause the anomaly detection system to monitor the current route traveled by the respective user by receiving location data from global positioning system (GPS) resources of the mobile device of the respective user.

14. The anomaly detection system of claim 1 , wherein the executed instructions further cause the anomaly detection system to:
compare the current route traveled by the respective user to a route inputted by a selected driver into a mapping resource of a mobile device of the selected driver.

15. The anomaly detection system of claim 1 , wherein the executed instructions further cause the anomaly detection system to:
provide a selectable emergency feature on a graphical user interface for the respective user during the current route; and
wherein a user selection of the selectable emergency feature triggers the probable anomaly.

16. A non-transitory computer readable medium storing instructions that, when executed by one or more processors of an anomaly detection system, cause the anomaly detection system to:
construct routine route profiles for users of a network service by, for each of the users, (i) collecting historical route data, and (ii) identifying correlated route shapes in the historical route data, wherein the correlated route shapes comprise a plurality of distinct routes between a common start point or region and a common end point or region;
monitor a current route traveled by a respective one of the users;
identify a matching routine route profile of the respective user;
determine a probable anomaly in the current route based on the matching routine route profile; and
in response to determining the probable anomaly, perform one or more actions in accordance with a safety protocol, wherein an initial action of the safety protocol comprises transmitting a status query to a mobile device of the respective user, and wherein a subsequent action in the safety protocol comprises contacting a predetermined user contact to report the probable anomaly.

17. The non-transitory computer readable medium of claim 16 , wherein the instructions further cause the anomaly detection system to, prior to monitoring the current route:
receive a request for service from the respective user; and
select a driver to provide the requested service.

18. A computer-implemented method for detecting route anomalies, the method performed by one or more processors of an anomaly detection system and comprising:
constructing routine route profiles for users of a network service by, for each of the users, (i) collecting historical route data, and (ii) identifying correlated route shapes in the historical route data, wherein the correlated route shapes comprise a plurality of distinct routes between a common start point or region and a common end point or region;
monitoring a current route traveled by a respective one of the users;
identifying a matching routine route profile of the respective user;
determining a probable anomaly in the current route based on the matching routine route profile; and
in response to determining the probable anomaly, performing one or more ac

Assignees

• Uber Technologies Inc

Inventors

• Truong Michael
• Purdy David
• Mawas Rami

Classifications

• E05F15/70 Primary

  with automatic actuation · CPC title

• H04L67/30

  Profiles · CPC title

• G06Q10/08

  Logistics, e.g. warehousing, loading or distribution; Inventory or stock management · CPC title

• G06Q10/083

  Shipping · CPC title

• H04W4/40

  for vehicles, e.g. vehicle-to-pedestrians [V2P] · CPC title