Adaptive and Personalized Navigation System

Patent metadata

Field	Value
Publication number	US-2016209228-A1
Application number	US-201615080781-A
Country	US
Kind code	A1
Filing date	Mar 25, 2016
Priority date	Nov 2, 2006
Publication date	Jul 21, 2016
Grant date	—

Abstract

Adaptive navigation techniques are disclosed that allow navigation systems to learn from a user's personal driving history. As a user drives, models are developed and maintained to learn or otherwise capture the driver's personal driving habits and preferences. Example models include road speed, hazard, favored route, and disfavored route models. Other attributes can be used as well, whether based on the user's personal driving data or driving data aggregated from a number of users. The models can be learned under explicit conditions (e.g., time of day/week, driver ID) and/or under implicit conditions (e.g., weather, drivers urgency, as inferred from sensor data). Thus, models for a plurality of attributes can be learned, as well as one or more models for each attribute under a plurality of conditions. Attributes can be weighted according to user preference. The attribute weights and/or models can be used in selecting a best route for user.

First claim

1 .- 20 . (canceled)

21 . A non-transitory machine-readable storage medium encoded with instructions that, when executed by one or more processors, cause the processor to carry out a process for generating one or more attribute models learned from a user's driving preferences, the process comprising:
receiving attribute data for a set of driving sessions for a user, wherein attribute data for each driving session includes measurements relevant to one or more target attributes, wherein each driving session is defined in terms of one or more road segments of one or more roads traversed by the user during the set of driving sessions; applying attribute estimation rules to the attribute data to compute an attribute value for each target attribute along each road segment covered by each driving session; assigning an attribute value for one or more unseen road segments of the one or more roads identified in each driving session, wherein unseen road segments correspond to road segments not yet traversed by the user during the driving sessions; determining and storing an attribute model comprising attribute values computed or assigned for each road segment of the one or more roads traversed by the user during the set of driving sessions; and accessing the attribute model to generate directions for use in navigation.

22 . The non-transitory machine-readable medium of claim 21 , wherein the instructions cause the processor to carry out a process further comprising:
determining if one or more explicit conditions applies to each target attribute, and when an explicit condition is determined to apply to a given target attribute, the instructions cause the processor to carry out a process further comprising:
defining a bucket for each explicit condition of each given target attribute;
storing attribute data for each road segment from each driving session into a corresponding bucket; and
determining and storing a conditional variant model for each bucket from attribute values computed or assigned from merged attribute data stored in each bucket.

23 . The non-transitory machine-readable medium of claim 21 , wherein the instructions cause the processor to carry out a process further comprising:
determining if one or more implicit conditions applies to each target attribute, and when an implicit condition is determined to apply to a given target attribute, the instructions cause the processor to carry out a process further comprising:
forming a mini-model for each given target attribute by merging attribute data for each driving session;
identifying pairs of like mini-models using a similarity metric;
merging attribute data from identified pairs of like mini-models; and
determining and storing a combined model from attribute values computed or assigned from attribute data merged from identified pairs of like mini-models.

24 . The non-transitory machine-readable storage medium of claim 21 , wherein the attribute estimation rules comprise one or more direct rules whereby an attribute value is measured directly.

25 . The non-transitory machine-readable storage medium of claim 21 , wherein the attribute estimation rules comprise one or more indirect rules whereby an attribute value is inferred or derived from one or more measurements.

26 . The non-transitory machine-readable storage medium of claim 21 , wherein assigning a default value for one or more unseen road segments comprises computing a summary statistic for one or more unseen road segments, the summary statistic assigning a statistically estimated attribute value for each unseen road segment.

27 . The non-transitory machine-readable storage medium of claim 26 , wherein computing a summary statistic for an unseen road segment comprises determining a mean or median attribute value observed in other driving sessions for the unseen road segment.

28 . The non-transitory machine-readable storage medium of claim 21 , wherein the one or more target attributes comprise one or more of road speed, favored routes, disfavored routes and hazardous routes.

29 . The non-transitory machine-readable medium of claim 21 , wherein
receiving attribute data for a set of driving sessions for a user comprises receiving sensor data.

30 . The non-transitory machine-readable medium of claim 29 , wherein the sensor data comprises data from one or more of a GPS receiver and a dead-reckoning sensor.

31 . The non-transitory machine-readable medium of claim 21 , wherein receiving attribute data for a set of driving sessions for a user comprises receiving user inputted data.

32 . A computer-implemented method of generating one or more attribute models learned from a user's driving preferences, comprising:
receiving, by one or more processors, attribute data for a set of driving sessions for a user, wherein attribute data for each driving session includes measurements relevant to one or more target attributes, wherein each driving session is defined in terms of one or more road segments of one or more roads traversed by the user during the set of driving sessions; applying, by the one or more processors, attribute estimation rules to the attribute data to compute an attribute value for each target attribute along each road segment covered by each driving session; assigning, by the one or more processors, an attribute value for one or more unseen road segments of the one or more roads identified in each driving session, wherein unseen road segments correspond to road segments not yet traversed by the user during the driving sessions; determining and storing, by the one or more processors, an attribute model comprising attribute values computed or assigned for each road segment of the one or more roads traversed by the user during the set of driving sessions; and accessing, by the one or more processors, the attribute model to generate directions for use in navigation.

33 . The computer-implemented method of claim 32 , further comprising:
determining, by the one or more processors, if one or more explicit conditions applies to each target attribute, and when an explicit condition is determined to apply to a given target attribute, the method further comprising:
defining, by the one or more processors, a bucket for each explicit condition of each given target attribute;
storing, by the one or more processors, attribute data for each road segment from each driving session into a corresponding bucket; and
determining and storing, by the one or more processors, a conditional variant model for each bucket from attribute values computed or assigned from merged attribute data stored in each bucket.

34 . The computer-implemented method of claim 32 , further comprising:
determining, by the one or more processors, if one or more implicit conditions applies to each target attribute, and when an implicit condition is determined to apply to a given target attribute, the method further comprising:
forming, by the one or more processors, a mini-model for each given target attribute by merging attribute data for each driving session;
identifying, by the one or more processors, pairs of like mini-models using a similarity metric;
merging, by the one or more processors, attribute data from identified pairs of like mini-models; and
determining and storing, by the one or more processors, a combined model from attribute values computed or assigned from attribute data merged from identified pairs of like mini-models.

35 . The computer-implemented method of claim 32 , wherein assigning a default value for one or more unseen road segments comprises computing a summar

Assignees

• Google Inc

Inventors

• Golding Andrew R
• Rasmussen Jens Eilstrup

Classifications

• G01C21/3484 Primary

  Personalized, e.g. from learned user behaviour or user-defined profiles · CPC title

• G01C21/36 Primary

  Input/output arrangements for on-board computers · CPC title

• G01C21/3492 Primary

  employing speed data or traffic data, e.g. real-time or historical (traffic control systems for road vehicles involving transmission of navigation instructions to the vehicle G08G1/0968) · CPC title