Correcting speed estimations using aggregated telemetry data

Patent metadata

Field	Value
Publication number	US-11686590-B2
Application number	US-201916718090-A
Country	US
Kind code	B2
Filing date	Dec 17, 2019
Priority date	Dec 17, 2019
Publication date	Jun 27, 2023
Grant date	Jun 27, 2023

Abstract

A method for correcting speed estimates for route planning using a machine-learned speed correction model trained on aggregated road data. Location and movement data collected from a plurality of mobile computing devices is aggregated on a server computer and used to train a speed correction model to correct estimated speeds corresponding to roads in one or more geographic regions. Speeds estimates for a road segment in a geographic region are corrected using a speed correction model trained on road data describing road segments in the same geographic region. In some embodiments, road data corresponding to one or more geographic regions is assigned to groups in training the speed correction model. The road data may be anonymized or segmented such that an originating device or route is unidentifiable. More fine-grained speed correction models may also be trained for different or additional factors than geographic region, such as day and/or time.

First claim

What is claimed is:

1. A computer-implemented method for correcting travel speed estimates on a computing device, the method comprising:
estimating, for an electronic map comprising road data specifying a plurality of roads in a plurality of geographic regions, a speed associated with a segment of a road of the plurality of roads;
identifying a speed correction mapping configured to transform estimated speeds to corrected speeds, the speed correction mapping determined from a subset of the road data, wherein the speed correction mapping is provided by a machine learned speed correction model trained to take estimated speeds as input and output corrected speeds;
transforming the estimated speed to a corrected speed based on the speed correction mapping; and
sending for display the corrected speed as part of an electronic map application.

2. The method of claim 1 , wherein the subset of the road data specifies a set of roads from the plurality of roads in one or more geographic regions, the one or more geographic regions including the geographic region of the segment of the road associated with the speed.

3. The method of claim 1 , wherein training the speed correction model comprises:
for each of a plurality of road segments in the one or more geographic regions:
determining one or more estimated speed and actual speed pairs on the road segment based on the road data;
determining an error for each of the one or more estimated speed and actual speed pairs;

encoding the errors of each of the plurality of road segments in one or more feature vectors; and
inputting the one or more feature vectors into the speed correction model, the speed correction model configured to transform a plurality of estimated speeds to a plurality of corrected speeds for the one or more geographic regions.

4. The method of claim 3 , wherein encoding the errors in the one or more feature vector comprises reducing one or more dimensions of the vector.

5. The method of claim 3 , wherein inputting the one or more feature vectors into the speed correction model comprises:
assigning the one or more feature vectors to a group of feature vectors based on a similarity of the errors of each of the plurality of road segments;
processing the feature vectors in the group into a combined feature vector; and
inputting the combined feature vector into the speed correction model.

6. The method of claim 5 , wherein processing the feature vectors in the group into the combined feature vector comprises:
determining a set of feature vectors from the group of features vectors derived from road data generated within a time range;
deriving the combined feature vector from the set of feature vectors.

7. The method of claim 5 , wherein the group of feature vectors correspond to geographic regions which are adjacent to each other on the electronic map.

8. The method of claim 3 , wherein each feature vector is a matrix with a first dimension corresponding to a range of estimated speeds and a second dimension corresponding to a range of actual speeds, each value in the matrix representing the frequency of a distinct estimated speed and actual speed pair.

9. The method of claim 1 , further comprising:
converting the corrected speed to a second corrected speed using a second speed correction mapping, the second model trained on a second set of the road data.

10. The method of claim 1 , wherein the speed correction mapping transforms a first range of estimated speeds to a second range of corrected speeds.

11. The method of claim 1 , wherein the road data is filtered to include only road data classified as having a given type of device movement.

12. The method of claim 1 , wherein determining the speed correction mapping comprises:
filtering the road data based on one or more of a set of attributes of the road data, the attributes selected from the group consisting of a geographic position, an estimation optimism bias, a time of day, a day of the week, a road segment type, and a traffic congestion coefficient.

13. The method of claim 1 , further comprising:
filtering the subset of road data to include road data generated within a time range.

14. The method of claim 1 , wherein the road data has been processed using one or more of the techniques selected from the group consisting of anonymizing, segmenting, or filtering.

15. A non-transitory computer-readable storage medium of a computing device storing executable computer instructions, the instructions configured to, when executed by a hardware processor on the mobile computing device, perform steps comprising:
estimating, for an electronic map comprising road data specifying a plurality of roads in a plurality of geographic regions, a speed associated with a segment of a road of the plurality of roads;
identifying a speed correction mapping configured to transform estimated speeds to corrected speeds, the speed correction mapping determined from a subset of the road data, wherein the speed correction mapping is provided by a machine learned speed correction model trained to take estimated speeds as input and output corrected speeds;
transforming the estimated speed to a corrected speed based on the speed correction mapping; and
sending for display the corrected speed as part of an electronic map application.

16. The computer readable storage medium of claim 15 , wherein the subset of the road data specifies a set of roads from the plurality of roads in one or more geographic regions, the one or more geographic regions including the geographic region of the segment of the road associated with the speed.

17. A computing device computer system, the computing device comprising a hardware processor, the hardware processor configured to perform steps comprising:
estimating, for an electronic map comprising road data specifying a plurality of roads in a plurality of geographic regions, a speed associated with a segment of a road of the plurality of roads;
identifying a speed correction mapping configured to transform estimated speeds to corrected speeds, the speed correction mapping determined from a subset of the road data, wherein the speed correction mapping is provided by a machine learned speed correction model trained to take estimated speeds as input and output corrected speeds;
transforming the estimated speed to a corrected speed based on the speed correction mapping; and
sending for display the corrected speed as part of an electronic map application.

18. The computing device of claim 17 , wherein the subset of the road data specifies a set of roads from the plurality of roads in one or more geographic regions, the one or more geographic regions including the geographic region of the segment of the road associated with the speed.

Assignees

• Mapbox Inc

Inventors

• Sanin Riano Camilo
• Tacchi Ruggero Altair
• Gold David Ariel

Classifications

• G08G1/0129

  for creating historical data or processing based on historical data · CPC title

• G08G1/0112

  from the vehicle, e.g. floating car data [FCD] · CPC title

• G01C21/3841

  Data obtained from two or more sources, e.g. probe vehicles · 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

• G01C21/16

  by integrating acceleration or speed, i.e. inertial navigation · CPC title