Method, apparatus, and system for determining a bicycle lane deviation for autonomous vehicle operation

Patent metadata

Field	Value
Publication number	US-12592149-B2
Application number	US-202117564944-A
Country	US
Kind code	B2
Filing date	Dec 29, 2021
Priority date	Dec 29, 2021
Publication date	Mar 31, 2026
Grant date	Mar 31, 2026

Abstract

An approach is provided for determining bicycle lane deviations for autonomous vehicle warning or operation. The approach, for example, involves retrieving probe data associated with a bicycle transportation mode. The approach also involves determining a plurality of probe points of the probe data that are map-matched outside of a bicycle lane. The approach further involves clustering the plurality of probe points into at least one location cluster. The approach further involves storing the one or more location clusters in a geographic database as respective one or more hazard areas where a plurality of bicycles deviates outside of the bicycle lane. By way of example, the approach can further involve using the at least one location cluster to perform at least one of providing a warning message or determining a driving parameter for an autonomous vehicle.

First claim

What is claimed is:

1 . A method comprising:
retrieving probe data associated with a bicycle transportation mode; determining a plurality of probe points of the probe data that are map-matched outside of a bicycle lane; clustering the plurality of probe points into at least one location cluster based on a clustering model configured to cluster archived probe points or selected probe points of the plurality of probe points for determining one or more geographic areas where a plurality of bicycles are not staying in a given bicycle lane, wherein the clustering model is configured to use a sliding window for selection of the archived probe points; storing the one or more location clusters in a geographic database as respective one or more hazard areas where a plurality of bicycles deviates outside of the bicycle lane; determining a confidence level of bicycle deviations for each of the one or more location clusters in the geographic databases based on a spatial relationship of the one or more location clusters to one or more bicycle lanes; determining a likelihood that a vehicle along a route is approaching the one or more location clusters; and providing an output signal for activation of one or more sensors coupled to the vehicle prior to the vehicle approaching the one or more location clusters and the determined confidence level, wherein activation of the one or more sensors includes one or more instructions to unretract a laser-based distance sensor from a retracted position to an unretracted position that enables the laser-based distance sensor to scan an area corresponding to the one or more location clusters.

2 . The method of claim 1 , further comprising:
using the at least one location cluster to perform at least one of: providing a warning message; or determining a driving parameter for an autonomous vehicle.

3 . The method of claim 1 , further comprising:
determining a confidence of bicycle deviations based on a location of the at least one location cluster with respect to the bicycle lane.

4 . The method of claim 3 , wherein the confidence of bicycle deviations is set to:
a low confidence based on determining there is no location cluster; a medium confidence based on determining that the at least one cluster is located partially within the bicycle lane and partially outside the bicycle lane; or a high confidence based on determining that the at least one location cluster is located completely outside the bicycle lane.

5 . The method of claim 4 , further comprising:
determining an autonomous vehicle driving response to approaching the at least one location cluster based on the confidence of bicycle deviations.

6 . The method of claim 5 , wherein the driving response is to move to another vehicle lane, to reduce vehicle speed, to engage a sensor to detect a bicyclist, or a combination thereof.

7 . The method of claim 1 , wherein the clustering is performed based on a specified minimum number of probe points, a specified distance radius, or a combination thereof.

8 . The method of claim 1 , further comprising:
providing data for generating a mapping display presenting a representation of the at least one cluster.

9 . The method of claim 1 , further comprising:
identifying a road blockage based on the at least one location cluster.

10 . The method of claim 1 , further comprising:
identifying that the bicycle lane is narrow based on the at least one cluster.

11 . An apparatus comprising:
at least one processor; and at least one memory including computer program code for one or more programs, the at least one memory and the computer program code configured to, with the at least one processor, cause the apparatus to perform at least the following, retrieve probe data associated with a bicycle transportation mode;
determine a plurality of probe points of the probe data that are map-matched outside of a bicycle lane;
cluster the plurality of probe points into at least one location cluster based on a clustering model configured to cluster archived probe points or selected probe points of the plurality of probe points for determining one or more geographic areas where a plurality of bicycles are not staying in a given bicycle lane, wherein the clustering model is configured to use a sliding window for selection of the archived probe points;
store the one or more location clusters in a geographic database as respective one or more hazard areas where a plurality of bicycles deviates outside of the bicycle lane;
determine a confidence level of bicycle deviations for each of the one or more location clusters in the geographic databases based on a spatial relationship of the one or more location clusters to one or more bicycle lanes;
determine a likelihood that a vehicle along a route is approaching the one or more location clusters; and
provide an output signal for activation of one or more sensors coupled to the vehicle prior to the vehicle approaching the one or more location clusters and the determined confidence level, wherein activation of the one or more sensors includes one or more instructions to unretract a laser-based distance sensor from a retracted position to an unretracted position that enables the laser-based distance sensor to scan an area corresponding to the one or more location clusters.

12 . The apparatus of claim 11 , wherein the apparatus is further caused to:
use the at least one location to perform at least one of: providing a warning message; or determining a driving parameter for an autonomous vehicle.

13 . The apparatus of claim 11 , wherein the apparatus is further caused to:
determine a confidence of bicycle deviations based on a location of the at least one location cluster with respect to the bicycle lane.

14 . The apparatus of claim 13 , wherein the confidence of bicycle deviations is set to:
a low confidence based on determining there is no location cluster; a medium confidence based on determining that the at least one cluster is located partially within the bicycle lane and partially outside the bicycle lane; or a high confidence based on determining that the at least one location cluster is located completely outside the bicycle lane.

15 . The apparatus of claim 14 , wherein the apparatus is further caused to:
determine an autonomous vehicle driving response to approaching the at least one location cluster based on the confidence of bicycle deviations.

16 . A non-transitory computer-readable storage medium carrying one or more sequences of one or more instructions which, when executed by one or more processors, cause an apparatus to perform:
retrieving probe data associated with a bicycle transportation mode; determining a plurality of probe points of the probe data that are map-matched outside of a bicycle lane; clustering the plurality of probe points into at least one location cluster based on a clustering model configured to cluster archived probe points or selected probe points of the plurality of probe points for determining one or more geographic areas where a plurality of bicycles are not staying in a given bicycle lane, wherein the clustering model is configured to use a sliding window for selection of the archived probe points; storing the one or more location clusters in a geographic database as respective one or more hazard areas where a plurality of bicycles deviates outside of the bicycle lane; determining a confidence level of bicycle deviations for each of the one or more location clusters in the geographic databases based on a spatial relationship of the one o

Assignees

• Here Global Bv

Inventors

• Stenneth Leon
• Nayak Amarnath
• Averbuch Alex
• Raut Advait Mohan
• Bernhardt Bruce

Classifications

• B60W2552/10

  Number of lanes · CPC title

• B60W2720/10

  Longitudinal speed · CPC title

• B60W2554/4041

  Position · CPC title

• B60W2554/4026

  Cycles · CPC title

• B60W2552/50

  Barriers · CPC title