Systems and methods for unprotected left turns in high traffic situations in autonomous vehicles

Patent metadata

Field	Value
Publication number	US-10198002-B2
Application number	US-201715820272-A
Country	US
Kind code	B2
Filing date	Nov 21, 2017
Priority date	Nov 21, 2017
Publication date	Feb 5, 2019
Grant date	Feb 5, 2019

Abstract

A method in an autonomous vehicle comprises determining to perform a left turn maneuver when the vehicle is in a stopping zone, the vehicle is clear of approaching vehicles, and a relevant traffic signal displays a go signal. The method further comprises determining to perform the left turn maneuver when the vehicle has entered a dilemma zone, the vehicle is clear of approaching vehicles, and the relevant traffic signal displays a go signal, a caution signal, or has displayed a stop signal for less than a predetermined amount of time. The method further comprises determining to perform the left turn maneuver when the vehicle has entered a cross-traffic zone, the vehicle is clear of approaching vehicles, and the relevant traffic signal displays a go signal, a caution signal, or a stop signal.

First claim

What is claimed is:

1. A processor-implemented method in an autonomous vehicle for executing an unprotected left turn maneuver, the method comprising:
determining, by a processor, a current zone of the vehicle to be at least one of a first zone, a second zone, or a third zone;
determining, by the processor, to perform a left turn maneuver when the vehicle is in the first zone, clear of approaching vehicles and a relevant traffic signal displays a go signal;
determining, by the processor, to perform the left turn maneuver when the vehicle is in the second zone, clear of approaching vehicles and the relevant traffic signal displays a go signal, a caution signal, or has displayed a stop signal for less than a predetermined amount of time;
determining, by the processor, to perform the left turn maneuver when, the vehicle is in the third zone or the velocity of the vehicle will propel the vehicle into the third zone, clear of approaching vehicles; and
communicating, by the processor, a message to vehicle controls conveying that a stop point may be lifted after it has been determined that the vehicle can perform the left turn maneuver.

2. The method of claim 1 , wherein:
the first zone comprises a stopping zone wherein the stopping zone ends at a stop line for the lane;
the second zone comprises a dilemma zone wherein the dilemma zone begins at the stop line for the lane and ends where the lane is intersected by cross-traffic; and
the third zone comprises a cross-traffic zone wherein the cross-traffic zone begins where the lane is intersected by cross-traffic.

3. The method of claim 1 , further comprising:
determining, by the processor, that the vehicle has approached an intersection at which the vehicle will attempt an unprotected left turn maneuver;
monitoring, by the processor, vehicle position data to determine whether the vehicle is positioned in the first zone, in the second zone, or the third zone;
monitoring, by the processor, for one or more approaching vehicles in one or more opposing lanes to determine if the vehicle is clear of the one or more approaching vehicles that will cross within a desired turn path of the vehicle; and
monitoring, by the processor, a traffic control device to determine if a relevant traffic signal for the lane in which the vehicle is located displays a go signal, a caution signal, or a stop signal.

4. The method of claim 3 , wherein monitoring vehicle position data comprises retrieving vehicle localization data that identifies the location of the vehicle on an internal vehicle map of the world.

5. The method of claim 4 , wherein the vehicle localization data is derived from GPS sensor data.

6. The method of claim 3 , wherein monitoring for one or more approaching vehicles comprises identifying one or more oncoming objects from camera image, radar, or lidar data.

7. The method of claim 6 , wherein determining that the vehicle is clear of approaching vehicles comprises:
predicting the movement of the one or more oncoming objects; and
predicting whether the one or more predicted objects are traveling slow enough or will come to a stop to allow the vehicle to complete the left turn maneuver.

8. The method of claim 3 , wherein monitoring a traffic control device comprises retrieving camera image data that identifies whether the relevant traffic signal displays a green signal, a yellow signal, or a red signal.

9. A system for controlling an autonomous vehicle comprising an intersection state machine that comprises one or more processors configured by programming instructions encoded in non-transient computer readable media, for an autonomous vehicle approaching an intersection at which the vehicle will attempt an unprotected left turn maneuver from a lane the intersection state machine configured to:
determine a current zone of the vehicle to be at least one of a first zone, a second zone, or a third zone;
determine to perform the left turn maneuver when the vehicle is in the first zone, clear of approaching vehicles from all opposing lanes, and a relevant traffic signal displays a go signal;
determine to perform the left turn maneuver when the vehicle is in the second zone, clear of approaching vehicles from all opposing lanes and the relevant traffic signal displays a go signal, a caution signal, or has displayed a stop signal for less than a predetermined amount of time; and
determine to perform the left turn maneuver when the vehicle is in the third zone or the velocity of the vehicle will propel the vehicle into the third zone, clear of approaching vehicles and the relevant traffic signal displays a go signal, a caution signal, or a stop signal.

10. The system of claim 9 , wherein:
the first zone comprises a stopping zone wherein the stopping zone ends at a stop line for the lane;
the second zone comprises a dilemma zone wherein the dilemma zone begins at the stop line for the lane and ends where the lane is intersected by cross-traffic; and
the third zone comprises a cross-traffic zone wherein the cross-traffic zone begins where the lane is intersected by cross-traffic.

11. The system of claim 9 , further comprising a world observer module that comprises one or more processors configured by programming instructions encoded in non-transient computer readable media, the world observer module configured to:
determine that the vehicle has approached an intersection at which the vehicle will attempt the unprotected left turn maneuver; and
invoke the intersection state machine to determine when the vehicle may perform the left turn maneuver.

12. The system of claim 10 , further comprising a vehicle position module that comprises one or more processors configured by programming instructions encoded in non-transient computer readable media, the vehicle position module configured to:
monitor vehicle position data to determine whether the vehicle is positioned in the stopping zone, the dilemma zone, or the cross-traffic zone; and
provide the identity of the zone type in which the vehicle is positioned to the intersection state machine.

13. The system of claim 12 , wherein the vehicle position module is configured to monitor vehicle position data by retrieving vehicle localization data derived from GPS sensor data that identifies the location of the vehicle on an internal vehicle map of the world.

14. The system of claim 9 , further comprising a predicted objects module that comprises one or more processors configured by programming instructions encoded in non-transient computer readable media, the predicted objects module configured to:
monitor for one or more approaching vehicles in one or more opposing lanes to determine if the vehicle is clear of approaching vehicles that will cross within a desired turn path of the vehicle; and
provide an indication as to whether the vehicle is clear of approaching vehicles that will cross within a desired turn path of the vehicle to the intersection state machine.

15. The system of claim 14 , wherein the predicted objects module is configured to monitor for one or more approaching vehicles using camera image, radar, or lidar data.

16. The system of claim 14 , wherein the predicted objects module is configured to determine that the vehicle is clear of approaching vehicles by:
predicting the movement of the one or more oncoming objects; and
predicting whether the one or more predicted objects are traveling slow enough or will come to a stop to allow the vehicle to complete the left turn maneuver.

17. The system of claim 9 , further comprising a traffic light monitoring module that co

Assignees

• Gm Global Tech Operations Llc

Inventors

• Holben Ryan

Classifications

• B60W2556/40

  High definition maps · CPC title

• B60W60/0011

  involving control alternatives for a single driving scenario, e.g. planning several paths to avoid obstacles · CPC title

• B60W2554/00

  Input parameters relating to objects · CPC title

• B60W2556/50

  of positioning data, e.g. GPS [Global Positioning System] data · CPC title

• B60W2555/60

  Traffic rules, e.g. speed limits or right of way · CPC title