Autonomous vehicle maneuvering based upon risk associated with occluded regions

1. A method executed by a computing system of an autonomous vehicle (AV), the method comprising:

defining an occluded region based upon output of a sensor of the AV, wherein the occluded region is a spatial region where the AV lacks visibility, and further wherein the occluded region comprises a portion of a roadway;

hypothesizing that a vehicle is travelling on the portion of the roadway in the occluded region towards the AV;

inferring that the vehicle will decelerate upon the AV becoming visible to a driver of the vehicle, wherein inferring that the vehicle will decelerate comprises computing an interaction point where the vehicle is expected to come to a stop after decelerating due to the AV becoming visible to the driver of the vehicle;

based upon the computed interaction point, identifying a driving maneuver for the AV to undertake; and

responsive to identifying the driving maneuver, causing the AV to perform the identified driving maneuver.

2. The method of claim 1 , wherein defining the occluded region comprises:

identifying an object in a field of view of the sensor; and

responsive to identifying the object, determining a location of the object relative to the AV, wherein the occluded region is defined based upon the determined location of the object relative to the AV.

3. The method of claim 1 , further comprising:

computing a position where the AV is expected to first be visible to the driver of the vehicle when the vehicle is at the position, wherein the position is within the occluded region, and further wherein the driving maneuver is identified based upon the computed position.

4. The method of claim 3 , further comprising:

estimating a reaction time of the vehicle, wherein the reaction time of the vehicle is time between when the AV is first visible to the driver of the vehicle and when the vehicle will begin to decelerate upon observing the AV, and further wherein the driving maneuver is identified based upon the estimated reaction time.

5. The method of claim 4 , further comprising:

estimating a velocity of the vehicle at the computed position, wherein the driving maneuver is identified based upon the estimated velocity of the vehicle at the computed position.

6. The method of claim 5 , further comprising:

assigning a deceleration to the vehicle based upon a profile of the vehicle that is retained in computer-readable storage;

computing a time value that is indicative of when the vehicle will reach a predefined location, wherein the time value is computed based upon the computed position of the vehicle, the reaction time of the vehicle, the estimated velocity of the vehicle, and the deceleration assigned to the vehicle; and

computing an amount of time required for the AV to complete the driving maneuver, wherein the driving maneuver is identified based upon the time value and the amount of time required for the AV to complete the driving maneuver.

7. The method of claim 6 , wherein the driving maneuver is a turn through an intersection.

8. The method of claim 5 , further comprising:

assigning a deceleration to the vehicle based upon a profile of the vehicle that is retained in computer-readable storage

wherein the computed interaction point is computed based upon the computed position of the vehicle, the reaction time of the vehicle, the estimated velocity of the vehicle, and the deceleration assigned to the vehicle.

9. The method of claim 8 , wherein identifying the driving maneuver comprises:

computing a velocity curve for the AV based upon the computed interaction point, wherein the velocity curve is indicative of velocity of the AV over time, wherein when the AV operates in accordance with the velocity curve the AV will decelerate and come to a stop prior to the computed interaction point, and further wherein the velocity curve is computed based upon:

a distance between the AV and the computed interaction point; and

a computer-readable profile of the AV, the computer-readable profile of the AV comprises data that is indicative of a maximum velocity of the AV, a maximum acceleration of the AV, and a maximum deceleration of the AV, and further wherein the driving maneuver is identified based upon the computed velocity curve for the AV.

10. The method of claim 9 , wherein the driving maneuver comprises exiting a first lane and entering into a second lane to navigate around an obstruction that causes the occluded region to exist from the perspective of the AV, wherein traffic in the first lane and the second lane flows in opposite directions.

11. An autonomous vehicle (AV) comprising:

a sensor that is configured to output a sensor signal; and

a computing system that is in communication with the sensor, the computing system programmed with instructions that, when executed by the computing system, cause the computing system to perform acts comprising:

based upon the sensor signal output by the sensor, identifying an occluded region from the perspective of the AV, wherein the occluded region is a spatial region where the AV lacks visibility, and further wherein the occluded region comprises a portion of a roadway;

responsive to identifying the occluded region, hypothesizing that a vehicle is in the occluded region on the portion of the roadway and travelling towards the AV;

inferring that the vehicle will decelerate upon the AV becoming visible to a driver of the vehicle, wherein inferring that the vehicle will decelerate comprises computing a position where the vehicle is expected to come to a stop after decelerating due to the AV becoming visible to the driver of the vehicle;

identifying a driving maneuver based upon the computed position; and

responsive to identifying the driving maneuver, causing the AV to perform the driving maneuver.

12. The AV of claim 11 , wherein identifying the occluded region comprises:

constructing a model of a geographic region based upon a map and the sensor signal; and

identifying a region in the model where information is lacking, wherein the occluded region corresponds to the region in the model.

13. The AV of claim 11 , wherein the AV is in a first lane of a roadway, the occluded region comprises a portion of a second lane of the roadway that is adjacent to the first lane of the roadway, and further wherein traffic in the first lane and the second lane flows in opposite directions.

14. The AV of claim 13 , wherein the driving maneuver comprises exiting the first lane of the roadway and entering the second lane of the roadway to navigate around an object in the first lane of the roadway.

15. The AV of claim 13 , wherein the driving maneuver comprises exiting the first lane and crossing over the second lane when performing a turn through an intersection.

16. The AV of claim 13 , wherein the vehicle is travelling in the second lane in a direction that is opposite a direction of travel of the AV, the acts further comprising estimating an amount of deceleration of the vehicle upon the AV becoming visible to the driver of the vehicle.

17. The AV of claim 16 , wherein the computed position is computed based upon the estimated amount of deceleration.

18. An autonomous vehicle (AV) comprising:

a computing system comprising:

a processor; and

memory storing instructions that, when executed by the processor, cause the processor to perform acts comprising:

defining an occluded region from the perspective of the AV, wherein the occluded region is a spatial region where the AV lacks visibility, wherein the occluded region comprises a portion of a traffic lane where vehicles are travelling in a direction that is opposite a direction of travel of the AV;

hypothesizing that a vehicle is in the occluded region and travelling towards the AV in the traffic lane;

inferring that the vehicle will decelerate upon the AV becoming visible to a driver of the vehicle;

computing a position where the vehicle is expected to come to a stop after decelerating due to the AV becoming visible to the driver of the vehicle; and

based upon the inferring that the vehicle will decelerate upon the AV becoming visible to the driver of the vehicle and the computed position, causing the AV to enter the traffic lane despite the occluded region possibly including a vehicle that is traveling in the traffic lane.

19. The AV of claim 18 , the acts further comprising assigning a deceleration value to the vehicle and a velocity value to the vehicle, wherein the position where the vehicle is expected to come to a stop after decelerating is computed based upon the deceleration value and the velocity value.