Sensor coverage analysis for automated driving scenarios involving intersections
Examples described herein provide a computer-implemented method that includes defining, by a processing device, a plurality of parameters so that any orthogonal intersection can be described. The method further includes building, by the processing device, an orthogonal parameterized model that can represent any orthogonal intersection based at least in part on the plurality of parameters that can describe any intersection of interest. The method further includes expanding, by the processing device, the orthogonal parameterized model to generate a fully parameterized intersection model that accounts for intersection complexities. The method further includes building, by the processing device, a low-fidelity analytical that computes various metrics based on the fully parameterized intersection model.
1. A computer-implemented method comprising:
defining, by a processing device, a plurality of parameters so that any orthogonal intersection can be described;
building, by the processing device, an orthogonal parameterized model that can represent any orthogonal intersection based at least in part on the plurality of parameters that can describe any intersection of interest;
expanding, by the processing device, the orthogonal parameterized model to generate a fully parameterized intersection model that accounts for intersection complexities including accounting for a total lateral distance from a stop bar associated with the vehicle to a center left target line in meters and a total longitudinal distance from the stop bar associated with the vehicle to the center left target line in meters; and
building, by the processing device, a low-fidelity analytical that computes various metrics based on the fully parameterized intersection model.
2. The computer-implemented method of claim 1 , wherein building the orthogonal parameterized model further comprises representing a vehicle as an ellipse.
3. The computer-implemented method of claim 2 , wherein expanding the orthogonal parameterized model further comprises calculating the ellipse is based on the following formula for the vehicle when the vehicle is turning left:
EllipC
=
π
(
a
+
b
)
(
3
(
a
-
b
)
2
(
a
+
b
)
2
(
-
3
(
a
-
b
)
2
(
a
+
b
)
2
+
4
+
1
0
)
+
1
)
where “a” is the total lateral distance from a stop bar associated with the vehicle to the center left target line in meters, and where “b” is the total longitudinal distance from the stop bar associated with the vehicle to the center left target line in meters.
4. The computer-implemented method of claim 3 , further comprising calculating an arcturn distance of the vehicle based on the ellipse.
5. The computer-implemented method of claim 3 , further comprising calculating an arcturn time of the vehicle based on the ellipse.
6. The computer-implemented method of claim 1 , wherein the intersection complexities comprise at least one of a positive road curvature, a negative road curvature, and an intersection angle.
7. The computer-implemented method of claim 1 , wherein the intersection complexities comprise a positive road curvature, a negative road curvature, and an intersection angle.
8. The computer-implemented method of claim 1 , further comprising:
using vehicle sensor semantic detections and a high-definition map to obtain intersection properties;
using the low-fidelity analytical model to estimate a time required for a host vehicle to complete a maneuver;
obtaining a velocity, an acceleration, and a location relative to the host vehicle for a plurality of actors;
using the low-fidelity analytical model to estimate time for actor to reach an intersection or the host vehicle; and
displaying a message to a user whether it is safe to perform the maneuver.
9. The computer-implemented method of claim 8 , wherein the intersection properties comprise a number of lanes, a lane width, and a road curvature.
10. The computer-implemented method of claim 1 , further comprising:
setting parameter values for the intersection of interest;
setting a velocity, an acceleration, and a location relative to a host vehicle for a plurality of actors;
using the low-fidelity analytical model to estimate how far away from the host vehicle the plurality of actors need to be detected so that the host vehicle can safely complete a maneuver; and
designing a sensor placement arrangement that defines a location for each of a plurality of sensors associated with the host vehicle.
11. A system comprising:
a memory comprising computer readable instructions; and
a processing device for executing the computer readable instructions, the computer readable instructions controlling the processing device to perform operations comprising:
defining, by the processing device, a plurality of parameters so that any orthogonal intersection can be described;
building, by the processing device, an orthogonal parameterized model that can represent any orthogonal intersection based at least in part on the plurality of parameters that can describe any intersection of interest, wherein building the orthogonal parameterized model further comprises representing a vehicle as an ellipse;
expanding, by the processing device, the orthogonal parameterized model to generate a fully parameterized intersection model that accounts for intersection complexities by calculating the ellipse is based on the following formula for the vehicle when the vehicle is turning left:
EllipC
=
π
(
a
+
b
)
(
3
(
a
-
b
)
2
(
a
+
b
)
2
(
-
3
(
a
-
b
)
2
(
a
+
b
)
2
+
4
+
1
0
)
+
1
)
where “a” is a total lateral distance from a stop bar associated with the vehicle to a center left target line in meters, and where “b” is a total longitudinal distance from the stop bar associated with the vehicle to the center left target line in meters; and
building, by the processing device, a low-fidelity analytical that computes various metrics based on the fully parameterized intersection model.
12. The system of claim 11 , wherein the intersection complexities comprise at least one of a positive road curvature, a negative road curvature, and an intersection angle.
13. The system of claim 11 , wherein the intersection complexities comprise a positive road curvature, a negative road curvature, and an intersection angle.
14. The system of claim 11 , wherein the operations further comprise:
using vehicle sensor semantic detections and a high-definition map to obtain intersection properties;
using the low-fidelity analytical model to estimate a time required for a host vehicle to complete a maneuver;
obtaining a velocity, an acceleration, and a location relative to the host vehicle for a plurality of actors;
using the low-fidelity analytical model to estimate time for actor to reach an intersection or the host vehicle; and
displaying a message to a user whether it is safe to perform the maneuver.
15. The system of claim 14 , wherein the intersection properties comprise a number of lanes, a lane width, and a road curvature.
16. The system of claim 11 , wherein the operations further comprise:
setting parameter values for the intersection of interest;
setting a velocity, an acceleration, and a location relative to a host vehicle for a plurality of actors;
using the low-fidelity analytical model to estimate how far away from the host vehicle the plurality of actors need to be detected so that the host vehicle can safely complete a maneuver; and
designing a sensor placement arrangement that defines a location for each of a plurality of sensors associated with the host vehicle.
17. A computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a processor to cause the processor to perform operations comprising:
defining, by the processing device, a plurality of parameters so that any orthogonal intersection can be described;
building, by the processing device, an orthogonal parameterized model that can represent any orthogonal intersection based at least in part on the plurality of parameters that can describe any intersection of interest, wherein building the orthogonal parameterized model further comprises representing a vehicle as an ellipse;
expanding, by the processing device, the orthogonal parameterized model to generate a fully parameterized intersection model that accounts for intersection complexities, and wherein expanding the orthogonal parameterized model further comprises calculating the ellipse is based on the following formula for the vehicle when the vehicle is turning left:
EllipC
=
π
(
a
+
b
)
(
3
(
a
-
b
)
2
(
a
+
b
)
2
(
-
3
(
a
-
b
)
2
(
a
+
b
)
2
+
4
+
1
0
)
+
1
)
where “a” is a total lateral distance from a stop bar associated with the vehicle to a center left target line in meters, and where “b” is a total longitudinal distance from the stop bar associated with the vehicle to the center left target line in meters; and
building, by the processing device, a low-fidelity analytical that computes various metrics based on the fully parameterized intersection model.
18. The method of claim 8 , wherein the vehicle sensor semantic detections comprise data results from a plurality of vehicle sensors, and the data results include identified and detected objects.
19. The method of claim 18 , wherein the identified and detected objects include at least one of vehicles motorcycles and pedestrians.
20. The method of claim 14 , wherein the vehicle sensor semantic detections comprise data results from a plurality of vehicle sensors, and the data results include identified and detected objects.