Planning for unknown objects by an autonomous vehicle
Among other things, a world model is maintained of an environment of a vehicle. A hypothetical object in the environment that cannot be perceived by sensors of the vehicle is included in the world model.
1. A method comprising:
receiving, using at least one processor and based on data collected by one or more sensors of a vehicle, a model of an environment comprising one or more objects that are detectable by the one or more sensors;
determining, using the at least one processor and based on at least the collected data, one or more hypothetical objects in the environment that are undetectable by the one or more sensors;
adding, using the at least one processor, the one or more hypothetical objects to the model; and
controlling, using the at least one processor and based on the model to which the one or more hypothetical objects have been added, movement of the vehicle in the environment.
2. The method of claim 1 , wherein the one or more objects of the environment are observable from the vehicle, wherein the one or more hypothetical objects are unobservable from the vehicle.
3. The method of claim 1 , wherein the one or more sensors comprise a light detection and ranging (LIDAR) sensor.
4. The method of claim 3 , wherein the determining of the one or more hypothetical objects comprises:
receiving, using the at least one processor, a collection of points output by the LIDAR sensor;
determining, using the at least one processor and based on a ground elevation, that a first point in a particular vertical slice of the collection of points that is perpendicular to a ground plane characterizes a ground location; and
determining, using the at least one processor, that a different second point in the particular vertical slice of the collection of points characterizes an object that obscures the ground location, wherein the object that obscures the ground location is one of the one or more hypothetical objects.
5. The method of claim 1 , wherein the adding of the one or more hypothetical objects to the model comprises:
receiving, using the at least one processor, data describing a configuration of a region that is undetectable by the one or more sensors; and
placing, using the at least one processor and based on the configuration of the region, the one or more hypothetical objects at one or more respective locations in the model.
6. The method of claim 5 , wherein the hypothetical object is a vehicle, wherein the undetectable region comprises an unobservable road in the environment, wherein placing the vehicle at a respective location comprises placing the vehicle on an unperceived segment of the road.
7. The method of claim 1 , wherein:
the controlling is based on motion characteristics of the one or more hypothetical objects;
the motion characteristics comprise at least one of an orientation, a direction of motion, a speed, or a trajectory of motion of the one or more hypothetical objects; and
the orientation, direction, or speed are based on at least one of map data, geographic data, traffic data, or traffic law data.
8. The method of claim 1 , wherein the controlling comprises avoiding a collision with the one or more objects and the one or more hypothetical objects.
9. A system comprising:
at least one programmable processor; and
a machine-readable medium storing instructions that, when executed by the at least one processor, cause the at least one programmable processor to perform operations comprising:
receiving, based on data collected by one or more sensors of a vehicle, a model of an environment comprising one or more objects that are detectable by the one or more sensors;
determining, based on at least the collected data, one or more hypothetical objects in the environment that are undetectable by the one or more sensors;
adding the one or more hypothetical objects to the model; and
controlling, based on the model to which the one or more hypothetical objects have been added, movement of the vehicle in the environment.
10. A non-transitory computer program product storing instructions that, when executed by at least one programmable processor, cause the at least one programmable processor to perform operations comprising:
receiving, based on data collected by one or more sensors of a vehicle, a model of an environment comprising one or more objects that are detectable by the one or more sensors;
determining, based on at least the collected data, one or more hypothetical objects in the environment that are undetectable by the one or more sensors;
adding the one or more hypothetical objects to the model; and
controlling, based on the model to which the one or more hypothetical objects have been added, movement of the vehicle in the environment.
11. The system of claim 9 , wherein the one or more objects of the environment are observable from the vehicle, wherein the one or more hypothetical objects are unobservable from the vehicle.
12. The system of claim 9 , wherein the one or more sensors comprise a light detection and ranging (LIDAR) sensor.
13. The system of claim 12 , wherein the determining of the one or more hypothetical objects comprises:
receiving, using the at least one processor, a collection of points output by the LIDAR sensor;
determining, using the at least one processor and based on a ground elevation, that a first point in a particular vertical slice of the collection of points that is perpendicular to a ground plane characterizes a ground location; and
determining, using the at least one processor, that a different second point in the particular vertical slice of the collection of points characterizes an object that obscures the ground location, wherein the object that obscures the ground location is one of the one or more hypothetical objects.
14. The system of claim 9 , wherein the adding of the one or more hypothetical objects to the model comprises:
receiving, using the at least one processor, data describing a configuration of a region that is undetectable by the one or more sensors; and
placing, using the at least one processor and based on the configuration of the region, the one or more hypothetical objects at one or more respective locations in the model.
15. The system of claim 14 , wherein the hypothetical object is a vehicle, wherein the undetectable region comprises an unobservable road in the environment, wherein placing the vehicle at a respective location comprises placing the vehicle on an unperceived segment of the road.
16. The system of claim 9 , wherein:
the controlling is based on motion characteristics of the one or more hypothetical objects;
the motion characteristics comprise at least one of an orientation, a direction of motion, a speed, or a trajectory of motion of the one or more hypothetical objects; and
the orientation, direction, or speed are based on at least one of map data, geographic data, traffic data, or traffic law data.
17. The system of claim 9 , wherein the controlling comprises avoiding a collision with the one or more objects and the one or more hypothetical objects.
18. The non-transitory computer program product of claim 10 , wherein the one or more objects of the environment are observable from the vehicle, wherein the one or more hypothetical objects are unobservable from the vehicle.
19. The non-transitory computer program product of claim 10 , wherein the one or more sensors comprise a light detection and ranging (LIDAR) sensor.
20. The non-transitory computer program product of claim 19 , wherein the determining of the one or more hypothetical objects comprises:
receiving, using the at least one processor, a collection of points output by the LIDAR sensor;
determining, using the at least one processor and based on a ground elevation, that a first point in a particular vertical slice of the collection of points that is perpendicular to a ground plane characterizes a ground location; and
determining, using the at least one processor, that a different second point in the particular vertical slice of the collection of points characterizes an object that obscures the ground location, wherein the object that obscures the ground location is one of the one or more hypothetical objects.