Automatically responding to emergency service vehicles by an autonomous vehicle
An autonomous vehicle that automatically responds to an emergency service vehicle. Perception data is captured or received by the autonomous vehicle (AV) cameras, sensors, and microphones. The perception data is used to detect lanes in a road, generate a list of objects, classify one or more of the objects, and determine a state for the classified objects. In response to detecting an emergency service vehicle such as a law enforcement vehicle, a responsive action may be planned by the AV. The responsive action may include sending an acknowledgement signal, notifying passengers of the AV, generating a planned trajectory based on the emergency service vehicle location, an emergency service vehicle detected intention, and other information. A safety check may be performed, and commands may be generated by a control module to navigate the AV along the planned trajectory. The commands may then be executed by a DBW module.
1. A system for automatically responding to an emergency service vehicle, comprising:
a data processing system comprising one or more processors, memory, a planning module, and a control module, the data processing system to:
identify an object on a road based on received sensor data;
classify the identified object as a type of emergency service vehicle based on the sensor data;
determine the intention of the type of emergency service vehicle at least in part based on a location of the type of emergency service vehicle;
generate a decision to respond to the emergency service vehicle based on the classification and the intention, wherein generating a decision includes:
generating a plurality of trajectories for an autonomous vehicle based on the emergency service vehicle intention and location, detected lane lines, and detected boundaries of the road,
generating a score for each of the plurality of trajectories, and
selecting the trajectory with the highest score; and
generate one or more commands to navigate along the selected trajectory.
2. The system of claim 1 , wherein the type of emergency service vehicle is a law enforcement vehicle.
3. The system of claim 2 , wherein determining the intention of the law enforcement vehicle includes:
detecting a state of the law enforcement vehicle; and
detecting the intention of the law enforcement vehicle based on the location of the law enforcement vehicle and the state of the law enforcement vehicle.
4. The system of claim 3 , wherein the state of the law enforcement vehicle includes detecting a siren emitting from the law enforcement vehicle.
5. The system of claim 3 , wherein the state of the law enforcement vehicle includes detecting flashing lights on law enforcement vehicle.
6. The system of claim 2 , wherein determining the intention of the law enforcement vehicle includes detecting the location of the law enforcement vehicle with respect to detected lane lines on the road.
7. The system of claim 1 , wherein the intention includes one of warning traffic, pulling over a car that is automatically responding to the emergency service vehicle, quickly navigating through traffic, and driving normally.
8. The system of claim 1 , wherein classifying includes:
performing a first classification to classify the identified object as a vehicle; and
performing a second classification to classify the vehicle as a law enforcement vehicle.
9. The system of claim 1 , wherein the type of emergency service vehicle is a fire truck or an ambulance, and the intention is stopping in a road, quickly navigating through traffic, and driving normally.
10. The system of claim 1 , wherein the data processing system generates an acknowledgement to the emergency service vehicle.
11. The system of claim 1 , wherein the data processing system generates a notification to one or more passengers inside the autonomous vehicle, the notification based on the decision to respond by the autonomous vehicle.
12. A non-transitory computer readable storage medium having embodied thereon a program, the program being executable by a processor to perform a method for automatically responding to an emergency service vehicle, the method comprising:
identifying an object on a road based on received sensor data;
classifying the identified object as a type of emergency service vehicle based on the sensor data;
determining the intention of the type of emergency service vehicle at least in part based on a location of the type of emergency service vehicle;
generating a decision to respond to the emergency service vehicle based on the classification and the intention, wherein generating a decision includes:
generating a plurality of trajectories for an autonomous vehicle based on the emergency service vehicle intention and location, detected lane lines, and detected boundaries of the road,
generating a score for each of the plurality of trajectories, and
selecting the trajectory with the highest score; and
generating one or more commands to navigate along the selected trajectory.
13. The non-transitory computer readable storage medium of claim 12 , wherein the type of emergency service vehicle is a law enforcement vehicle.
14. The non-transitory computer readable storage medium of claim 13 , wherein determining the intention of the law enforcement vehicle includes:
detecting a state of the law enforcement vehicle; and
detecting the intention of the law enforcement vehicle based on a location of the law enforcement vehicle and the state of the law enforcement vehicle.
15. The non-transitory computer readable storage medium of claim 14 , wherein the state of the law enforcement vehicle includes detecting a siren emitting from the law enforcement vehicle.
16. The non-transitory computer readable storage medium of claim 14 , wherein the state of the law enforcement vehicle includes detecting flashing lights on law enforcement vehicle.
17. The non-transitory computer readable storage medium of claim 13 , wherein determining the intention of the law enforcement vehicle includes detecting the location of the law enforcement vehicle with respect to detected lane lines on the road.
18. The non-transitory computer readable storage medium of claim 12 , wherein the intention includes one of warning traffic, pulling over a car that is automatically responding to the emergency service vehicle, quickly navigating through traffic, and driving normally.
19. The non-transitory computer readable storage medium of claim 12 , wherein the data processing system generates an acknowledgement to the emergency service vehicle.
20. The non-transitory computer readable storage medium of claim 12 , wherein the data processing system generates a notification to one or more passengers inside the autonomous vehicle, the notification based on the decision to respond by the autonomous vehicle.
21. A method for automatically responding to an emergency service vehicle, comprising:
identifying, by a data processing system stored in memory and executed by one or more processors, an object on a road based on received sensor data;
classifying, by a data processing system, the identified object as a type of emergency service vehicle based on the sensor data;
determining, by a data processing system, the intention of the type of emergency service vehicle at least in part based on a location of the type of emergency service vehicle;
generating a decision to respond to the emergency service vehicle based on the classification and the intention, wherein generating a decision includes:
generating a plurality of trajectories for an autonomous vehicle based on the emergency service vehicle intention and location, detected lane lines, and detected boundaries of the road,
generating a score for each of the plurality of trajectories, and
selecting the trajectory with the highest score; and
generating one or more commands to navigate along the selected trajectory.
22. The method of claim 21 , wherein the type of emergency service vehicle is a law enforcement vehicle.
23. The method of claim 22 , wherein determining the intention of the law enforcement vehicle includes:
detecting a state of the law enforcement vehicle; and
detecting the intention of the law enforcement vehicle based on a location of the law enforcement vehicle and the state of the law enforcement vehicle.