Autonomous vehicle interaction and profile sharing
Techniques for determining behavior profiles and unique identities for objects observed in an environment surrounding an autonomous vehicle are described. The behavior profiles may be generated and associated with unique identifiers that are stored and used by a fleet of vehicles for planning and navigating in traffic environments. The behavior profiles may be used to control operation of the vehicle and increase safety of the vehicle by identifying potentially risky or erratic objects and navigating through the environment accordingly. The behavior profiles may be stored at a central database and propagated across a fleet to increase observations used to build profiles and thereby increase profile confidence.
1. A system comprising:
one or more processors; and
non-transitory computer readable media storing instructions that when executed by the one or more processors, cause an autonomous vehicle system to perform operations comprising:
receiving first sensor data associated with an environment of a first vehicle, the first sensor data indicative of a presence of an instant object in proximity to the first vehicle and including a characteristic of the instant object;
determining a behavior profile of the instant object based at least in part on the first sensor data, the behavior profile indicative of a predicted response by the instant object for a condition;
determining an identifier that is unique to the instant object based at least in part on the first sensor data; and
storing the behavior profile and the identifier in a database, the database accessible by a second vehicle.
2. The system of claim 1 , wherein:
the first sensor data includes location data; and
the behavior profile is stored in the database based at least in part on the location data.
3. The system of claim 2 , further comprising conveying to the second vehicle, the behavior profile and the identifier in response to determining that the second vehicle is within a threshold distance of a geographic location indicated by the location data.
4. The system of claim 1 , wherein the first sensor data is associated with a first time, and wherein the operations further comprise:
receiving second sensor data associated with a second time after the first time, the second sensor data indicative of a presence of a second object in the environment at the second time and including the characteristic of the instant object;
determining the identifier of the instant object based at least in part on the first sensor data;
accessing a stored behavior profile in the database based at least in part on the identifier; and
modifying the behavior profile for the instant object based at least in part on the second sensor data.
5. The system of claim 1 , wherein storing the behavior profile comprises:
generating a hash based at least in part on the identifier and a geographic location of the instant object; and
storing the behavior profile in the database using the hash as a profile identifier.
6. The system of claim 1 , further comprising receiving second sensor data associated with the environment, the second sensor data indicative of a presence of a second object operating in the environment and including a second characteristic of the second object;
determining a second behavior profile of the second object based at least in part on the second sensor data;
determining a second identifier that is unique to the second object based at least in part on the second sensor data; and
storing the second behavior profile and the second identifier in the database.
7. A method, comprising:
determining a behavior profile corresponding to an instant object within an environment in which a fleet of vehicles operate based at least in part on sensor data associated with the environment;
determining an identifier that is unique to the instant object derived from the sensor data associated with the environment;
storing the behavior profile with the identifier; and
conveying the behavior profile to at least one vehicle of the fleet of vehicles, wherein the behavior profile is accessible via the identifier.
8. The method of claim 7 , wherein the behavior profile is associated with a location of the instant object and conveying the behavior profile to the fleet of vehicles comprising conveying the behavior profile to a subset of the fleet of vehicles within a predetermined distance of the location.
9. The method of claim 8 , wherein the identifier comprises a hash value that includes the location of the instant object.
10. The method of claim 7 , wherein conveying the behavior profile comprises providing the behavior profile in response to receiving a request for the behavior profile identified by the identifier from a vehicle of the fleet of vehicles.
11. The method of claim 10 , wherein the request includes a location of the vehicle, and the behavior profile is determined based at least in part on the location.
12. The method of claim 7 , further comprising determining a likelihood for the instant object to diverge from an expected behavior, wherein conveying the behavior profile to the fleet is based at least in part on the likelihood for the instant object to diverge from the expected behavior meeting or exceeding a threshold value.
13. The method of claim 7 , wherein determining the behavior profile comprises:
accessing a database and selecting the behavior profile for the instant object based on the identifier; and
updating the behavior profile based at least in part on the sensor data.
14. The method of claim 7 , wherein the behavior profile includes a confidence score, the confidence score indicative of a confidence of the behavior profile to predict behavior of the instant object, wherein the confidence score is based at least in part on a temporal proximity of the behavior profile to a present time and additional observations of the instant object.
15. A non transitory computer readable medium storing instructions executable by a processor, wherein the instructions, when executed, cause the processor to perform operations comprising:
determining a behavior profile corresponding to an instant object within an environment in which a fleet of vehicles operate based at least in part on sensor data associated with the environment;
determining an identifier that is unique to the instant object derived from the sensor data associated with the environment;
storing the behavior profile with the identifier; and
conveying the behavior profile to at least one vehicle of the fleet of vehicles, wherein the behavior profile is accessible via the identifier.
16. The non transitory computer readable medium of claim 15 , wherein the identifier comprises a hash value that includes a characteristic of the instant object.
17. The non transitory computer readable medium of claim 15 , wherein conveying the behavior profile to the fleet of vehicles comprises communicating directly with one or more vehicles of the fleet of vehicles based on the one or more vehicles being within a threshold distance of the instant object.
18. The non transitory computer readable medium of claim 15 , wherein determining the behavior profile comprises:
accessing a database and selecting the behavior profile for the instant object based on the identifier; and
updating the behavior profile based at least in part on the sensor data.
19. The non transitory computer readable medium of claim 15 , wherein the operations further comprise determining a likelihood for the instant object to diverge from an expected behavior, wherein conveying the behavior profile to the fleet is based at least in part on the likelihood for the instant object to diverge from the expected behavior meeting or exceeding a threshold value.
20. The non transitory computer readable medium of claim 15 , wherein conveying the behavior profile comprises updating a cloud-based database of behavior profiles accessible by the fleet of vehicles.