Automatic navigation based on traffic management vehicles and road signs
Methods, systems, and non-transitory computer readable media are configured to perform operations comprising determining an occurrence of a sign in an environment of a vehicle; determining that at least a portion of the sign is unrecognized by a local machine learning model of the vehicle; and providing sensor data associated with the at least one portion of the sign to an operations center remote from the vehicle.
1. A computer-implemented method comprising:
determining, by a computing system, an occurrence of a sign in an environment of a vehicle;
determining, by the computing system, that a first portion of the sign including a first type of content is unrecognized and a second portion of the sign including a second type of content is recognized by a local machine learning model of the vehicle;
providing, by the computing system, sensor data associated with the first portion of the sign to an operations center remote from the vehicle; and
causing, by the computing system, the vehicle to perform an action based on navigation guidance provided by the operations center.
2. The computer-implemented method of claim 1 , wherein the portion of the sign is a section of a plurality of sections of the sign constituting the entirety of the sign, each section indicating an associated type of information.
3. The computer-implemented method of claim 1 , wherein the determining that a first portion of the sign is unrecognized comprises:
extracting from the sign a first sign element that is recognized by the local machine learning model; and
extracting from the sign a second sign element that is unrecognized by the local machine learning model.
4. The computer-implemented method of claim 1 , further comprising:
providing to the operations center a request for navigation guidance to plan motion for the vehicle in response to the first portion of the sign and associated scenario context data; and
slowing down or stopping motion of the vehicle behind the sign pending receipt of the navigation guidance.
5. The computer-implemented method of claim 4 , further comprising:
Receiving from the operations center the navigation guidance including an action to be performed by the vehicle in response to the first portion of the sign.
6. The computer-implemented method of claim 5 , further comprising:
performing the action;
generating assessment data including safety considerations of the performed action; and
providing the assessment data to the operations center.
7. The computer-implemented method of claim 6 , wherein a central machine learning model configured to recognize variable signage and associated with the operations center is retrained based on the assessment data generated by the vehicle and other assessment data generated by other vehicles in a fleet.
8. The computer-implemented method of claim 7 , wherein the retrained central machine learning model is periodically distributed to the vehicle and the other vehicles in the fleet to update local machine learning models maintained by the vehicle and the other vehicles.
9. The computer-implemented method of claim 1 , wherein the sign includes a plurality of sign elements, the first portion of the sign includes a first sign element that is unrecognized by the vehicle, and the second portion of the sign includes a second sign element that is recognized by the vehicle, the method further comprising:
performing an action responsive to the second sign element when the first sign element is unrecognized if performance of the action does not violate a safety rule.
10. The computer-implemented method of claim 1 , wherein the sign includes a plurality of sign elements, the first portion of the sign includes a first sign element that is unrecognized by the vehicle, and second portion of the sign includes a second sign element that is recognized by the vehicle, the method further comprising:
not performing an action responsive to the second sign element until navigation guidance from the operations center is provided for the first sign element.
11. A system comprising:
at least one processor; and
a memory storing instructions that, when executed by the at least one processor, cause the system to perform operations comprising:
determining an occurrence of a sign in an environment of a vehicle;
determining that a first portion of the sign including a first type of content is unrecognized and a second portion of the sign including a second type of content is recognized by a local machine learning model of the vehicle;
providing sensor data associated with the first portion of the sign to an operations center remote from the vehicle; and
causing the vehicle to perform an action based on navigation guidance provided by the operations center.
12. The system of claim 11 , wherein the first portion of the sign is a section of a plurality of sections of the sign constituting the entirety of the sign, each section indicating an associated type of information.
13. The system of claim 11 , wherein the determining that a first portion of the sign is unrecognized comprises:
extracting from the sign a first sign element that is recognized by the local machine learning model; and
extracting from the sign a second sign element that is unrecognized by the local machine learning model.
14. The system of claim 11 , the operations further comprising:
providing to the operations center a request for navigation guidance to plan motion for the vehicle in response to the first portion of the sign and associated scenario context data; and
slowing down or stopping motion of the vehicle behind the sign pending receipt of the navigation guidance.
15. The system of claim 14 , the operations further comprising:
receiving from the operations center the navigation guidance including an action to be performed by the vehicle in response to the first portion of the sign.
16. A non-transitory computer-readable storage medium including instructions that, when executed by at least one processor of a computing system, cause the computing system to perform operations comprising:
determining an occurrence of a sign in an environment of a vehicle;
determining that a first portion of the sign including a first type of content is unrecognized and a second portion of the sign including a second type of content is recognized by a local machine learning model of the vehicle;
providing sensor data associated with the portion of the sign to an operations center remote from the vehicle; and
causing the vehicle to perform an action based on navigation guidance provided by the operations center.
17. The non-transitory computer-readable storage medium of claim 16 , wherein the first portion of the sign is a section of a plurality of sections of the sign constituting the entirety of the sign, each section indicating an associated type of information.
18. The non-transitory computer-readable storage medium of claim 16 , wherein the determining that a first portion of the sign is unrecognized comprises:
extracting from the sign a first sign element that is recognized by the local machine learning model; and
extracting from the sign a second sign element that is unrecognized by the local machine learning model.
19. The non-transitory computer-readable storage medium of claim 16 , the operations further comprising:
providing to the operations center a request for navigation guidance to plan motion for the vehicle in response to the first portion of the sign and associated scenario context data; and
slowing down or stopping motion of the vehicle behind the sign pending receipt of the navigation guidance.
20. The non-transitory computer-readable storage medium of claim 19 , the operations further comprising:
receiving from the operations center the navigation guidance including an action to be performed by the vehicle in response to the first portion of the sign.