Navigation of autonomous vehicles to enhance safety under one or more fault conditions
Various embodiments relate generally to autonomous vehicles and associated mechanical, electrical and electronic hardware, computing software, including autonomy applications, image processing applications, etc., and computing systems, and wired and wireless network communications to facilitate autonomous control of vehicles, and, more specifically, to systems, devices, and methods configured to navigate autonomous vehicles under one or more fault conditions. In some examples, a method may include localizing an autonomous vehicle, accessing map data to identify safe zones, computing drive parameters and alternate drive parameters, detecting an anomalous event, and apply alternate drive parameters to a vehicle control unit.
1. A method comprising:
determining, by a processor, a position of an autonomous vehicle relative to a roadway over which the autonomous vehicle is configured to transit via a path of travel, wherein the path of travel comprises predetermined travel points corresponding to discrete geographic locations disposed along the path of travel;
accessing, by the processor, map data to identify locations of one or more safe zones associated with each of the predetermined travel points along the path of travel;
determining, by the processor, a set of driving instructions for a vehicle controller to autonomously control the autonomous vehicle along the path of travel;
determining, by the processor prior to detecting any anomalous events associated with the autonomous vehicle or an environment external to the autonomous vehicle, one or more subsets of alternate driving instructions for the vehicle controller to autonomously control the autonomous vehicle to a safe zone of the one or more safe zones via a recovery path of travel from at least one of the predetermined travel points;
controlling, autonomously by the vehicle controller, the autonomous vehicle along the path of travel based on the set of driving instructions;
detecting, by the processor, an anomalous event associated with the autonomous vehicle or the environment external to the autonomous vehicle; and
controlling, autonomously by the vehicle controller and in response to detecting the anomalous event, the autonomous vehicle based on a subset of the one or more subsets of alternate driving instructions to guide the autonomous vehicle to the safe zone of the one or more safe zones via the recovery path of travel.
2. The method of claim 1 further comprising:
receiving, by the processor, images of the locations of the one or more safe zones captured via one or more cameras; and
analyzing, by the processor, the images and the map data to confirm the locations of the one or more safe zones.
3. The method of claim 2 further comprising:
analyzing, by the processor, the images and the map data for multiple travel points of the predetermined travel points along the path of travel;
identifying, by the processor, a candidate safe zone of the one or more safe zones based on the map data;
determining, by the processor, data representing an occlusion associated with the candidate safe zone; and
excluding, by the processor, the candidate safe zone from being included in the one or more safe zones prior to detecting the anomalous event.
4. The method of claim 2 wherein analyzing the images and the map data comprises:
receiving, by the processor, images of the locations as high definition (“HD”) images; and
analyzing, by the processor, the HD images and HD map data, which together constitutes the map data.
5. The method of claim 1 further comprising:
selecting, by the processor, a location for one of the one or more safe zones prior to reaching the at least one of the predetermined travel points.
6. The method of claim 5 wherein selecting the location comprises:
identifying, by the processor, a reference point representing a location of the autonomous vehicle; and
predicting, by the processor, a subset of actions to guide the autonomous vehicle autonomously to the safe zone, at least one action including the subset of the one or more subsets of alternate driving instructions.
7. The method of claim 6 wherein predicting the subset of actions comprises:
determining, by the processor, for the at least one action the subset of the one or more alternate driving instruction to navigate the autonomous vehicle over a portion of the recovery path of travel that excludes an object detected by the autonomous vehicle on the roadway.
8. The method of claim 1 further comprising:
classifying, by the processor, the detected anomalous event based on a state of operation of a sensor;
identifying, by the processor, one or more other sensors responsive to the state of operation of the sensor; and
determining, by the processor, an event-specific subset of actions utilizing the one or more other sensors based on the detected anomalous event.
9. The method of claim 8 further comprising:
controlling, by the vehicle controller, the autonomous vehicle based on the event-specific subset of actions to navigate the autonomous vehicle to the safe zone via the recovery path of travel.
10. The method of claim 1 further comprising:
identifying, by the processor, data representing a glide path associated with the safe zone, the data representing the glide path including waypoints each of which is associated with executable instructions configured to receive sensor data and to guide the autonomous vehicle from a waypoint to the safe zone.
11. The method of claim 1 further comprising:
receiving, by the processor, a portion of the map data as updated map data including the executable instructions of at least one waypoint of the glide path; and
controlling, autonomously by the vehicle controller, the autonomous vehicle based on the executable instructions to navigate the autonomous vehicle via the glide path.
12. The method of claim 1 further comprising:
receiving, by the processor, images of the locations of the one or more safe zones captured via one or more cameras;
analyzing, by the processor, the images and the map data at multiple points along the path of travel;
identifying, by the processor, a candidate safe zone based on the map data;
determining, by the processor, data representing an obstacle associated with the candidate safe zone;
determining, by the processor, that the obstacle is an animal;
excluding, by the processor, the candidate safe zone from being included in the one or more safe zones prior to detecting the anomalous event.
13. The method of claim 3 further comprising:
transmitting, via a communication network, the data representing the occlusion to a computing device implemented as a portion of a vehicular autonomy platform to cause an update to a standard map.
14. The method of claim 1 further comprising:
detecting, by the processor, a manual vehicular drive control after detecting the anomalous event; and
preventing, by the processor and based on the detected manual vehicular drive control, control of the autonomous vehicle by the vehicle controller based on the subset of the one or more subsets of alternate driving instructions.
15. The method of claim 14 , wherein detecting the manual vehicular drive control further comprises:
identifying, by the processor, a state in which a driver is intervening with the autonomous control of the autonomous vehicle while traveling via the recovery path of travel.
16. The method of claim 1 further comprising:
receiving, by the processor, images of regions associated with the path of travel captured via one or more cameras;
analyzing, by the processor, the images of at least one of the regions;
detecting, by the processor, one or more objects in the at least one region to form an obstructed region; and
identifying, by the processor, the safe zone as another region of the regions in which objects are absent.
17. A controller for an autonomous vehicle, comprising:
a memory including executable instructions; and
a processor, responsive to executing the instructions, is programmed to:
determine a position of the autonomous vehicle relative to a roadway to transit via a path of travel, wherein the path of travel comprises predetermined travel points corresponding to discrete geographic locations disposed along the path of travel;
access map data to identify locations of one or more safe zones associated with each of the predetermined travel points along the path of travel;
determine driving instructions for autonomously controlling the autonomous vehicle along the path of travel;
determine, prior to detecting any anomalous events associated with the autonomous vehicle or an environment external to the autonomous vehicle, one or more subsets of alternate driving instructions for the autonomous vehicle to autonomously navigate to a safe zone of the one or more safe zones via a recovery path of travel from at least one of the predetermined travel points;
control the autonomous vehicle along the path of travel based on the driving instructions;
detect an anomalous event associated with the autonomous vehicle or the environment external to the autonomous vehicle; and
control autonomously by the vehicle controller and in response to detecting the anomalous event, the autonomous vehicle based on a subset of the one or more subsets of alternate driving instructions to guide the autonomous vehicle to the safe zone of the one or more safe zones via the recovery path of travel.
18. The apparatus of claim 17 , wherein the processor is further configured to:
receive images of the locations of the one or more safe zones captured via one or more cameras; and
analyze the images and the map data to confirm the locations of the one or more safe zones.
19. The apparatus of claim 18 , wherein the processor is further configured to:
analyze the images and the map data at multiple points along the path of travel;
identify a candidate safe zone based on the map data;
determine data representing an occlusion associated with the candidate safe zone; and
exclude the candidate safe zone from being included in the one or more safe zones prior to detecting the anomalous event.
20. The apparatus of claim 17 , wherein the processor is further configured to:
classify the detected anomalous event based on a state of operation of a sensor;
identify one or more other sensors responsive to the state of operation of the sensor;
determine an event-specific subset of actions implementing the one or more other sensors based on the detected anomalous event; and
control the autonomous vehicle based on the event-specific subset of actions to navigate the autonomous vehicle to the safe zone via the recovery path of travel.