Dual mode map for autonomous vehicle
An autonomous vehicle control system and method may utilize a dual mode map including sparse map data for some portions of an environment that lacks some of the data maintained in dense map data for other portions of the environment. Sparse map data may be used, for instance, to address a recently-established construction area on a roadway that is incompatible with dense map data that was previously used to operate on the roadway, enabling operation of an autonomous vehicle in the construction area to proceed even in the absence of dense map data for the construction area, e.g., by dynamically augmenting the sparse map data to incorporate additional data sensed by a perception system of the autonomous vehicle.
1 . An autonomous vehicle control system for an autonomous vehicle, comprising:
one or more processors; and
memory storing instructions that, when executed by the one or more processors, cause the autonomous vehicle control system to:
store dense map data describing a ground surface of a first portion of an environment within which the autonomous vehicle operates, the dense map data further including semantic content for the first portion of the environment;
store sparse map data describing a ground surface of a second portion of the environment within which the autonomous vehicle operates, wherein the first and second portions are different from one another, and wherein the sparse map data describing the ground surface of the second portion of the environment is generated remote from the autonomous vehicle based on log data collected from one or more other autonomous vehicles that are operated within the second portion of the environment; and
in response to determining that the autonomous vehicle is located in the second portion of the environment:
receive perception data describing at least one perceived boundary for a roadway in the environment that is sensed by at least one perception sensor of the autonomous vehicle during operation of the autonomous vehicle on the roadway;
augment the sparse map data to generate augmented sparse map data, wherein the augmented sparse map data describes a pathway for use in operating the autonomous vehicle in a perceived lane defined by the at least one perceived boundary; and
control the autonomous vehicle using the augmented sparse map data.
2 . The autonomous vehicle control system of claim 1 , wherein the one or more processors are further configured to, in response to determining that the autonomous vehicle is located in the first portion of the environment, control the autonomous vehicle using the dense map data and the semantic content thereof for the first portion of the environment.
3 . The autonomous vehicle control system of claim 1 , wherein the sparse map data lacks semantic content for the second portion of the environment.
4 . The autonomous vehicle control system of claim 1 , wherein the dense map data describing the ground surface of the first portion of the environment includes one or more outer road boundaries for one or more roadways in the first portion of the environment and the sparse map data describing the ground surface of the second portion of the environment includes one or more outer road boundaries for one or more roadways in the second portion of the environment.
5 . The autonomous vehicle control system of claim 1 , wherein the dense map data describing the ground surface of the first portion of the environment includes localization data for pose determination within the first portion of the environment and the sparse map data describing the ground surface of the second portion of the environment includes localization data for pose determination in the second portion of the environment.
6 . The autonomous vehicle control system of claim 1 , wherein the semantic content in the dense map data includes one or more mapped boundaries, one or more mapped lanes, one or more speed limits, one or more mapped signs and/or one or more traffic signals.
7 . The autonomous vehicle control system of claim 1 , wherein the sparse map data is generated in response to detection of one or more modified road boundaries in the second portion of the environment, and wherein the sparse map data describing the ground surface of the second portion of the environment includes localization data for pose determination in the second portion of the environment that is reused from dense map data describing the ground surface of the second portion of the environment.
8 . The autonomous vehicle control system of claim 1 , wherein the sparse map data is generated in response to a global pose fault in the second portion of the environment, and wherein the sparse map data describing the ground surface of the second portion of the environment includes localization data for pose determination in the second portion of the environment.
9 . The autonomous vehicle control system of claim 1 , wherein the one or more processors are configured to disable lane alignment localization when controlling the autonomous vehicle using the augmented sparse map data.
10 . The autonomous vehicle control system of claim 1 , wherein the one or more processors are configured to enable a perceived boundary control mode in response to determining that the autonomous vehicle is located in the second portion of the environment, and to disable the perceived boundary control mode in response to determining that the autonomous vehicle is located in the first portion of the environment.
11 . The autonomous vehicle control system of claim 1 , wherein the one or more processors are further configured to:
store dense map data describing a ground surface of the second portion of the environment and including semantic content for the second portion of the environment;
control the autonomous vehicle using the dense map data and the semantic content thereof for the second portion of the environment;
after storing the dense map data for the second portion of the environment and controlling the autonomous vehicle using the dense map data and the semantic content thereof for the second portion of the environment, receive the sparse map data; and
after receiving the sparse map data, invalidate the dense map data for the second portion of the environment.
12 . The autonomous vehicle control system of claim 11 , wherein the sparse map data is received wirelessly via an over-the-air map update.
13 . The autonomous vehicle control system of claim 12 , wherein the sparse map data is received subsequent to detection of an event in the second portion of the environment.
14 . The autonomous vehicle control system of claim 13 , wherein the event is initiation of road construction within the second portion of the environment.
15 . The autonomous vehicle control system of claim 14 , wherein the road construction requires vehicle traffic to drive outside of a mapped outer road boundary defined in the dense map data describing the ground surface of the second portion of the environment.
16 . The autonomous vehicle control system of claim 14 , wherein the road construction requires vehicle traffic to drive within a mapped oncoming lane defined in the dense map data describing the ground surface of the second portion of the environment.
17 . The autonomous vehicle control system of claim 13 , wherein the event is detected by another autonomous vehicle, and wherein the sparse map data is generated using one or more logs collected by the other autonomous vehicle and/or one or more additional vehicles.
18 . The autonomous vehicle control system of claim 13 , wherein the sparse map data is received as a temporary lightweight map update, and wherein the one or more processors are further configured to, after receiving the sparse map data:
receive updated dense map data describing the ground surface of the second portion of the environment as a map release update, the updated dense map data further including updated semantic content for the second portion of the environment; and
in response to receiving the updated dense map data, store the updated dense map data, invalidate the sparse map data for the second portion of the environment, and control the autonomous vehicle using the updated dense map data and the semantic content thereof for the second portion of the environment.
19 . A method of operating an autonomous vehicle, comprising:
storing dense map data describing a ground surface of a first portion of an environment within which the autonomous vehicle operates, the dense map data further including semantic content for the first portion of the environment;
storing sparse map data describing a ground surface of a second portion of the environment within which the autonomous vehicle operates, wherein the first and second portions are different from one another, and wherein the sparse map data describing the ground surface of the second portion of the environment is generated remote from the autonomous vehicle based on log data collected from one or more other autonomous vehicles that are operated within the second portion of the environment; and
in response to determining that the autonomous vehicle is located in the second portion of the environment:
receiving perception data describing at least one perceived boundary for a roadway in the environment that is sensed by at least one perception sensor of the autonomous vehicle during operation of the autonomous vehicle on the roadway;
augmenting the sparse map data to generate augmented sparse map data, wherein the augmented sparse map data describes a pathway for use in operating the autonomous vehicle in a perceived lane defined by the at least one perceived boundary; and
controlling the autonomous vehicle using the augmented sparse map data.
20 . A non-transitory computer readable storage medium storing computer instructions executable by one or more processors to perform a method of operating an autonomous vehicle, the method comprising:
storing dense map data describing a ground surface of a first portion of an environment within which the autonomous vehicle operates, the dense map data further including semantic content for the first portion of the environment;
storing sparse map data describing a ground surface of a second portion of the environment within which the autonomous vehicle operates, wherein the first and second portions are different from one another, and wherein the sparse map data describing the ground surface of the second portion of the environment is generated remote from the autonomous vehicle based on log data collected from one or more other autonomous vehicles that are operated within the second portion of the environment; and
in response to determining that the autonomous vehicle is located in the second portion of the environment:
receiving perception data describing at least one perceived boundary for a roadway in the environment that is sensed by at least one perception sensor of the autonomous vehicle during operation of the autonomous vehicle on the roadway;
augmenting the sparse map data to generate augmented sparse map data, wherein the augmented sparse map data describes a pathway for use in operating the autonomous vehicle in a perceived lane defined by the at least one perceived boundary; and
controlling the autonomous vehicle using the augmented sparse map data.