Road condition deep learning model
The technology relates to using on-board sensor data, off-board information and a deep learning model to classify road wetness and/or to perform a regression analysis on road wetness based on a set of input information. Such information includes on-board and/or off-board signals obtained from one or more sources including on-board perception sensors, other on-board modules, external weather measurement, external weather services, etc. The ground truth includes measurements of water film thickness and/or ice coverage on road surfaces. The ground truth, on-board and off-board signals are used to build the model. The constructed model can be deployed in autonomous vehicles for classifying/regressing the road wetness with on-board and/or off-board signals as the input, without referring to the ground truth. The model can be applied in a variety of ways to enhance autonomous vehicle operation, for instance by altering current driving actions, modifying planned routes or trajectories, activating on-board cleaning systems, etc.
1 . A system configured to operate a vehicle in an autonomous driving mode, the system comprising:
one or more processors operatively coupled to a driving system of the vehicle and a perception system of the vehicle, the driving system being configured to perform driving operations of the vehicle, the perception system being configured to detect objects or conditions in an environment around the vehicle during operation in the autonomous driving mode, and the one or more processors being configured to:
receive sensor data from the perception system of the vehicle, the sensor data including at least one of a lidar return, imagery information, radar information or acoustical information;
apply a stored machine learning model to the received sensor data to identify a weather condition classification; and
cause the vehicle to perform a selected operation while the vehicle is operating in the autonomous driving mode, in response to the identified weather condition classification.
2 . The system of claim 1 , wherein the identified weather condition classification comprises a rain condition or water spray condition.
3 . The system of claim 1 , wherein the one or more processors are further configured to:
receive environmental information regarding the environment from a remote system; and
application of the stored machine learning model to identify the weather condition classification is further based on the received environmental information.
4 . The system of claim 3 , wherein the environmental information is received from the remote system according to information gathered by one or more other vehicles that operated in the environment.
5 . The system of claim 3 , wherein the remote system comprises at least one other vehicle that obtained the environmental information while operating in the environment.
6 . The system of claim 1 , wherein the selected operation is a current driving action to be performed by the driving system.
7 . The system of claim 1 , wherein the selected operation is modification of a planned route or a trajectory.
8 . The system of claim 1 , wherein the selected operation includes activation of an on-board cleaning system.
9 . The system of claim 8 , wherein the on-board cleaning system includes a defogger or a defroster.
10 . The system of claim 1 , wherein application of the stored machine learning model to identify the weather condition classification includes determination of whether there is a puddle along a roadway in the environment.
11 . The system of claim 1 , wherein application of the stored machine learning model to identify the weather condition classification includes determination of a probability regarding how wet or dry a portion of a lane is.
12 . The system of claim 11 , wherein the selected operation is a lane change operation based on the probability.
13 . A method for operating a vehicle in an autonomous driving mode, the method comprising:
receiving, by one or more processors of the vehicle while operating in the autonomous driving mode, sensor data from a perception system of the vehicle, the sensor data including at least one of a lidar return, imagery information, radar information or acoustical information, the perception system being configured to detect objects or conditions in an environment around the vehicle while operating in the autonomous driving mode;
applying, by the one or more processors, a stored machine learning model to the received sensor data to identify a weather condition classification; and
causing the vehicle to perform a selected operation while the vehicle is operating in the autonomous driving mode, in response to the identified weather condition classification.
14 . The method of claim 13 , further comprising:
receiving, by the one or more processors, environmental information regarding the environment from a remote system;
wherein applying the stored machine learning model to identify the weather condition classification is further based on the received environmental information.
15 . The method of claim 13 , wherein the selected operation is a current driving action to be performed by a driving system of the vehicle while operating in the autonomous driving mode.
16 . The method of claim 13 , wherein the selected operation comprises modifying a planned route or a trajectory.
17 . The method of claim 13 , wherein the selected operation comprises activating an on-board cleaning system to clean one or more sensors of the perception system.
18 . A system configured to operate a vehicle in an autonomous driving mode, the system comprising:
one or more processors operatively coupled to a driving system of the vehicle and a perception system of the vehicle, the driving system being configured to perform driving operations of the vehicle, the perception system being configured to detect objects or conditions in an environment around the vehicle during operation in the autonomous driving mode, and the one or more processors being configured to:
receive sensor data from the perception system of the vehicle, the sensor data including at least one of a lidar return, imagery information, radar information or acoustical information;
detect, from the sensor data, a weather condition in the environment around the vehicle;
determine, from the detected weather condition, an expected road condition; and
cause the vehicle to perform a selected operation while the vehicle is operating in the autonomous driving mode, in response to the expected road condition.
19 . The system of claim 18 , wherein:
the one or more processors are further configured to receive environmental information regarding the environment from a remote system; and
wherein the determination of the expected road condition is further based on the received environmental information.
20 . The system of claim 18 , wherein the selected operation is a lane change operation.