Data-driven planning for automated driving

Patent metadata

Field	Value
Publication number	US-9933781-B1
Application number	US-201615359651-A
Country	US
Kind code	B1
Filing date	Nov 23, 2016
Priority date	Nov 23, 2016
Publication date	Apr 3, 2018
Grant date	Apr 3, 2018

Abstract

A system and method are provided and include a subject vehicle having vehicle actuation systems. A driving context database stores driving trends associated with driving locations. A state estimation module determines a current location of the subject vehicle. A global route planner module determines a route to an inputted destination. An action primitive planning module retrieves a driving trend associated with the current location, selects a sequence of action primitives based on the driving trend associated with the current location and based on the determined route, and generates waypoints for the subject vehicle to travel to based on the sequence of action primitives, each waypoint including location coordinates and a direction. A trajectory planner module determines a trajectory for the subject vehicle based on the waypoints. A vehicle control module controls the vehicle actuation systems based on the determined trajectory.

First claim

What is claimed is:

1. A system comprising:
an action primitive database that stores a listing of possible actions for a subject vehicle having vehicle actuation systems including at least one of a steering system, a braking system, a throttle system, a horn system, a headlight system, and a turn signal system;
a driving context database that stores driving trends, each driving trend associated with a driving location;
a controller having a state estimation module, a global route planner module, an action primitive planning module, a trajectory planner module, and a vehicle control module,
wherein:
the state estimation module determines a current state of the subject vehicle, the current state of the vehicle including a current location of the subject vehicle;
the global route planner module determines a route to an inputted destination, the determined route including a plurality of road segments;
the action primitive planning module (i) receives the current location, (ii) retrieves a driving trend associated with the current location from the driving context database based on the current location of the subject vehicle, (iii) selects a sequence of action primitives from the action primitive database based on the driving trend associated with the current location and based on the determined route to the inputted destination, and (iv) generates a plurality of waypoints for the subject vehicle to travel to based on the sequence of action primitives, each waypoint including location coordinates and a direction;
the trajectory planner module determines a trajectory for the subject vehicle based on the plurality of waypoints; and
the vehicle control module controls at least one of the vehicle actuation systems based on the determined trajectory.

2. The system of claim 1 , further comprising:
at least one vehicle sensor, the at least one vehicle sensor including at least one of a vehicle speed sensor, a vehicle acceleration sensor, an image sensor, a Lidar sensor, a radar sensor, a stereo sensor, and an ultrasonic sensor;
wherein the state estimation module determines the current state of the subject vehicle based on data from the at least one vehicle sensor.

3. The system of claim 1 , further comprising:
at least one vehicle sensor, the at least one vehicle sensor including at least one of an image sensor, a Lidar sensor, a stereo sensor, a radar sensor, and an ultrasonic sensor;
wherein the controller further comprises an environmental estimator module that receives data from the at least one vehicle sensor and determines information about a surrounding environment of the subject vehicle; and
wherein the action primitive planning module selects the sequence of action primitives from the action primitive database additionally based on the information about the surrounding environment of the subject vehicle.

4. The system of claim 1 , wherein the vehicle actuation systems include at least one of the horn system, the headlight system, and the turn signal system and wherein the selected sequence of action primitives includes action primitives to actuate at least one of the horn system, the headlight system, and the turn signal system.

5. The system of claim 1 , wherein the driving context database is based on vehicle data and location data collected from a plurality of data-collection vehicles driven by human drivers, the vehicle data including data from at least one of a steering system, a throttle system, a braking system, a horn system, a headlight system, a turn signal system, a vehicle speed sensor, and a vehicle acceleration sensor.

6. The system of claim 5 , wherein the vehicle data and location data is communicated from the plurality of data-collection vehicles to a server and wherein the server generates the driving context database based on the vehicle data and location data communicated from the plurality of data-collection vehicles.

7. The system of claim 6 , wherein the server determines a plurality of action primitives associated with each data-collection vehicle, each action primitive corresponding to a driving action for the data-collection vehicle at an associated time interval and an associated location of the data-collection vehicle, and wherein the server generates the driving context database by determining the driving trend for the driving location based on a frequency of action primitives at that driving location.

8. A method comprising:
storing, with an action primitive database, a listing of possible actions for a subject vehicle having vehicle actuation systems including at least one of a steering system, a braking system, a throttle system, a horn system, a headlight system, and a turn signal system;
storing, with a driving context database, driving trends, each driving trend associated with a driving location;
determining, with a state estimation module of a controller, a current state of the subject vehicle, the current state including a current location of the subject vehicle;
determining, with a global route planner module of the controller, a route to an inputted destination, the determined route including a plurality of road segments;
receiving, with an action primitive planning module of the controller, the current location;
retrieving, with the action primitive planning module, a driving trend associated with the current location from the driving context database based on the current location of the subject vehicle;
selecting, with the action primitive planning module, a sequence of action primitives from the action primitive database based on the driving trend associated with the current location and based on the determined route to the inputted destination;
generating, with the action primitive planning module, a plurality of waypoints for the subject vehicle to travel to based on the sequence of action primitives, each waypoint including location coordinates and a direction;
determining, with a trajectory planner module of the controller, a trajectory for the subject vehicle based on the plurality of waypoints; and
controlling, with a vehicle control module of the controller, at least one of the vehicle actuation systems based on the determined trajectory.

9. The method of claim 8 , further comprising:
generating vehicle data with at least one vehicle sensor, the at least one vehicle sensor including at least one of a vehicle speed sensor, a vehicle acceleration sensor, an image sensor, a Lidar sensor, a stereo sensor, a radar sensor, and an ultrasonic sensor; and
receiving, with the state estimation module of the controller, the vehicle data;
wherein the state estimation module determines the current state of the subject vehicle based on data from the at least one vehicle sensor.

10. The method of claim 8 , further comprising:
generating vehicle data with at least one vehicle sensor, the at least one vehicle sensor including at least one of an image sensor, a Lidar sensor, a stereo sensor, a radar sensor, and an ultrasonic sensor;
receiving, with an environmental estimator module of the controller, the vehicle data; and
determining, with the environmental estimator module of the controller, information about a surrounding environment of the subject vehicle;
wherein the environmental estimator module of the controller selects the sequence of action primitives from the action primitive database additionally based on the information about the surrounding environment of the subject vehicle.

11. The method of claim 8 , wherein the vehicle actuation systems include at least one of the horn system, the headlight system, and the turn signal system and wherein the selected sequence of action primitives includes action primitives to actuate a

Assignees

• Denso Int America Inc
• Denso Corp
• RobotWits LLC
• RobotWits LLC

Inventors

• Bando Takashi
• Xu Yunfei
• Likhachev Maxim
• Ratner Ellis Marshal

Classifications

• B60W60/001

  Planning or execution of driving tasks · CPC title

• B60W2556/50

  of positioning data, e.g. GPS [Global Positioning System] data · CPC title

• B60W2520/105

  Longitudinal acceleration · CPC title

• B60W2710/18

  Braking system · CPC title

• B60W30/16

  Control of distance between vehicles, e.g. keeping a distance to preceding vehicle · CPC title