EP-4742216-A2 - COLLISION PREDICTION AND AVOIDANCE FOR VEHICLES

EP4742216A2EP 4742216 A2EP4742216 A2EP 4742216A2EP-4742216-A2

Abstract

A vehicle computing system may implement techniques to control a vehicle to avoid collisions between the vehicle and agents (e.g., dynamic objects) in an environment. The techniques may include generating a representation of a path of the vehicle through an environment as a polygon. The vehicle computing system may compare the two-dimensional path with a trajectory of an agent determined using sensor data to determine a collision zone between the vehicle and the agent. The vehicle computing system may determine a risk of collision based on predicted velocities and probable accelerations of the vehicle and the agent approaching and traveling through the collision zone. Based at least in part on the risk of collision, the vehicle computing system may cause the vehicle to perform an action.

Inventors

PACKER, JEFFERSON BRADFIELD
SILVA, WILLIAM ANTHONY
HUANG, Zhenqi

Assignees

Zoox, Inc.

Dates

Publication Date: 20260513
Application Date: 20190918

Claims (15)

A method comprising: determining a first probability density function associated with a vehicle in an environment; determining a second probability density function associated with an object in the environment; determining a likelihood of collision in a potential collision zone between the vehicle and the object based at least in part on an overlap between the first probability density function and the potential collision zone and the second probability density function and the potential collision zone; determining an action for the vehicle to perform; and causing the vehicle to perform the action.
The method of claim 1, further comprising determining that the likelihood of collision is above a threshold value; and wherein the action comprises modifying a vehicle trajectory based at least in part on the likelihood of collision being above the threshold value.
The method of any one of the preceding claims, wherein the potential collision zone is based at least in part on a first parameter of the vehicle and a second parameter of the object, and the likelihood of collision is based at least in part on a size of the potential collision zone.
The method of any one of the preceding claims, wherein the overlap is further associated with: position cones associated with the vehicle and the object.
The method of any one of the preceding claims, wherein the action for the vehicle to perform is based at least in part on the likelihood of collision.
The method of any one of the preceding claims, wherein the action comprises at least one of: maintaining a first trajectory associated with the vehicle; modifying a speed associated with the first trajectory; or modifying a direction of travel associated with the first trajectory.
The method of any one of the preceding claims, wherein determining the potential collision zone comprises: determining a vehicle polygon based at least in part on a first trajectory associated with the vehicle, wherein the vehicle polygon is representative of a planned path of the vehicle; and determining an object polygon based at least in part on a second trajectory associated with the object, wherein the object polygon is representative of a predicted path of the object in the environment, wherein the potential collision zone comprises an intersection of the vehicle polygon and the object polygon.
The method of any one of the preceding claims, further comprising determining an object entry time into the potential collision zone; determining an object exit time from the potential collision zone; determining a vehicle entry time into the potential collision zone; and determining a vehicle exit time from the potential collision zone, wherein the period of time comprises a range of times during which the vehicle and the object are predicted to be located in the potential collision zone based at least in part on the object entry time, the object exit time, the vehicle entry time and the vehicle exit time.
The method according to any one of the preceding claims, wherein determining the action for the vehicle to perform comprises: generating a second polygon of the vehicle based on the action; determining a second potential collision zone between the vehicle and the object based on the second polygon; and determining a second likelihood of collision, wherein causing the vehicle to perform the action is based at least in part on the second likelihood of collision.
The method according to any one of the preceding claims, wherein the first probability density function is associated with probable positions of the vehicle in the environment over time; wherein the second probability density function is associated with probable positions of the object in the environment over time;
The method of claim 3, wherein the first parameter of the vehicle is a vehicle enter point and a vehicle exit point; wherein the second parameter of the object is an object enter point and an object exit point.
The method according to any one of the preceding claims, further comprising determining a first trajectory associated with the vehicle traveling in the environment; determining a second trajectory associated with the object traveling in the environment; determining that the likelihood of collision between the vehicle and the object is between a first threshold value and a second threshold value; determining, based at least in part on the likelihood of collision being between the first threshold value and the second threshold value, a third trajectory for the vehicle, the third trajectory comprising a first deceleration; and causing the vehicle to perform the action based at least in part on the third trajectory.
The method of any one of the preceding claims, wherein the likelihood of collision is a first probability of collision determined at a first time, the method further comprising: determining a second probability of collision at a second time after the first time; determining that the second probability of collision meets or exceeds the first threshold value and the second threshold value; and determining a fourth trajectory for the vehicle, the fourth trajectory comprising a second deceleration with a greater magnitude than the first deceleration.
A system comprising: one or more processors; and one or more computer-readable media storing instructions executable by the one or more processors, wherein the instructions, when executed, cause the system to perform the method of any one of the preceding claims.
A computer-readable medium storing instructions that, when executed, cause one or more processors to perform the method of any one of claims 1-13.

Description

CROSS REFERENCE TO RELATED APPLICATIONS This application is a divisional application of EP19780094.9. BACKGROUND Vehicles may be equipped with collision avoidance systems configured to detect and avoid objects in an operating environment. The objects may include mobile objects, such as other vehicles, cyclists, pedestrians, etc. Traditional collision avoidance systems may avoid collisions by simply identifying the presence of surfaces in an environment and adjusting a velocity of a vehicle to avoid collision with a surface. However, these traditional systems may cause the vehicle to yield in situations in which it is unnecessary and unnatural, thereby potentially causing traffic delays. BRIEF DESCRIPTION OF THE DRAWINGS The detailed description is described with reference to the accompanying figures. In the figures, the left-most digit(s) of a reference number identifies the figure in which the reference number first appears. The use of the same reference numbers in different figures indicates similar or identical components or features. FIG. 1 is an illustration of an autonomous vehicle in an environment, wherein a path polygon of the autonomous vehicle and estimated agent trajectories are overlaid in the illustration representing a two-dimensional map of the environment generated by a collision avoidance system of the autonomous vehicle to determine whether a potential collision zone exists between the autonomous vehicle and an agent, in accordance with embodiments of the disclosure.FIG. 2 is an illustration of an autonomous vehicle in an environment, in which a collision avoidance system of the autonomous vehicle may determine a potential collision zone between a path polygon representing a planned path of the autonomous vehicle and an estimated agent trajectory associated with an agent.FIGS. 3A and 3B are illustrations of example time-space overlaps of position cones associated with agent trajectories and planned speeds of an autonomous vehicle relative to a potential collision zone, in which a collision avoidance system of the autonomous vehicle may determine a possible collision between the autonomous vehicle and an agent based on the time-space overlap, in accordance with embodiments of the disclosure. FIG. 3A is an illustration of a time-space overlap in which the collision avoidance system may determine a high risk of collision between the vehicle and the agent based on a position cone associated with an agent overlapping with the autonomous vehicle along the planned path. FIG. 3B is an illustration of a time-space overlap in which the collision avoidance system may determine a low risk of collision between the vehicle and the agent based on a position cone associated with the agent not overlapping with the autonomous vehicle along the planned path.FIG. 4 is an illustration of a collision zone between a path polygon of an autonomous vehicle and an estimated agent trajectory, in which a collision avoidance system of the autonomous vehicle may determine a possible collision point between the autonomous vehicle and an agent based on one or more probability density functions of probable speeds and actions associated with the agent, in accordance with embodiments of the disclosure.FIG. 5 is a block diagram of an example system for implementing the techniques described herein.FIG. 6 depicts an example process for determining an action to perform to avoid a collision between an autonomous vehicle and an object in an environment, in accordance with embodiments of the disclosure.FIG. 7 depicts an example process for determining an action to perform to avoid a collision between a vehicle and an object in an environment, in accordance with embodiments of the disclosure.FIG. 8 depicts an example process for determining a collision zone between an autonomous vehicle and an agent in an environment, in accordance with embodiments of the disclosure. DETAILED DESCRIPTION This disclosure is directed to techniques for improving collision prediction and avoidance between a vehicle and agents (e.g., dynamic objects) in an environment. The vehicle may include an autonomous or semi-autonomous vehicle. The agents may include other vehicles (e.g., cars, trucks, motorcycles, mopeds, etc.), pedestrians, bicyclists, or the like. A vehicle computing system may be configured to determine possible collision zones between the vehicle and the agents based on probable paths and velocities associated therewith. The vehicle computing system may then be configured to determine an action to perform to avoid a collision in the one or more possible collision zones. The vehicle computing system may be configured to generate a point path polygon (path polygon) representing a two-dimensional path of the vehicle (e.g., vehicle path) through the environment. The path polygon may include a plurality of point pairs (or simply points) along a planned path of the vehicle. In various examples, the points may include a representation of a left and