Collision avoidance with trajectory evaluation

1 . A system comprising:

one or more processors; and

one or more non-transitory computer-readable media storing computer-executable instructions that, when executed, cause the one or more processors to perform operations comprising:

receiving, from a first computing device, a primary trajectory and an alternative trajectory for an autonomous vehicle, wherein the primary trajectory comprises a first lateral profile and the alternative trajectory comprises a second lateral profile different from the first lateral profile, and wherein the alternative trajectory is determined based on the primary trajectory;

determining, by a second computing device, that a probability that the primary trajectory will lead to collision between the autonomous vehicle and an object within a threshold time period falls below a probability threshold, wherein the threshold time period is determined based at least in part on a time period associated with the first computing device generating a new primary trajectory and includes a buffer time period; and

based at least in part on determining that the probability falls below the probability threshold, controlling the autonomous vehicle based on the primary trajectory.

2 . The system of claim 1 , the operations further comprising:

determining that the primary trajectory will lead to collision at a first time;

determining that the first time meets or exceeds the threshold time period; and

determining that the probability falls below the probability threshold based on determining that the first time meets or exceeds the threshold time period.

3 . The system of claim 1 , the operations further comprising:

determining the probability that the primary trajectory will lead to collision between the autonomous vehicle and the object within the threshold time period based at least in part on at least one of a first speed of the autonomous vehicle, a second speed of the object, a classification of the object, or an environment of the autonomous vehicle.

4 . The system of claim 1 , wherein the second lateral profile is within a threshold lateral offset from the first lateral profile.

5 . The system of claim 1 , the operations further comprising:

determining the probability that the primary trajectory will lead to collision between the autonomous vehicle and the object within the threshold time period based at least in part on a predicted trajectory of the object.

6 . One or more non-transitory computer-readable media storing computer-executable instructions that, when executed, cause one or more processors to perform operations comprising:

receiving, from a first computing device, a first trajectory and a second trajectory, wherein the first trajectory is based on a reference trajectory and the second trajectory deviates from the reference trajectory by less than or equal to a swerve threshold, and wherein the second trajectory is determined based on the first trajectory;

determining, by a second computing device, whether the first trajectory is predicted to lead to collision with an object within a threshold time period, wherein the threshold time period is determined based at least in part on a time period associated with the first computing device generating a new trajectory; and

based at least in part on determining that the first trajectory is predicted to lead to collision, controlling a vehicle based on the second trajectory.

7 . The one or more non-transitory computer-readable media of claim 6 , wherein controlling the vehicle comprises:

based at least in part on determining that the first trajectory is predicted to lead to collision with the object within the threshold time period and that the second trajectory is associated with a least collision probability among a plurality of alternative trajectories, controlling the vehicle based on the second trajectory, wherein the plurality of alternative trajectories comprises the second trajectory and a third trajectory configured to control the vehicle to stop.

8 . The one or more non-transitory computer-readable media of claim 7 , the operations further comprising:

determining, by the second computing device, the third trajectory by determining a stopping point along the third trajectory.

9 . The one or more non-transitory computer-readable media of claim 6 , the operations further comprising:

controlling the vehicle based on the second trajectory based at least in part on determining that a probability that the first trajectory is predicted to lead to collision within the threshold time period falls below a probability threshold.

10 . The one or more non-transitory computer-readable media of claim 9 , the operations further comprising:

determining the probability based on at least one of a speed of the vehicle, an environment feature of an environment of the vehicle, a deceleration profile of the vehicle, a time-to-collision measure, a distance-to-collision measure, or a classification of the object.

11 . The one or more non-transitory computer-readable media of claim 9 , the operations further comprising:

determining the probability based on a predicted trajectory associated with the object.

12 . The one or more non-transitory computer-readable media of claim 6 , the operations further comprising:

controlling the vehicle based on the second trajectory based at least in part on determining that:

a probability that the first trajectory is predicted to lead to collision within the threshold time period falls below a probability threshold, and

that a fourth trajectory has not been received from the first computing device after receiving the first trajectory.

13 . The one or more non-transitory computer-readable media of claim 6 , wherein the second trajectory comprises a stopping profile associated with controlling the vehicle to stop.

14 . The one or more non-transitory computer-readable media of claim 6 , wherein the second computing device comprises at least one of a kinematics-based model or a dynamics-based model for evaluating whether at least one of the first trajectory or the second trajectory is predicted to intersect with a predicted trajectory of the object at a window of time.

15 . The one or more non-transitory computer-readable media of claim 6 , the operations further comprising:

determining a lane width associated with a lane of the vehicle; and

determining the swerve threshold based on the lane width.

16 . A method comprising:

receiving, from a first computing device, a first trajectory and a second trajectory, wherein the first trajectory is based on a reference trajectory and the second trajectory deviates from the reference trajectory by less than or equal to a swerve threshold, and wherein the second trajectory is determined based on the first trajectory;

determining, by a second computing device, whether the first trajectory is predicted to lead to collision with an object within a threshold time period, wherein the threshold time period is determined based at least in part on a time period associated with the first computing device generating a new trajectory; and

based at least in part on determining that the first trajectory is predicted to lead to collision, controlling a vehicle based on the second trajectory.

17 . The method of claim 16 , wherein controlling the vehicle comprises:

based at least in part on determining that the first trajectory is predicted to lead to collision and that the second trajectory is associated with a least collision probability among a plurality of alternative trajectories, controlling the vehicle based on the second trajectory, wherein the plurality of alternative trajectories comprises the second trajectory and a third trajectory configured to control the vehicle to stop.

18 . The method of claim 16 , further comprising:

controlling the vehicle based on the second trajectory based at least in part on determining that a probability that the first trajectory is predicted to lead to collision within the threshold time period falls below a probability threshold.

19 . The system of claim 1 , the operations further comprising:

determining, by the second computing device, that a second probability that the primary trajectory will lead to collision between the autonomous vehicle and the object outside of the threshold time period exceeds the probability threshold; and

controlling the autonomous vehicle based on the primary trajectory based at least in part on determining that the probability falls below the probability threshold and the second probability falls above the probability threshold.

20 . The one or more non-transitory computer-readable media of claim 6 , the operations further comprising:

determining the swerve threshold based on:

a distance between the vehicle and a lane boundary, and

a buffer distance.