Path planning using delta cost volume generated from movement restrictions and observed driving behavior
AssigneeWoven Planet North America, Inc.
InventorsTsung-Han Lin, Sammy Omari, Peter Ondruska, Matthew Swaner Vitelli
Life of this patent
Estimating the term…
Abstract
In one embodiment, a method includes determining an initial cost volume associated with a plurality of potential trajectories of a vehicle in an environment based on a set of movement restrictions of the vehicle, generating a delta cost volume using the initial cost volume and environment data associated with the environment, wherein the delta cost volume is generated by determining adjustments to the initial cost volume that incorporate observed driving behavior, and scoring a trajectory of the plurality of potential trajectories for the vehicle based on the initial cost volume and the delta cost volume.
Claims20
A method comprising:
- determining an initial cost volume associated with potential trajectories of a vehicle in an environment based on a set of movement restrictions of the vehicle, wherein the initial cost volume is based upon hand-engineered costs derived from the set of movement restrictions for the vehicle;
- generating, using a machine learning model, a delta cost volume using the initial cost volume and environment data of the environment, wherein the machine learning model is trained based upon observed driving behavior of vehicles;
- scoring a trajectory of the potential trajectories based on the initial cost volume and the delta cost volume;
- controlling the vehicle based upon the trajectory;
- and
- in response to computed measurements from the delta cost volume resulting in a finalized cost volume to satisfy a threshold measurement, adjusting the initial cost volume with the finalized cost volume associated with the potential trajectories, wherein the finalized cost volume comprises combining parts of the initial cost volume and the delta cost volume.
A system comprising:
- one or more processors and one or more computer-readable non-transitory storage media, the one or more computer-readable non-transitory storage media comprising instructions operable when executed by the one or more processors to cause the system to perform operations comprising:
- determining an initial cost volume associated with potential trajectories of a vehicle in an environment based on a set of movement restrictions of the vehicle, wherein the initial cost volume is based upon hand-engineered costs derived from the set of movement restrictions for the vehicle;
- generating, using a machine learning model, a delta cost volume using the initial cost volume and environment data of the environment, wherein the machine learning model is trained based upon observed driving behavior of vehicles;
- scoring a trajectory of the potential trajectories based on the initial cost volume and the delta cost volume;
- controlling the vehicle based upon the trajectory;
- and
- in response to computed measurements from the delta cost volume resulting in a finalized cost volume to satisfy a threshold measurement, adjusting the initial cost volume with the finalized cost volume associated with the potential trajectories, wherein the finalized cost volume comprises combining parts of the initial cost volume and the delta cost volume.
A computer-readable non-transitory storage media embodying software that is operable when executed to cause operations comprising:
- determining an initial cost volume associated with potential trajectories of a vehicle in an environment based on a set of movement restrictions of the vehicle, wherein the initial cost volume is based upon hand-engineered costs derived from the set of movement restrictions for the vehicle;
- generating, using a machine learning model, a delta cost volume using the initial cost volume and environment data of the environment, wherein the machine learning model is trained based upon observed driving behavior of vehicles;
- scoring a trajectory of the potential trajectories based on the initial cost volume and the delta cost volume;
- controlling the vehicle based upon the trajectory;
- and
- in response to computed measurements from the delta cost volume resulting in a finalized cost volume to satisfy a threshold measurement, adjusting the initial cost volume with the finalized cost volume associated with the potential trajectories, wherein the finalized cost volume comprises combining parts of the initial cost volume and the delta cost volume.
Dependencies are inferred from the claim text.