Granted Patent B2

US 12,535,827 · App. 18/393,564 · Granted Jan 27, 2026

Inventors: Gerrit Bagschik (Lower Saxony, DE); Andrew Scott Crego (Palo Alto, CA); Avery Wagner Faller (San Francisco, CA); Matthew Ramin Hamedani Heffernan (San Francisco, CA); Francis Indaheng (East Palo Alto, CA); Andraz Kavalar (San Francisco, CA); Aditya Pramod Khadilkar (Foster City, CA); Deepan Subrahmanian Palguna (Santa Clara, CA)

Assignee: Zoox, Inc.

G05D1/225

Abstract

Claims (67)

1 . A system comprising,

one or more processors; and

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

receiving a driving simulation scenario specifying:

a maneuver for a simulated vehicle,

a position of a simulated agent proximate the simulated vehicle, and

a road context to be simulated;

generating a database query based at least in part on the driving simulation scenario;

retrieving from a database, based at least in part on the database query, a plurality of driving events associated with the driving simulation scenario, comprising:

identifying a set of driving events in the database associated with the maneuver; and

retrieving the plurality of driving events from the set of driving events based at least in part on similarity metrics between the driving events in the set of driving events and the driving simulation scenario;

determining, based at least in part on the plurality of driving events, a metric associated with a likelihood of occurrence of the driving simulation scenario;

initiating a simulation based at least in part on the driving simulation scenario;

determining, based at least in part on the simulation and the metric, a safety metric; and

transmitting, based at least in part on the safety metric meeting or exceeding a threshold safety value, a signal to a vehicle of a fleet of vehicles in a real-world environment, the signal configured to control the vehicle to one or more of drive or refrain from driving in a region of the real-world environment associated with the driving simulation scenario.

2 . The system of claim 1 , wherein the safety metric comprises a first safety metric, the operations comprising:

modifying a parameter associated with the driving simulation scenario to create an altered driving simulation scenario;

determining, based at least in part on the altered driving simulation scenario, a modified metric;

instantiating a second simulation based at least in part on the altered driving simulation scenario; and

determining, based at least in part on the modified metric and the second simulation, a second safety metric,

wherein transmitting the signal is further based at least in part on the second safety metric.

3 . The system of claim 2 , wherein:

the parameter comprises an object type comprising one or more of a car, a truck, a pedestrian, a motorcycle, or a bicycle; and

determining the modified metric comprises scaling the metric by a scalar determined based at least in part on the object type.

4 . The system of claim 1 , wherein the operations comprise identifying a scenario type associated with the driving simulation scenario, the scenario type associated with one or more of: continuous driving, entering a road, exiting a road, turning, approaching an intersection, traversing an intersection, picking up a passenger, or dropping off a passenger.

5 . A method comprising:

receiving a driving simulation scenario associated with a simulated vehicle;

determining, based at least in part on the driving simulation scenario and a dataset of recorded driving associated with operation of a test vehicle in an environment, a plurality of driving events present in the dataset, comprising:

determining the plurality of driving events based at least in part on similarity metrics between the driving events present in the dataset and the driving simulation scenario;

determining, based at least in part on the plurality of driving events, a first metric associated with a likelihood of occurrence of the driving simulation scenario;

instantiating a simulation based at least in part on the driving simulation scenario;

determining, based at least in part on the simulation, a safety metric indicative of the safe performance of the simulated vehicle in the simulation;

determining, based at least in part on the safety metric and the first metric, an aggregate safety metric; and

based at least in part on the aggregate safety metric, transmitting a signal to a vehicle of a fleet of vehicles to be controlled autonomously.

6 . The method of claim 5 , wherein the dataset comprises abstracted representations of log data comprising one or more vehicle maneuvers over periods of time.

7 . The method of claim 5 , wherein determining the plurality of driving events comprises:

determining a similarity metric between a first driving event in the dataset and the driving simulation scenario; and

including the first driving event in the plurality of driving events based at least in part on the similarity metric meeting or exceeding a similarity threshold.

8 . The method of claim 5 , wherein:

the driving simulation scenario is associated with a road context to be simulated, and

the method comprises identifying the plurality of driving events based at least in part on the road context.

9 . The method of claim 5 , wherein the driving simulation scenario specifies, as simulation information, one or more of: a vehicle mode, a position of the simulated vehicle, or a maneuver to be performed by the simulated vehicle, and

wherein determining the plurality of driving events comprises determining a plurality of driving events associated with information similar to the simulation information.

10 . The method of claim 5 , wherein the driving simulation scenario specifies simulated agent information associated with simulated agents, and

wherein the method comprises identifying the plurality of driving events based at least in part on the simulated agent information.

11 . The method of claim 5 , comprising:

determining, based at least in part on map data and the driving simulation scenario, a road segment of the map data;

wherein the simulation includes the road segment.

12 . The method of claim 5 , comprising updating the first metric based at least in part on one or more of: a change associated with the driving simulation scenario; a change associated with the dataset; a change associated with the environment; or a predetermined schedule.

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

receiving a driving simulation scenario associated with a simulated vehicle;

determining, based at least in part on the driving simulation scenario and a dataset of recorded driving associated with operation of a test vehicle in an environment, a plurality of driving events present in the dataset, comprising:

determining the plurality of driving events based at least in part on similarity metrics between the driving events present in the dataset and the driving simulation scenario;

determining, based at least in part on the plurality of driving events, a first metric associated with a likelihood of occurrence of the driving simulation scenario;

instantiating a simulation based at least in part on the driving simulation scenario;

determining, based at least in part on the simulation, a safety metric indicative of the safe performance of the simulated vehicle in the simulation;

determining, based at least in part on the safety metric and the first metric, an aggregate safety metric; and

based at least in part on the aggregate safety metric, transmitting a signal to a vehicle of a fleet of vehicles to be controlled autonomously.

14 . The one or more non-transitory computer-readable media of claim 13 , the operations comprising determining one or more statistical models representing parameters associated with the plurality of driving events.

15 . The one or more non-transitory computer-readable media of claim 13 , wherein the signal is configured to control the vehicle to one or more of drive or refrain from driving in a region of the real-world environment associated with the driving simulation scenario; or to one or more of allow the vehicle to perform a maneuver or prevent the vehicle from performing the maneuver, the maneuver associated with the driving simulation scenario.

16 . The one or more non-transitory computer-readable media of claim 13 , wherein the dataset comprises abstracted representations associated with the driving events in the dataset, the operations comprising:

determining the plurality of driving events based at least in part on a similarity between abstracted information associated with the driving simulation scenario and the abstracted representations associated with the driving events.

17 . The one or more non-transitory computer-readable media of claim 16 , wherein the abstracted representations comprise one or more of: vehicle maneuvers associated with the test vehicle; agent maneuvers associated with agents in the environment proximate to the test vehicle; or a road context associated with the environment.

18 . The one or more non-transitory computer-readable media of claim 13 , the operations comprising:

identifying a scenario type associated with the driving simulation scenario; and

determining the plurality of driving events based at least in part on the scenario type.

19 . The one or more non-transitory computer-readable media of claim 18 , wherein the scenario type is associated with one or more of: continuous driving, entering a road, exiting a road, turning, approaching an intersection, traversing an intersection, picking up a passenger, or dropping off a passenger.

Assignments (1)

Continuity (1)

References Cited (27)