IP Library Granted Patent US 12,511,997
Granted Patent B2
US 12,511,997 · App. 18/912,014 · Granted Dec 30, 2025

Techniques for autonomous vehicle accident mitigation

Inventors: Liu Liu (San Jose, CA); Yefei Peng (Palo Alto, CA); Hao Zheng (Saratoga, CA); Yu Sun (Santa Clara, CA)
Assignee: PlusAI, Inc.
G08G1/096725B60W60/001G08G1/0116G08G1/0133G08G1/096783G08G1/166B60W2556/50B60W2720/10
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,511,997
App. No.
18/912,014
Granted
Dec 30, 2025
Kind
B2
Abstract

Techniques are described for scenario data generation. An example method can include processing location-specific sensor information of a traffic accident at a location on a road network, the location-specific sensor information received from a sensor arranged on an infrastructure at the road network. The method can further include determining a location-specific element based on the location-specific sensor information. The method can further include determining a location-specific contributing factor of the traffic accident based on the location-specific element. The method can further include generating a location-specific model for simulating the traffic accident based on the location-specific contributing factor. The method can further include determining control instructions for a location-specific driving maneuver for an autonomous vehicle to mitigate a future traffic accident based on the location-specific model. The method can further include causing the control instructions to be transmitted to an actuator of the autonomous vehicle to perform the location-specific driving maneuver.

Claims (74)

1 . A method, comprising:

processing, by a computing system, location-specific sensor information of a traffic accident at a location on a road network, the location-specific sensor information received from a sensor arranged on an infrastructure at the road network, the sensor having a field of view based on being arranged on the infrastructure, and the location-specific sensor information based on the field of view of the sensor;

determining, by the computing system, a location-specific element based on the location-specific sensor information;

determining, by the computing system, a location-specific contributing factor of the traffic accident based on the location-specific element;

generating, by the computing system, a location-specific model for simulating the traffic accident based on the location-specific contributing factor;

accessing, by the computing system, the location-specific sensor information based at least in part on generating the location-specific model;

determining, by the computing system, a fidelity of the location-specific model to the location-specific sensor information;

determining, by the computing system, an acceptability of the location-specific model based at least in part on the fidelity;

incorporating, by the computing system, a representation of a first hypothetical autonomous vehicle into the location-specific model;

determining, by the computing system, control instructions for a location-specific driving maneuver for a second autonomous vehicle to mitigate a future traffic accident based on the location-specific model, the control instructions determined based on incorporating the representation of the first hypothetical autonomous vehicle into the location-specific model; and

causing, by the computing system, the control instructions to be transmitted to an actuator of the second autonomous vehicle to perform the location-specific driving maneuver thereby controlling the second autonomous vehicle at the location based on the location-specific model.

2 . The method of claim 1 , wherein the location-specific element is a first location-specific element, and wherein the method further comprises:

receiving real time location-specific sensor information from the sensor; and

determining a second location-specific element indicating a traffic accident based on the real time location-specific sensor information, wherein the control instructions are transmitted based on the determining the second location-specific element indicating the traffic accident.

3 . The method of claim 1 , wherein the method further comprises;

generating synthetic information associated with the autonomous vehicle, wherein the synthetic information incorporates a representation of the autonomous vehicle into the location-specific model, and wherein the control instructions are determined based on incorporating the representation of the autonomous vehicle into the location-specific model.

4 . The method of claim 1 , wherein the location-specific sensor information is associated with a stationary space based on the sensor being a stationary sensor, and wherein the method further comprises:

determining a transformation function for converting the location-specific sensor information from a stationary space representation to a moving space representation, and wherein the location-specific model is determined based on the moving space representation.

5 . The method of claim 1 , wherein the location is a first location along a route, wherein the location-specific sensor information is first location specific sensor information, wherein the infrastructure is a first infrastructure, and wherein the method further comprises:

identifying second location-specific sensor information from a second sensor arranged on a second infrastructure at a second location along the route, wherein the location-specific model is determined based on the first location specific sensor information and second location-specific sensor information.

6 . The method of claim 5 , wherein the method further comprises:

determining a route of the autonomous vehicle based on the road network;

determining that the second location is on the route;

determining that a third sensor is located at a third location outside of the route; and

determining to use a second location-specific element associated with the second location to determine the location-specific model instead of a third location-specific element associated with the third location based on the second location being on the route and the third location being located outside of the route.

7 . The method of claim 1 , wherein the location-specific element is determined based on information from a driver assistance system.

8 . The method of claim 1 , wherein the location-specific element is a road element describing a characteristic of a road upon which the traffic accident occurred.

9 . The method of claim 8 , wherein the method further comprises;

determining a vehicle element based on the location-specific sensor information, the vehicle element describing a characteristic of a vehicle involved in the traffic accident, wherein the location-specific model is determined based on the vehicle element.

10 . The method of claim 1 , wherein mitigating the future traffic accident comprises avoiding the future traffic accident.

11 . The method of claim 1 , wherein the method further comprises:

receiving an identification of an element not described in the location-specific sensor information; and

generating a location-specific synthetic element representing the element, and wherein the location-specific model is determined based on the location-specific synthetic element.

12 . A computing system, comprising:

one or more processors; and

one or more non-transitory, computer-readable media including instructions that, when executed, cause one or more processors to:

process location-specific sensor information of a traffic accident at a location on a road network, the location-specific sensor information received from a sensor arranged on an infrastructure at the road network, the sensor having a field of view based on being arranged on the infrastructure, and the location-specific sensor information based on the field of view of the sensor;

determine a location-specific element based on the location-specific sensor information;

determine a location-specific contributing factor of the traffic accident based on the location-specific element;

generate a location-specific model for simulating the traffic accident based on the location-specific contributing factor;

access the location-specific sensor information based at least in part on generating the location-specific model;

determine a fidelity of the location-specific model to the location-specific sensor information;

determine an acceptability of the location-specific model based at least in part on the fidelity;

incorporate a representation of a first hypothetical autonomous vehicle into the location-specific model;

determine control instructions for a location-specific driving maneuver for a second autonomous vehicle to mitigate a future traffic accident based on the location-specific model, the control instructions determined based on incorporating the representation of the first hypothetical autonomous vehicle into the location-specific model; and

cause the control instructions to be transmitted to an actuator of the second autonomous vehicle to perform the location-specific driving maneuver thereby controlling the second autonomous vehicle at the location based on the location-specific model.

13 . The computing system of claim 12 , wherein the location-specific element is a first location-specific element, and wherein the instructions that, when executed, further cause the one or more processors to:

receive real time location-specific sensor information from the sensor; and

determine a second location-specific element indicating a traffic accident based on the real time location-specific sensor information, wherein the control instructions are transmitted based on determining the second location-specific element indicating the traffic accident.

14 . The computing system of claim 12 , wherein the instructions that, when executed, further cause the one or more processors to:

generating synthetic information associated with the autonomous vehicle, wherein the synthetic information incorporates the autonomous vehicle into the location-specific model, and wherein the control instructions are determined based on incorporates the autonomous vehicle into the location-specific model.

15 . The computing system of claim 12 , wherein the location-specific sensor information is associated with a stationary space based on the sensor being a stationary sensor, and wherein the instructions that, when executed, further cause the one or more processors to:

determining a transformation function for converting the location-specific sensor information from a stationary space representation to a moving space representation, and wherein the location-specific model is determined based on the moving space representation.

16 . The computing system of claim 12 , wherein the location is a first location, wherein the location-specific sensor information is first location specific sensor information, wherein the infrastructure is a first infrastructure, and wherein the instructions that, when executed, further cause the one or more processors to:

identify a second location-specific element from a second sensor arranged on a second infrastructure at a second location on the road network, wherein the location-specific model is determined based on the second location-specific element.

17 . The computing system of claim 16 , wherein the instructions that, when executed, further cause the one or more processors to:

determine a route of the autonomous vehicle based on the road network;

determine that the second location is on the route; and

determine to use the second location-specific element to determine the location-specific model instead of third location-specific element associated with a third sensor that is outside of the route based on the second location being on the route.

18 . The computing system of claim 12 , wherein the location-specific element is determined based on information from a driver assistance system.

19 . One or more non-transitory computer-readable media having stored thereon a sequence of instructions which, when executed, cause one or more processors to:

process location-specific sensor information of a traffic accident at a location on a road network, the location-specific sensor information received from a sensor arranged on an infrastructure at the road network, the sensor having a field of view based on being arranged on the infrastructure, and the location-specific information based on the field of view of the sensor;

determine a location-specific element based on the location-specific sensor information;

determine a location-specific contributing factor of the traffic accident based on the location-specific element;

generate a location-specific model for simulating the traffic accident based on the location-specific contributing factor;

access the location-specific sensor information based at least in part on generating the location-specific model;

determine a fidelity of the location-specific model to the location-specific sensor information;

determine an acceptability of the location-specific model based at least in part on the fidelity;

incorporate a representation of a first hypothetical autonomous vehicle into the location-specific model;

determine control instructions for a location-specific driving maneuver for a second autonomous vehicle to avoid a future traffic accident based on the location-specific model, the control instructions determined based on incorporating the representation of the first hypothetical autonomous vehicle into the location-specific model; and

cause the control instructions to be transmitted an actuator of the second autonomous vehicle to perform the location-specific driving maneuver thereby controlling the second autonomous vehicle at the location based on the location-specific model.

20 . The one or more non-transitory, computer-readable media of claim 19 , wherein the location-specific element is a first location-specific element, and wherein the instructions that, when executed, further cause the one or more processors to:

receive real time location-specific sensor information from the sensor; and

determine a second location-specific element indicating a traffic accident based on the real time location-specific sensor information, wherein the control instructions are transmitted based on determining the second location-specific element indicating the traffic accident.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Oct 10, 2024
From: LIU, LIU; PENG, YEFEI; ZHENG, HAO; SUN, YU
To: PLUSAI, INC.
Reel/Frame 068866/0643 →
Continuity (2)
Provisional Application 63590335 · Oct 13, 2023
Related Publication 20250124787A1 · Apr 17, 2025
References Cited (44)
US 8760314B2 · Ramachandran · 2014 [cited by examiner]
US 9646428B1 · Konrardy · 2017 [cited by examiner]
US 9805601B1 · Fields · 2017 [cited by examiner]
US 10782693B2 · Zhang · 2020 [cited by examiner]
US 10782694B2 · Zhang · 2020 [cited by examiner]
US 11714971B2 · Wray · 2023 [cited by examiner]
US 11772666B2 · Zhu · 2023 [cited by examiner]
US 12174027B2 · Konrardy · 2024 [cited by examiner]
US 20140278034A1 · Marko · 2014 [cited by examiner]
US 20150239413A1 · Kozloski · 2015 [cited by examiner]
US 20150242552A1 · Kozloski · 2015 [cited by examiner]
US 20160059855A1 · Rebhan · 2016 [cited by examiner]
US 20170256164A1 · Haag · 2017 [cited by examiner]
US 20170341513A1 · Wu · 2017 [cited by examiner]
US 20180059669A1 · Madigan et al. · 2018 [cited by applicant]
US 20180299284A1 · Wang · 2018 [cited by examiner]
US 20190012909A1 · Mintz · 2019 [cited by applicant]
US 20190384303A1 · Muller · 2019 [cited by examiner]
US 20200132497A1 · Glebov et al. · 2020 [cited by applicant]
US 20200326719A1 · Tram · 2020 [cited by examiner]
US 20210005085A1 · Cheng · 2021 [cited by examiner]
US 20210233396A1 · Guo et al. · 2021 [cited by applicant]
US 20210287536A1 · Siltanen · 2021 [cited by examiner]
US 20220048533A1 · Ödblom · 2022 [cited by applicant]
US 20220068123A1 · Guo et al. · 2022 [cited by applicant]
US 20220126864A1 · Moustafa et al. · 2022 [cited by applicant]
US 20220269279A1 · Redford · 2022 [cited by examiner]
US 20220309920A1 · Yang · 2022 [cited by examiner]
US 20220351613A1 · Hu et al. · 2022 [cited by applicant]
US 20220375340A1 · Pittman · 2022 [cited by examiner]
US 20230004165A1 · Sun · 2023 [cited by examiner]
US 20230079196A1 · Hari · 2023 [cited by examiner]
US 20230139072A1 · Haas et al. · 2023 [cited by applicant]
US 20230256999A1 · Donderici · 2023 [cited by examiner]
US 20230386325A1 · Tseng · 2023 [cited by examiner]
US 20240092390A1 · Philion · 2024 [cited by examiner]
US 20240174242A1 · Schaaf · 2024 [cited by applicant]
US 20240233523A1 · Vogel · 2024 [cited by applicant]
US 20240326814A1 · Al-Stouhi et al. · 2024 [cited by applicant]
US 20240371264A1 · Budan · 2024 [cited by examiner]
US 20240416907A1 · Barrera · 2024 [cited by examiner]
US 20250124788A1 · Liu · 2025 [cited by examiner]
U.S. Appl. No. 18/912,274, Non-Final Office Action, Mailed on Jan. 17, 2025, 15 pages. [cited by applicant]
U.S. Appl. No. 18/912,274 , Final Office Action, Mailed on May 27, 2025, 6 pages. [cited by applicant]