Granted Patent B2

US 12,654,746 · App. 18/823,001 · Granted Jun 16, 2026

Inventors: Ali Nasr (Waterloo, CA); Frederic Risacher (Kitchener, CA)

Assignee: Yinwang Intelligent Technologies Co., Ltd.

B60W60/0015B60W50/14B60W60/0053B60W2420/403B60W2420/408B60W2556/40B60W2556/50

Abstract

Claims (148)

1 . A computer-implemented method for controlling a vehicle, the method comprising:

executing a first processing pipeline, including:

triggering control over the vehicle; and

executing, in parallel to the first processing pipeline, a risk management processing pipeline, including:

acquiring first monitoring data from a first monitoring source used in the first processing pipeline;

acquiring second monitoring data from a second monitoring source used in the first processing pipeline;

generating a cumulative risk assessment value based on a combination of the first monitoring data and the second monitoring data;

in response to the cumulative risk assessment value being above a pre-determined main threshold, generating a warning notification;

in response to a frequency of occurrence of the generating the warning notification being above a pre-determined frequency threshold, triggering, independently from the first processing pipeline, a remedial action to be performed by the vehicle; and

acquiring external factor data indicative of external factors;

generating, for at least one of the first and second monitoring data, a respective severity level based on the external factor data;

wherein the cumulative risk assessment value is generated based on the following:

R

=

∑

(

F

×

S

)

,

wherein F is a respective frequency value of one of the first and second monitoring data exceeding a respective threshold value; and

S is the respective severity level associated with the one of the first and second monitoring data.

2 . The method of claim 1 , wherein the

external factors influence the at least one of the first monitoring data and the second monitoring data.

3 . The method of claim 2 , wherein the external factors comprise at least one of a current weather condition, a current traffic situation, a current environment of the vehicle, an ego-speed of the vehicle, and a current time of day.

4 . The method of claim 1 , wherein a given one of the first and second monitoring sources comprises one of a hardware component of a plurality of hardware components, a learning-enabled component (LEC) of a plurality of LECs, and a non-learning-enabled component (NLEC) of a plurality of NLECs of the vehicle that are used in the first processing pipeline for controlling the vehicle.

5 . The method of claim 4 , wherein the plurality of hardware components comprise a plurality of sensors of the vehicle and an electronic device communicatively coupled thereto,

the plurality of sensors comprising:

a camera sensor;

a LIDAR sensor;

a RADAR sensor;

a Global Positioning System (GPS) sensor; and

an Inertial Measurement Unit (IMU) sensor; and

the electronic device comprising:

a processor;

a read-access memory (RAM); and

a CAN-bus.

6 . The method of claim 4 , wherein the plurality of LECs comprises:

a camera object detection or tracking machine-learning (ML) model;

a LIDAR object detection or tracking ML model;

a RADAR object detection or tracking ML model;

a lane line detecting ML model;

a traffic light recognizing ML model; and

a trajectory predicting ML model.

7 . The method of claim 4 , wherein the plurality of NLECs comprises:

a trajectory planning model;

an object tracking model;

a global localization model; and

a high definition (HD) map.

8 . The method of claim 4 , wherein the given one of the first and second monitoring data is indicative of at least one of:

a connection status of each one of the plurality of hardware components;

a confidence level of a respective output of each one of the plurality of LECs;

consistency among respective outputs of the plurality of LECs and the plurality of NLECs;

a smoothness of a current trajectory traversed by the vehicle; and

a spatial and temporal proximity of the vehicle to at least one object in a surrounding area.

9 . The method of claim 1 , wherein the remedial action comprises at least one of generating a safety alert, handing over control of the vehicle to a human driver present in the vehicle, stopping the vehicle, and logging events occurring in the first processing pipeline.

10 . A system for controlling a vehicle, the system comprising at least one processor, at least one non-transitory computer-readable memory storing instructions, which, when executed by the at least one processor, cause the system to perform:

executing a first processing pipeline, including:

triggering control over the vehicle; and

executing, in parallel to the first processing pipeline, a risk management processing pipeline, including:

acquiring first monitoring data from a first monitoring source used in the first processing pipeline;

acquiring second monitoring data from a second monitoring source used in the first processing pipeline;

generating a cumulative risk assessment value based on a combination of the first monitoring data and the second monitoring data;

in response to the cumulative risk assessment value being above a pre-determined main threshold, generating a warning notification;

in response to a frequency of occurrence of the generating the warning notification being above a pre-determined frequency threshold, triggering, independently from the first processing pipeline, a remedial action to be performed by the vehicle; and

acquiring external factor data indicative of external factors;

generating, for at least one of the first and second monitoring data, a respective severity level based on the external factor data;

wherein the cumulative risk assessment value is generated based on the following:

R

=

∑

(

F

×

S

)

,

wherein F is a respective frequency value of one of the first and second monitoring data exceeding a respective threshold value; and

S is the respective severity level associated with the one of the first and second monitoring data.

11 . The system of claim 10 , wherein the

external factors influence the at least one of the first monitoring data and the second monitoring data.

12 . The system of claim 11 , wherein the external factors comprise at least one of a current weather condition, a current traffic situation, a current environment of the vehicle, an ego-speed of the vehicle, and a current time of day.

13 . The system of claim 10 , wherein a given one of the first and second monitoring sources comprises one of a hardware component of a plurality of hardware components, a learning-enabled component (LEC) of a plurality of LECs, and a non-learning-enabled component (NLEC) of a plurality of NLECs of the vehicle that are used in the first processing pipeline for controlling the vehicle.

14 . The system of claim 13 , wherein the plurality of hardware components comprise a plurality of sensors of the vehicle and an electronic device communicatively coupled thereto,

the plurality of sensors comprising:

a camera sensor;

a LIDAR sensor;

a RADAR sensor;

a Global Positioning System (GPS) sensor; and

an Inertial Measurement Unit (IMU) sensor; and

the electronic device comprising:

a processor;

a read-access memory (RAM); and

a CAN-bus.

15 . The system of claim 13 , wherein the plurality of LECs comprises:

a camera object detection or tracking machine-learning (ML) model;

a LIDAR object detection or tracking ML model;

a RADAR object detection or tracking ML model;

a lane line detecting ML model;

a traffic light recognizing ML model; and

a trajectory predicting ML model.

16 . The system of claim 13 , wherein the plurality of NLECs comprises:

a trajectory planning model;

an object tracking model;

a global localization model; and

a high definition (HD) map.

17 . The system of claim 13 , wherein the given one of the first and second monitoring data is indicative of at least one of:

a connection status of each one of the plurality of hardware components;

a confidence level of a respective output of each one of the plurality of LECs;

consistency among respective outputs of the plurality of LECs and the plurality of NLECs;

a smoothness of a current trajectory traversed by the vehicle; and

a spatial and temporal proximity of the vehicle to at least one object in a surrounding area.

18 . A non-transitory computer-readable medium storing executable instructions for causing one or more computer processors to:

execute a first processing pipeline, including:

triggering control over a vehicle; and

execute, in parallel to the first processing pipeline, a risk management processing pipeline, including:

acquiring first monitoring data from a first monitoring source used in the first processing pipeline;

acquiring second monitoring data from a second monitoring source used in the first processing pipeline;

generating a cumulative risk assessment value based on a combination of the first monitoring data and the second monitoring data;

in response to the cumulative risk assessment value being above a pre-determined main threshold, generating a warning notification;

in response to a frequency of occurrence of the generating the warning notification being above a pre-determined frequency threshold, triggering, independently from the first processing pipeline, a remedial action to be performed by the vehicle; and

acquiring external factor data indicative of external factors;

generating, for at least one of the first and second monitoring data, a respective severity level based on the external factor data;

wherein the cumulative risk assessment value is generated based on the following:

R

=

∑

(

F

×

S

)

,

wherein F is a respective frequency value of one of the first and second monitoring data exceeding a respective threshold value; and

S is the respective severity level associated with the one of the first and second monitoring data.

19 . The non-transitory computer-readable medium of claim 18 , wherein a given one of the first and second monitoring sources comprises one of a hardware component of a plurality of hardware components, a learning-enabled component (LEC) of a plurality of LECs, and a non-learning-enabled component (NLEC) of a plurality of NLECs of the vehicle that are used in the first processing pipeline for controlling the vehicle.

20 . The non-transitory computer-readable medium of claim 19 , wherein the plurality of hardware components comprise a plurality of sensors of the vehicle and an electronic device communicatively coupled thereto,

the plurality of sensors comprising:

a camera sensor;

a LIDAR sensor;

a RADAR sensor;

a Global Positioning System (GPS) sensor; and

an Inertial Measurement Unit (IMU) sensor; and

the electronic device comprising:

a processor;

a read-access memory (RAM); and

a CAN-bus.

Assignments (3)

Continuity (1)

References Cited (14)