# US 12,594,967 B2 · Method and system for addressing failure in an autonomous agent

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 12,594,967 B2
- Document ID: US12594967B2
- Status: Granted
- Kind: B2
- Application number: 18/792,340
- Filed: 2024-08-01
- Published: 2024-12-05
- Granted: 2026-04-07
- Earliest claimed: 2021-12-16
- Assignees: Gatik AI Inc.
- Applicants: Gatik AI Inc.
- Inventors: Apeksha Kumavat; Brian McLean; Siddharth Ram; Arjun Narang; Gautam Narang
- Web page: https://patentgrep.com/patent/US12594967B2

## CPC

- B60W 50/023: Avoiding failures by using redundant parts
- B60W 60/00186: {related to the vehicle}
- B60W 50/0098: {Details of control systems ensuring comfort, safety or stability not otherwise provided for}
- B60W 50/0225: {Failure correction strategy}
- B60W 60/00: Drive control systems specially adapted for autonomous road vehicles
- B60W 60/0015: {specially adapted for safety}
- B60W 2050/0006: {Digital architecture hierarchy}
- B60W 2050/0215: {Sensor drifts or sensor failures}
- B60W: CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT

## Abstract

A system for addressing failure in an autonomous agent includes a driving subsystem, a control subsystem, a central computing subsystem, and an autonomous vehicle (AV) sensor subsystem. The system can optionally additionally include a power subsystem, a vehicle chassis subsystem, a communication subsystem, a distributed computing and/or processing subsystem, a supplementary sensor subsystem, and/or any other components. A method for addressing failure can include any or all of: detecting and responding to a failure; and operating the vehicle. Additionally or alternatively, the method 200 can include any other processes.

## Claims

### Claim 1

A method for targeted failure management of an autonomous vehicle, the method comprising, at an autonomous vehicle:
during a trip, checking for a 1st set of failures with a high-level processing subsystem configured to process environmental information from an environmental sensor subsystem;
checking for a 2nd set of failures with a set of distributed processing subsystems arranged within a set of vehicle subsystems of the autonomous vehicle, the set of vehicle subsystems comprising at least one of:
the environmental sensor subsystem;
a control subsystem;
a drive-by-wire subsystem comprising a set of actuators; or
a vehicle chassis subsystem;
in response to detecting one or more of the 1st set of failures, triggering one of a 1st set of failure responses, wherein at least a portion of the 1st set of failure responses comprises waiting until the autonomous vehicle has completed the trip before triggering an action; and
in response to detecting one or more of the 2nd set of failures, triggering one of a 2nd set of failure responses.

### Claim 2 (depends on 1)

The method of claim 1, wherein the one or more of the 1st set of failures is detected based on uncertainty values for the environmental information.

### Claim 3 (depends on 1)

The method of claim 1, wherein the high-level processing subsystem comprises:
a first high-level computing subsystem configured to determine a first trajectory;
a second high-level computing subsystem configured to determine a second trajectory; and
a control subsystem comprising:
a first controller configured to determine a first control instruction based on the first trajectory and the second trajectory; and
a second controller configured to determine a second control instruction based on the first trajectory and the second trajectory; and
wherein the autonomous vehicle is configured to operate based on the first control instruction and the second control instruction.

### Claim 4 (depends on 3)

The method of claim 3, wherein the control subsystem is configured to compare the first trajectory to the second trajectory, and wherein the first controller is configured to determine the first control instruction based on the comparison of the first trajectory to the second trajectory.

### Claim 5 (depends on 4)

The method of claim 4, wherein the control subsystem is further configured to compare the first control instruction to the second control instruction, and wherein the autonomous vehicle is operated based on the comparison of the first control instruction to the second control instruction.

### Claim 6 (depends on 1)

The method of claim 1, wherein the one of a 1st set of failure responses is determined based on a location of the autonomous vehicle.

### Claim 7 (depends on 1)

The method of claim 1, wherein the one or more of the 1st set of failures is separate and distinct from the one or more of the 2nd set of failures, and wherein the 1st set of failure responses is separate and distinct from the 2nd set of failure responses.

### Claim 8 (depends on 1)

The method of claim 1, wherein the one of the 2nd set of failure responses is determined based on a magnitude of the one or more of the 1st set of failures.

### Claim 9 (depends on 1)

The method of claim 1, further comprising reaching consensus between a threshold subset of the vehicle subsystems before triggering one of the 2nd set of failure responses.

### Claim 10 (depends on 1)

The method of claim 1, wherein a failure response within the 2nd set of failure responses comprises entering a minimal risk condition (MRC).

### Claim 11

An autonomous vehicle, comprising:
a high-level processing system configured to:
check for a set of high-level failures based on a set of autonomous vehicle data; and
when a high-level failure is detected during a trip, wait until the autonomous vehicle has completed the trip before triggering a first response action; and
a set of vehicle subsystems, each comprising a low-level processing system configured to:
check for a set of low-level failures based on data from the respective vehicle subsystem; and
when a low-level failure is detected during the trip, trigger a second of response action during the trip.

### Claim 12 (depends on 11)

The autonomous vehicle of claim 11, wherein the high-level failure is detected based on uncertainty values for information collected by a vehicle subsystem of the set of vehicle subsystems.

### Claim 13 (depends on 11)

The autonomous vehicle of claim 11, wherein the high-level processing system comprises:
a set of redundant high-level computing subsystems, each configured to determine a trajectory for the autonomous vehicle based on shared information from the set of vehicle subsystems;
a control subsystem comprising a set of redundant controllers, each configured to determine a control instruction set based on the trajectories determined by the redundant high-level computing subsystems;
wherein the autonomous vehicle is controlled based on each of the control instruction sets.

### Claim 14 (depends on 13)

The autonomous vehicle of claim 13, wherein the control subsystem further comprises a watchdog controller configured to selectively determine a third response action based on a comparison between the control instruction sets generated by the set of redundant controllers.

### Claim 15 (depends on 11)

The autonomous vehicle of claim 11, wherein the second response action is determined based on a geolocation of the autonomous vehicle.

### Claim 16 (depends on 11)

The autonomous vehicle of claim 11, wherein the first response action is determined based on a geolocation of the autonomous vehicle.

### Claim 17 (depends on 11)

The autonomous vehicle of claim 11, wherein at least one of the set of high-level failures or the set of low-level failures are determined based on a geolocation of the autonomous vehicle.

### Claim 18 (depends on 11)

The autonomous vehicle of claim 11, wherein the second response action is determined based on a magnitude of the low-level failure.

### Claim 19 (depends on 11)

The autonomous vehicle of claim 11, wherein consensus from a threshold subset of the set of vehicle subsystems is required before triggering the second response action.

### Claim 20 (depends on 11)

The autonomous vehicle of claim 11, wherein the second response action comprises entering a minimal risk condition (MRC).
