System and method for error handling of an uncalibrated sensor

Patent metadata

Field	Value
Publication number	US-11899140-B2
Application number	US-202217892403-A
Country	US
Kind code	B2
Filing date	Aug 22, 2022
Priority date	May 8, 2019
Publication date	Feb 13, 2024
Grant date	Feb 13, 2024

Abstract

Provided herein is a system and method for determining whether a sensor is calibrated and error handling of an uncalibrated sensor. The system comprises a sensor system comprising a sensor and an analysis engine configured to determine whether the sensor is uncalibrated. The system further comprises an error handling system configured to perform an error handling in response to the sensor system determining that the sensor is uncalibrated. The method comprises determining, by a sensor system, whether the sensor is uncalibrated, and performing, by an error handling system, an error handling in response to the sensor system determining that the sensor is uncalibrated.

First claim

What is claimed is:

1. A system comprising:
a sensor; and
one or more processors configured to:
determine a probability of whether the sensor is uncalibrated based on one or more sensor parameters or a comparison of a proportion of features from data captured by the sensor and historical data at a same location, wherein the sensor parameters comprising any of a skew, a depth of field, an angle of view, a beam angle, an aspect ratio, and a pixel number;
in response to determining that the probability of whether the sensor is uncalibrated satisfying a threshold probability, select a first error handling mechanism based on a previous rate of success of recalibration relative to a specific location, a specific landmark, or a type of landmark, wherein the selecting of the first error handling mechanism comprises:
selecting a recalibration mechanism in response to the previous rate of success satisfying a threshold; and
selecting a different error handling mechanism in response to the previous rate of success failing to satisfy the threshold;

perform the first error handling mechanism in accordance with the selection of the error handling mechanism;
determine whether the sensor is validated following the first error handling mechanism;
in response to determining that the sensor is not validated following the first error handling mechanism, perform a second error handling mechanism and a second validation;
upon determining that the sensor is validated after the first error handling mechanism or the second error handling mechanism, capture sensor data using the sensor after the first error handling mechanism or the second error handling mechanism; and
navigate a vehicle using the sensor data.

2. The system of claim 1 , wherein:
the determination of whether the sensor is uncalibrated comprises determining a type of sensor error, the type of the sensor error comprising any of an offset, a sensitivity or a slope error, and a deviation from linearity; and
the selecting of the recalibration mechanism, in response to the previous rate of success satisfying a threshold, is based on the type of the sensor error.

3. The system of claim 1 , wherein the threshold probability comprises a first threshold probability, and the selecting of the first error handling mechanism is based on the probability of whether the sensor is uncalibrated.

4. The system of claim 3 , wherein the selecting of the first error handling mechanism comprises:
in response to the probability failing to satisfy a second threshold probability, selecting the first error handling mechanism as one of eliminating erroneous sensor data or issuing an alert; and
in response to the probability satisfying the second threshold probability, selecting the first error handling mechanism as one of recalibrating the sensor or initiating a backup sensor.

5. The system of claim 1 , wherein:
the probability comprises a first probability determined prior to the first error handling mechanism; and
the performing of the second error handling mechanism comprises determining the second error handling mechanism based on the first probability, a second probability that the sensor is uncalibrated following the first error handling mechanism, or a difference between the first probability and the second probability.

6. The system of claim 1 , wherein:
the performing of the second error handling mechanism comprises determining the second error handling mechanism based on a density of moving objects at a current location.

7. The system of claim 1 , wherein:
the performing of the second error handling mechanism comprises determining the second error handling mechanism based on respective load consumptions of potential error handling mechanisms.

8. The system of claim 1 , wherein:
the performing of the second error handling mechanism comprises determining the second error handling mechanism based on an availability of a system load.

9. The system of claim 1 , wherein:
the selecting of the first error handling mechanism is based on respective load consumptions of potential error handling mechanisms.

10. The system of claim 1 , wherein:
the selecting of the first error handling mechanism is based on an availability of a system load.

11. An error handling method for a sensor of a system, performed by one or more processors, the error handling method comprising:
determining a probability of whether the sensor is uncalibrated based on one or more sensor parameters or a comparison of a proportion of features from data captured by the sensor and historical data at a same location, wherein the sensor parameters comprising any of a skew, a depth of field, an angle of view, a beam angle, an aspect ratio, and a pixel number;
in response to determining that the probability of whether the sensor is uncalibrated satisfying a threshold probability, selecting a first error handling mechanism based on a previous rate of success of recalibration relative to a specific location, a specific landmark, or a type of landmark, wherein the selecting of the first error handling mechanism comprises:
selecting a recalibration mechanism in response to the previous rate of success satisfying a threshold; and
selecting a different error handling mechanism in response to the previous rate of success failing to satisfy the threshold;

performing the first error handling mechanism in accordance with the selection of the error handling mechanism;
determining whether the sensor is validated following the first error handling mechanism;
in response to determining that the sensor is not validated following the first error handling mechanism, performing a second error handling mechanism and a second validation;
upon determining that the sensor is validated after the first error handling mechanism or the second error handling mechanism, capturing sensor data using the sensor after the first error handling mechanism or the second error handling mechanism; and
navigating a vehicle using the sensor data.

12. The error handling method of claim 11 , wherein:
the determination of whether the sensor is uncalibrated comprises determining a type of sensor error, the type of the sensor error comprising any of an offset, a sensitivity or a slope error, and a deviation from linearity; and
the selecting of the recalibration mechanism, in response to the previous rate of success satisfying a threshold, is based on the type of the sensor error.

13. The method of claim 11 , wherein the threshold probability comprises a first threshold probability, and the selecting of the first error handling mechanism is based on the probability of whether the sensor is uncalibrated.

14. The method of claim 13 , wherein the selecting of the first error handling mechanism comprises:
in response to the probability failing to satisfy a second threshold probability, selecting the first error handling mechanism as one of eliminating erroneous sensor data or issuing an alert; and
in response to the probability satisfying the second threshold probability, selecting the first error handling mechanism as one of recalibrating the sensor or initiating a backup sensor.

15. The method of claim 11 , wherein:
the probability comprises a first probability determined prior to the first error handling mechanism; and
the performing of the second error handling mechanism comprises determining the second error handling mechanism based on the first probability, a second probability that the sensor is uncalibrated following the first error handling mechanism, or a difference between the first probability and the second probability.

Assignees

• Pony Ai Inc

Inventors

• Meng Pingfan
• Pan Zhenhao

Classifications

• G01S7/4972 Primary

  Alignment of sensor · CPC title

• G01S7/40

  Means for monitoring or calibrating · CPC title

• G01S7/52004 Primary

  Means for monitoring or calibrating (short-range imaging G01S7/5205) · CPC title

• H04L67/12

  specially adapted for proprietary or special-purpose networking environments, e.g. medical networks, sensor networks, networks in vehicles or remote metering networks · CPC title

• G01S15/931

  of land vehicles · CPC title