Granted Patent B2

US 11,094,148 · App. 16/010,646 · Granted Aug 17, 2021

Abstract

Claims (39)

1. A method for an autonomous vehicle, the method comprising:

detecting a collision of the vehicle with an object, the detecting comprising receiving first data used for object detection by the vehicle from at least one sensor of the vehicle, and comparing the first data to a threshold, wherein the vehicle comprises at least one computing device that controls the object detection, wherein the at least one computing device includes a controller and system memory, and wherein detecting the collision further comprises analyzing an output from an artificial neural network to identify the object based at least in part on the first data; and

in response to detecting the collision:

determining a first portion of the system memory to download;

detecting a type of event;

determining a priority based on the type of event; and

downloading the first data from portions of the system memory to a non-volatile memory device of the vehicle, wherein the downloaded portions include the first portion, and the first data is downloaded portion by portion from the system memory in an order based on the priority.

2. The method of claim 1 , further comprising analyzing, by a second computing device, the downloaded first data.

3. The method of claim 2 , further comprising:

extracting, by a third computing device, the downloaded first data from the non-volatile memory device; and

sending, over a wireless network, the extracted data to the second computing device for analysis.

4. The method of claim 2 , wherein analyzing the downloaded first data comprises analyzing a function of software used by the vehicle that is related to object detection.

5. The method of claim 4 , wherein the software controls navigation of the vehicle.

6. The method of claim 1 , wherein the system memory includes a volatile memory device that stores the first data prior to downloading the first data from the system memory to the non-volatile memory device.

7. The method of claim 1 , wherein the output from the artificial neural network classifies the type of event based at least in part on the first data.

8. An autonomous vehicle comprising:

at least one sensor configured to collect sensor data;

a volatile memory device to store data used by software during operation of the vehicle;

a non-volatile memory device configured to receive data from the volatile memory device;

at least one computing device configured to control at least one operation of the vehicle; and

memory storing instructions configured to instruct the at least one computing device to:

detect a collision of the vehicle with an object, wherein detecting the collision comprises analyzing an output from an artificial neural network to identify the object based on at least a portion of the sensor data; and

in response to detecting the collision:

download the sensor data from the volatile memory device to the non-volatile memory device, wherein the sensor data is downloaded in an order based on a priority; and

download program counter data from the volatile memory device to the non-volatile memory device, the program counter data indicating operational state of the software while executing during the collision.

9. The autonomous vehicle of claim 8 , wherein downloading the sensor data from the volatile memory device to the non-volatile memory device is completed in less than five seconds after detecting the collision.

10. The autonomous vehicle of claim 8 , wherein the sensor data is stored in system memory of the at least one computing device.

11. The autonomous vehicle of claim 8 , wherein the volatile memory device is a dynamic random access memory device.

12. The autonomous vehicle of claim 8 , wherein the non-volatile memory device is configured to provide the sensor data to a first computing device configured to analyze the sensor data.

13. The autonomous vehicle of claim 8 , wherein the at least one computing device includes a first computing device executing the software prior to detecting the collision, and wherein after detecting the collision, the software is analyzed by emulating the collision using the sensor data.

14. A non-transitory computer storage medium storing instructions which, when executed on a computing device, cause the computing device to perform a method for an autonomous vehicle, the method comprising:

detecting a collision of the vehicle with an object, wherein the vehicle comprises at least one computing device configured to use sensor data stored in a volatile memory device for object detection, and wherein detecting the collision comprises analyzing an output from an artificial neural network to identify the object based on the sensor data; and

in response to detecting the collision:

determining a first portion of the volatile memory to download;

detecting a type of event;

determining a priority based on the type of event; and

downloading the sensor data from the first portion of the volatile memory device to a non-volatile memory device in an order based on the priority.

15. The non-transitory computer storage medium of claim 14 , wherein the method further comprises analyzing, based on the sensor data, at least one function of software executing on the at least one computing device at the time of detecting the collision.

16. The non-transitory computer storage medium of claim 14 , wherein the method further comprises, in response to detecting the collision, selecting a type of data to download, wherein the sensor data comprises data of the selected type.

Assignments (6)