Method and apparatus for changing route when error occurs in autonomous driving artificial intelligence

1 . A method for changing a route of an autonomous driving vehicle when an error occurs in an autonomous driving artificial intelligence, the method comprising:

determining whether an error has occurred in an autonomous driving artificial intelligence;

collecting error information of the artificial intelligence when the error of the artificial intelligence has occurred;

extracting, from a storage, past error information about a same kind and version of another artificial intelligence as a kind and version of the artificial intelligence based on the error information of the artificial intelligence;

generating an error analysis result based on the past error information;

generating an error analysis result message based on the error analysis result;

determining whether driving of the autonomous driving vehicle needs to be stopped based on the error analysis result message; and

moving the autonomous driving vehicle to a safe place and stopping the driving of the autonomous driving vehicle when it is determined that the driving of the autonomous driving vehicle needs to be stopped,

wherein the determining whether an error has occurred in the autonomous driving artificial intelligence includes determining whether the artificial intelligence has accurately recognized an object based on a comparison among recognition results of a plurality of sensors,

wherein the past error information includes degree of error, number of error occurrences, number of accidents, and a vehicle accident risk caused by errors of the same kind and version of the other artificial intelligence as the kind and version of the artificial intelligence in which the error has occurred,

wherein the error analysis result includes a vehicle accident risk caused by the error of the artificial intelligence,

wherein the error analysis result includes error avoidance information,

wherein the generating of the error analysis result message includes including an autonomous driving operation availability information indicating ‘operation continuation’ and the error avoidance information into the error analysis result message when the vehicle accident risk caused by the error of the artificial intelligence is less than a first predetermined threshold value,

wherein the method further comprises resetting a driving route of the autonomous driving vehicle based on the error avoidance information when it is determined that it is not necessary to stop the driving of the autonomous driving vehicle,

wherein the generating of the error analysis result includes generating the error avoidance information based on the past error information,

wherein the error avoidance information includes an error occurrence cause and an error occurrence place at a time of error occurrence in the past error information, and

wherein the resetting includes resetting the driving route so as not to include an error occurrence place associated with an error occurrence cause having a frequency that is above a second predetermined threshold value.

2 . The method for changing a route of an autonomous driving vehicle when an error occurs in an autonomous driving artificial intelligence according to claim 1 , wherein the extracting of the past error information from the storage includes extracting, from the storage, the past error information about the same kind and version of the other artificial intelligence as the kind and version of the artificial intelligence in which the error has occurred.

3 . The method for changing a route of an autonomous driving vehicle when an error occurs in an autonomous driving artificial intelligence according to claim 1 , wherein the error analysis result includes the vehicle accident risk caused by the error of the artificial intelligence, and

wherein the generating of the error analysis result message includes including an autonomous driving operation availability information indicating ‘operation interruption’ into the error analysis result message when the vehicle accident risk caused by the error of the artificial intelligence is greater than or equal to the first predetermined threshold value.

4 . The method for changing a route of an autonomous driving vehicle when an error occurs in an autonomous driving artificial intelligence according to claim 3 , wherein the determining of whether the driving of the autonomous driving vehicle needs to be stopped includes determining that the driving of the autonomous driving vehicle needs to be stopped when the error analysis result message includes the autonomous driving operation availability information indicating ‘operation interruption’.

5 . The method for changing a route of an autonomous driving vehicle when an error occurs in an autonomous driving artificial intelligence according to claim 1 , wherein the error analysis result message includes the vehicle accident risk caused by the error of the artificial intelligence, and

wherein the determining of whether the driving of the autonomous driving vehicle needs to be stopped includes determining that the driving of the autonomous driving vehicle needs to be stopped when the vehicle accident risk caused by the error of the artificial intelligence is greater than or equal to the first predetermined threshold value.

6 . The method for changing a route of an autonomous driving vehicle when an error occurs in an autonomous driving artificial intelligence according to claim 1 , wherein the determining of whether the driving of the autonomous driving vehicle needs to be stopped includes determining that it is not necessary to stop the driving of the autonomous driving vehicle when the error analysis result message includes the autonomous driving operation availability information indicating ‘operation continuation’.

7 . The method for changing a route of an autonomous driving vehicle when an error occurs in an autonomous driving artificial intelligence according to claim 1 , further comprising resetting the driving route of the autonomous driving vehicle when it is determined that it is not necessary to stop the driving of the autonomous driving vehicle.

8 . A method for changing a route in an autonomous driving system installed on an autonomous driving vehicle when an error occurs in an autonomous driving artificial intelligence, the method comprising:

determining whether an error has occurred in an autonomous driving artificial intelligence;

collecting error information of the artificial intelligence when the error of the artificial intelligence has occurred;

transmitting the error information of the artificial intelligence to a cloud server;

receiving an error analysis result message corresponding to the error information of the artificial intelligence from the cloud server;

determining whether driving of the autonomous driving vehicle needs to be stopped based on the error analysis result message;

moving the autonomous driving vehicle to a safe place and stopping the driving of the autonomous driving vehicle when it is determined that the driving of the autonomous driving vehicle needs to be stopped; and

resetting a driving route of the autonomous driving vehicle when it is not necessary to stop the driving of the autonomous driving vehicle,

wherein the determining whether an error has occurred in the autonomous driving artificial intelligence includes determining whether the artificial intelligence has accurately recognized an object based on a comparison among recognition results of a plurality of sensors,

wherein the resetting includes resetting the driving route of the autonomous driving vehicle based on error avoidance information included in the error analysis result message,

wherein the error avoidance information includes an error occurrence cause and an error occurrence place at a time of error occurrence in past error information generated in a same kind and version of another artificial intelligence as a kind and version of the artificial intelligence, and

wherein the resetting includes resetting the driving route so as not to include an error occurrence place associated with an error occurrence cause having a frequency that is above a first predetermined threshold.

9 . The method for changing a route in an autonomous driving system installed on an autonomous driving vehicle when an error occurs in an autonomous driving artificial intelligence according to claim 8 , wherein the determining of whether the driving of the autonomous driving vehicle needs to be stopped includes determining that the driving of the autonomous driving vehicle needs to be stopped when a vehicle accident risk included in the error analysis result message is greater than or equal to a second predetermined threshold value.

10 . The method for changing a route in an autonomous driving system installed on an autonomous driving vehicle when an error occurs in an autonomous driving artificial intelligence according to claim 8 , wherein the determining of whether the driving of the autonomous driving vehicle needs to be stopped includes determining that the driving of the autonomous driving vehicle needs to be stopped when the error analysis result message includes the autonomous driving operation availability information indicating ‘operation interruption’.

11 . An autonomous driving system installed on an autonomous driving vehicle, the system comprising a processor and memory configured to implement:

a recognition module configured to detect an object based on sensor data, determine a kind of the object using an artificial intelligence with respect to the detected object, and generate object information;

an error detection module configured to determine whether an error has occurred in the artificial intelligence based on the object information;

an error information collection module configured to collect error information of the artificial intelligence when the error of the artificial intelligence has occurred;

a transmission and reception module configured to transmit the error information to a cloud server, and receive an error analysis result message corresponding to the error information from the cloud server;

an error response module configured to respond to the error of the artificial intelligence based on the error analysis result message;

a vehicle control module configured to move the autonomous driving vehicle to a safe place and stop the driving of the autonomous driving vehicle when it is determined that the driving of the autonomous driving vehicle needs to be stopped; and

a determination module configured to set a driving route of the autonomous driving vehicle,

wherein the error detection module is configured to determine the artificial intelligence has accurately recognized an object based on a comparison among recognition results of a plurality of sensors,

wherein the error response module delivers error avoidance information included in the error analysis result message to the determination module when a vehicle accident risk is included in the error analysis result message and the vehicle accident risk is less than a first predetermined threshold value,

wherein the error avoidance information includes an error occurrence cause and an error occurrence place at a time of error occurrence in past error information generated in a same kind and version of another artificial intelligence as a kind and version of the artificial intelligence,

wherein the determination module resets the driving route of the autonomous driving vehicle based on the error avoidance information so as not to include an error occurrence place associated with an error occurrence cause having a frequency that is above a second predetermined threshold value.

12 . The autonomous driving system installed on an autonomous driving vehicle according to claim 11 , wherein the error response module checks whether the error analysis result message includes autonomous driving operation availability information indicating ‘operation interruption’, and determines that it is necessary to stop the driving of the vehicle when the autonomous driving operation availability information indicating ‘operation interruption’ is included in the error analysis result message.

13 . The autonomous driving system installed on an autonomous driving vehicle according to claim 11 , wherein the error response module determines that it is necessary to stop the driving of the vehicle when the error analysis result message includes a vehicle accident risk and the vehicle accident risk is greater than or equal to the first predetermined threshold value.