Autonomous driving system
An autonomous driving system includes an external sensor, an internal sensor, a machine learning model configured to receive a detection result of the external sensor and the internal sensor as input values and output an instruction value of autonomous driving, an operation unit configured to perform the autonomous driving based on the instruction value, a user interface, and a control unit, wherein the control unit includes, determining at least one risk candidate scene, and extracting the input value and the instruction value in the at least one risk candidate scene, and causes the user interface to notify the occupant of monitoring information including information indicating that the input value and the instruction value have been extracted.
1 . An autonomous driving system comprising:
an external sensor configured to detect an external environment of a vehicle;
an internal sensor configured to detect a traveling state of the vehicle;
a machine learning model configured to receive a detection result of the external sensor and the detection result of the internal sensor as input values and output an instruction value of autonomous driving;
an operation unit configured to perform the autonomous driving of the vehicle based on the instruction value output by the machine learning model;
a user interface configured to exchange information with an occupant of the vehicle;
a transmission unit; and
a control unit connected to the user interface,
wherein the control unit is configured to:
determine at least one risk candidate scene based on at least one selected from a group of the detection result of the external sensor, the detection result of the internal sensor, and an operating state of an occupant of the vehicle;
extract at least one of the input values and the instruction value of the machine learning model in the at least one risk candidate scene;
cause the user interface to notify the occupant of the vehicle of monitoring information including information related to the at least one of the input values and the instruction value; and
cause the transmission unit to transmit the monitoring information to a device outside of the vehicle;
wherein the monitoring information includes request information for selecting whether or not to perform transmission to the device outside of the vehicle; and
wherein the control unit is configured to cause the transmission unit to transmit extraction data, which is the at least one extracted input value and the extracted instruction value of the machine learning model, to the device outside of the vehicle in response to reception of a response permitting transmission to the device outside of the vehicle via the user interface.
2 . The autonomous driving system according to claim 1 , wherein the control unit is configured to present the at least one risk candidate scene to the occupant of the vehicle through the user interface, and causes the transmission unit to transmit the extraction data corresponding to the risk candidate scene selected by the occupant of the vehicle through the user interface.
3 . The autonomous driving system according to claim 2 , wherein the control unit is configured to receive a ranking of the at least one risk candidate scene designated by the occupant of the vehicle through the user interface, and determine the extraction data to be transmitted by the transmission unit based on the ranking.
4 . The autonomous driving system according to claim 2 , wherein
the control unit is configured to:
calculate a similarity between risk candidate scenes based on the at least one of the input values and the instruction value in the risk candidate scenes;
classify the risk candidate scenes into categories based on the similarity calculated; and
prevent the risk candidate scenes to the occupant of the vehicle for each of the categories classified, through the user interface.
5 . The autonomous driving system according to claim 2 , wherein the control unit is configured to extract the data such that extraction data corresponding to the risk candidate scene instructed from the occupant of the vehicle through the user interface includes more information than extraction data corresponding to the risk candidate scene non-instructed.