KR20210063244A - System for event data recording of autonomous vehicle - Google Patents

Abstract

Disclosed is a system for event data recording of an autonomous vehicle. The present invention proposes an operating method of an EDR system suitable for properly identifying the cause of an accident of a vehicle. A method according to an embodiment of the present invention includes: a step of receiving a trigger signal indicating occurrence of one of a plurality of predefined events from subsystems of an autonomous driving system; a step of collecting event data including at least data related to a recognition-judgment-control process of the autonomous driving system before and after the occurrence of the event; and a step of recording the event data in an internal storage and uploading the event data to a remote server communicatively connected to the vehicle.

Description

SYSTEM FOR EVENT DATA RECORDING OF AUTONOMOUS VEHICLE

The present invention relates to a system for recording event data in an autonomous vehicle.

The content described in this section merely provides background information on the present invention and does not constitute prior art.

Event Data Recorder (EDR) can record, store, and extract driving information or crash information for a certain period of time before and after the accident when a crash or event under certain conditions occurs while driving a vehicle. As a data recording device, it is usually built into the airbag control module (ACU) of the vehicle.

EDR includes vehicle speed, brake operating status, engine RPM, accelerator pedal, throttle valve operating status, steering wheel angle, and seat belt ) Wearing status, Crash Severity (Delta-V or Acceleration), Tire Pressure, Gear Position, Air-bag Deployment Data, etc. are recorded. do.

Based on the information collected from inside and outside the vehicle, the autonomous driving system operates in such a way that the autonomous driving system recognizes the environment, determines (determines) actions for each situation, and controls actuators, etc. If an error occurs in the recognition-judgment-control process in an autonomous vehicle, it can lead to an accident. The record conditions or record items of the EDR system used in general vehicles are not suitable for properly identifying the cause of an accident in an autonomous vehicle. may not be

The present invention proposes an operation method of an EDR system suitable for properly identifying the cause of an accident of an autonomous vehicle.

According to one aspect of the present disclosure, a method for collecting and recording event data by a vehicle (ie, an autonomous vehicle) operating fully or partially in an autonomous driving mode is provided. The method includes receiving a trigger signal instructing occurrence of one of a plurality of predefined events from subsystems of the autonomous driving system, and a recognition-judgment-control process of the autonomous driving system before and after the occurrence of the event. Collecting event data including at least data related to , recording the event data in an internal storage, and uploading the event data to a remote server communicatively connected to the autonomous vehicle.

In some embodiments, the event data includes at least one of a camera image inside/outside the autonomous vehicle, occupant recognition data of the autonomous vehicle, software version information of ECUs mounted in the vehicle, and information related to recently used V2X messages. It may include one more. In some embodiments, the event data may further include object recognition information (eg, object location, class, relative speed, etc.) suitable for reconstructing an external environment map surrounding the autonomous vehicle before and after the occurrence of the event. can In some embodiments, a data item or recording period constituting the event data may be different for at least some of the plurality of events.

In some embodiments, the receiving of the trigger signal includes receiving, from a subsystem that performs an autonomous driving function based at least in part on lane recognition information, a trigger signal indicating that the autonomous vehicle has departed from a lane may include steps. In some embodiments, the receiving of the trigger signal comprises: from a subsystem that performs an autonomous driving function based at least in part on a Time To Collision (TCC), via braking based on a current speed of the vehicle. It may include receiving a trigger signal indicating that collision cannot be avoided. In some embodiments, the receiving of the trigger signal comprises receiving a trigger signal indicating that an Emergency Maneuver (EM) or Minimal Risk Maneuver (MRM) has been initiated from a subsystem of the autonomous driving system. may include steps. In some embodiments, the receiving of the trigger signal may include receiving a trigger signal indicating that an intrusion into the in-vehicle network has been detected from a subsystem of the autonomous driving system.

In some embodiments, when a new trigger signal instructing the occurrence of a new event is received before the end of the recording period of the event data, the recording period is extended while the type of the event that occurred and the occurrence time thereof are additionally recorded It may further include the step of

According to an aspect of the present disclosure, an autonomous driving system that controls a vehicle to fully or partially operate in an autonomous driving mode, a wireless communication system that enables communication between the vehicle and an external system, and EDR that collects and manages event data A vehicle system comprising the system provides. The EDR system (1) receives a trigger signal instructing occurrence of one of a plurality of predefined events from subsystems of the autonomous driving system, and (2) recognition of the autonomous driving system before and after the occurrence of the event - collect event data including at least data related to a decision-control process, (3) record the event data in an internal storage, and upload the event data to a remote server via the wireless communication system.

1 is a functional block diagram depicting a vehicle according to an exemplary embodiment of the present invention;
2A, 2B, and 2C are conceptual diagrams illustrating exemplary architectures in which an EDR system may be deployed in an autonomous vehicle.
3 is a conceptual diagram illustrating a change in parameter D according to a recognized lane geometry and whether or not to depart from a lane.
4 is a flowchart illustrating a method for determining, by the LKAS system, the occurrence of an event to trigger the EDR system, according to an embodiment of the present invention.
5 is a flowchart illustrating a method of determining occurrence of an event to trigger an EDR system based on TTC, according to an embodiment of the present invention.
6A-6D illustrate some example scenarios in which emergency maneuver (EM) occurs in an autonomous driving system.
7 is a flowchart illustrating a method of triggering recording of an EDR system upon initiation of a minimum risk maneuver (MRM), according to an embodiment of the present invention.
8 illustrates an external environment map surrounding an autonomous vehicle.
9 illustrates autonomous driving data to be stored by an EDR system, according to an embodiment of the present invention.

Hereinafter, some embodiments of the present invention will be described in detail through exemplary drawings. In adding reference numerals to elements of each drawing, it should be noted that the same elements are assigned the same numerals as possible, even if they are indicated on different drawings. In addition, in describing the present invention, when it is determined that a detailed description of a related known configuration or function may obscure the subject matter of the present invention, a detailed description thereof will be omitted.

In addition, in describing the constituent elements of the present invention, terms such as first, second, A, B, (a) and (b) may be used. These terms are for distinguishing the constituent element from other constituent elements, and the nature, order, or order of the constituent element is not limited by the term. Throughout the specification, when a part'includes' or'includes' a certain element, it means that other elements may be further included rather than excluding other elements unless otherwise stated. . In addition, the '... Terms such as 'unit' and 'module' mean a unit that processes at least one function or operation, which may be implemented as hardware or software or a combination of hardware and software.

1 is a functional block diagram depicting a vehicle according to an exemplary embodiment of the present invention;

The vehicle is configured to operate fully or partially in an autonomous driving mode and may therefore be referred to as an “autonomous driving vehicle”. For example, the autonomous driving system 120 may receive information from the sensor system 110 and, in an automated manner, one (eg, set steering to avoid a detected obstacle) based on the received information. More than one control process can be executed.

Vehicles may be fully autonomous or partially autonomous. In partially autonomous vehicles, some functions may be temporarily or continuously controlled manually by the driver. Further, the partially autonomous vehicle may be configured to be switchable between a fully manual operating mode and a partially autonomous and/or fully autonomous operating mode.

The vehicle may include various functional systems such as the sensor system 110 , the autonomous driving system 120 , the wireless communication system 130 , and the intrusion detection system 140 . A vehicle may include more or fewer (sub) systems and each (sub) system may include a plurality of components. In addition, (sub)systems of the vehicle can be connected to each other. Accordingly, one or more of the described functionality of a vehicle may be divided into additional functional or physical components or combined into fewer functional or physical components.

The sensor system 110 includes one or more sensors configured to detect information about the environment surrounding the vehicle. For example, the sensor system 110 may include a global positioning system (GPS), a radar unit, a lidar (LIDAR) unit, a camera, an inertial measurement unit (IMU), a microphone, and the like. The sensor system 110 may further include a sensor configured to monitor an internal system of the vehicle (eg, a fuel gauge, engine oil temperature, wheel speed sensor, etc.).

The autonomous driving system 120 may be configured to control the operation of the vehicle and its components. The autonomous driving system 120 may include a recognition subsystem 121 , a determination subsystem 122 , and a control subsystem 123 .

Recognition subsystem 121 provides data captured by the sensor system (eg, images, video) to identify objects and/or features within the environment in which the vehicle is placed, including lane information, traffic signals, other vehicles, pedestrians, obstacles, etc. , depth data, etc.) to be processed and analyzed. Recognition subsystem 121 may use sensor fusion algorithms, object recognition algorithms, video tracking, or other computer vision techniques. The sensor fusion algorithm can provide various evaluations based on data from the sensor system. Evaluation may include evaluation of individual objects and/or features in the environment in which the vehicle is placed, evaluation of specific situations, and/or evaluation of possible effects on driving based on said specific situations.

Based on the various evaluations of the recognition subsystem 121 , the determination subsystem 122 performs actions (eg, lane keeping, lane change, left and right turn, low-speed vehicle overtaking, U-turn, emergency stop, shoulder stop) for each driving situation (scenario) based on the various evaluations of the recognition subsystem 121 . , parking, etc.). The decision subsystem 122 may also plan a route to a destination or plan an obstacle avoidance route.

The control subsystem 123 controls the movement of the vehicle by controlling actuators necessary for the vehicle to travel according to the behavior for each driving situation determined by the determination subsystem 122 . For example, the control subsystem 123 may control a steering unit configured to adjust the steering of the vehicle. As another example, the control subsystem 123 may control a throttle unit that controls the operating speed of the engine and/or motor, thereby controlling the speed of the vehicle. As another example, the control subsystem 123 may control a braking unit configured to slow the vehicle.

Subsystems 121 , 122 , and 123 of the autonomous driving system 120 cooperate with each other to perform various autonomous driving functions or ADAS (Advanced Driver Assistance System) functions, for example, adaptive cruise control (ACC), lane keeping (LKAS). Assist system), Forward Collision-Avoidance Assist (FCA), Autonomous Emergency Braking (AEB), and the like may be supported. The sub-systems 121 , 122 , and 123 of the autonomous driving system 120 may be implemented in a plurality of electronic control units (ECUs) or computer systems in a vehicle that perform one or more autonomous driving functions.

The wireless communication system 130 enables communication between the vehicle and an external system (eg, another vehicle or server, etc.). For example, the wireless communication system 130 may wirelessly communicate with one or more devices directly or through a communication network. The wireless communication system 130 is a cellular communication (eg, GSM, CDMA, LTE, 5G, etc.), IEEE 802.11 communication protocol (eg, WiFi, Bluetooth, ZigBee, etc.), one or a plurality of various wireless communication technologies such as DRSC can be used

An intrusion detection system (IDS) 140 is configured to detect and respond to security threats to the vehicle internal network.

The autonomous driving system 120 , its subsystems 121 , 122 , 123 , as well as other subsystems not shown in FIG. 1 , may generate data indicating whether one or more autonomous driving functions are currently active. For example, the autonomous driving system 120 may generate data indicating whether an adaptive cruise control (ACC) function is currently activated. As another example, the autonomous driving system 120 may generate data indicating whether driving of the vehicle is currently being controlled fully automatically rather than manually.

The vehicle further includes an EDR system 150 configured to receive data from various (sub) systems in the vehicle including the sensor system 110 . One or more (sub) systems may provide data to the EDR system 150 via a data bus (eg, CAN bus, Ethernet bus, etc.). The EDR system 150 may collect data and/or analog signals provided or shared by each (sub) system from the data bus in real time. The EDR system 150 may periodically sample data from (sub) systems. The EDR system 150 may generate a time stamp for each sampled data point.

The one or more (sub) systems detect the occurrence of one or more preset events, and provide trigger signals notifying the occurrence of the event to the EDR system 150, thereby causing the EDR system to generate data related to the event (hereinafter 'EDR data'). ) can be triggered to write to the non-volatile memory. Each trigger signal may include a unique identifier capable of uniquely identifying a related event or trigger condition. For at least some trigger signals, the EDR system may apply different write periods and/or data items to be written.

The vehicle may be configured to upload EDR data to a remote server (not shown) that collects and manages EDR data from multiple vehicles via the wireless communication system 130 . The remote server may be operated by a vehicle manufacturer or may be operated by a service provider providing an EDR data management service.

2A, 2B, and 2C are conceptual diagrams illustrating exemplary architectures in which an EDR system may be deployed in an autonomous vehicle.

As illustrated in FIG. 2A , in the autonomous vehicle, an EDR system specialized for the autonomous vehicle may be embedded as a sub-module in the main controller (integrated ECU) of the autonomous driving system. The EDR system may receive trigger signals notifying the occurrence of an event from an airbag control module (ACM), an intrusion detection system (IDS), etc. located outside the main controller as well as the main controller. In addition, the autonomous vehicle may further include an EDR system that is built in as a sub-module in the airbag controller (ACU) and operates in the same manner as a conventional typical EDR system.

As illustrated in FIG. 2B , the EDR system in the autonomous vehicle may be embedded as a sub-module in the airbag controller (ACU), like a conventional typical EDR system. The EDR system may receive trigger signals notifying the occurrence of an event from the airbag controller (ACU) as well as the main controller of the autonomous driving system located outside the airbag controller (ACU), the IDS system, and the like.

As illustrated in FIG. 2C , the autonomous vehicle may be equipped with an independent electronic control unit (ECU) that performs an EDR function. The EDR system may receive trigger signals notifying the occurrence of an event from an airbag controller (ACU) on a vehicle network, a main controller of an autonomous driving system, an IDS system, and the like.

Diversification of EDR trigger conditions

A conventional EDR system is configured to record predefined data items upon receiving an event trigger signal from an Airbag Control Unit (ACU). Such an event may in particular be a traffic collision. A crash may be detected, for example, when deployment of an irreversible safety device such as an airbag or pretensioner is triggered. A crash accident may also be detected when an acceleration/deceleration exceeding a predefined threshold (eg, a speed change of 8 km/h or more within 150 ms) occurs. These trigger conditions used in general vehicles may not be suitable for properly identifying the cause of an accident in an autonomous vehicle. The present disclosure presents various event trigger conditions suitable for an autonomous vehicle and data items suitable for reconstructing the environment inside and outside the vehicle surrounding the events. According to the present disclosure, subsystems of the autonomous driving control system may determine whether an event satisfying a preset event condition has occurred, and when the occurrence of the event is detected, transmit an event trigger signal to the EDR system.

(1) Lane departure

In a situation where the Lane Keeping Assist System (LKAS) among the ADAS (Advance Driver Assistance System) technology is operating, lane recognition is correct, but when it is determined that the vehicle has deviated significantly from the lane, the EDR system record is triggered can be

The computer vision system may recognize a lane by analyzing image data output from the camera sensor, and may provide lane recognition information to the lane keeping system. The lane recognition information includes not only a distance to a lane marking, but also lane geometry parameters such as a heading angle and a curvature. Lane recognition information may be provided as a road equation in the form of a cubic function (y = Ax^3 + Bx^2 + Cx + D) describing the relative motion of the vehicle and the lane. Here, A is a curvature rate, B is a curvature, C is a heading angle, and D is a lateral offset. Also, the lane recognition information may include a quality signal indicating reliability of lane geometry measurement. The quality signal may be divided into a total of four levels (Very Low Quality, Low Quality, High Quality, and Very High Quality).

3 is a conceptual diagram illustrating a change in a parameter D (= 0, 0.5) according to a recognized lane geometry and whether or not to depart from a lane. According to an embodiment of the present invention, the degree of lane departure is determined using the size of the parameter (D) included in lane recognition information indicating the lateral distance between the vehicle and the lane, and when excessive lane departure is detected A trigger signal can be generated that triggers the recording of the EDR system.

4 is a flowchart illustrating a method for determining, by the LKAS system, the occurrence of an event to trigger the EDR system, according to an embodiment of the present invention.

The LKAS system receives lane recognition information from the computer vision system while driving (S410). The LKAS system may determine whether the quality signal indicating the reliability of the lane geometry measurement included in the lane recognition information is equal to or greater than a preset level (eg, 'High Quality') ( S420 ).

In the LKAS system, if the quality signal is higher than or equal to a preset level (eg, 'High Quality') ('Yes' in S420), the parameter (D) indicating the lateral distance between the vehicle and the lane included in the lane recognition information is It is determined whether or not a preset threshold value (thr) is exceeded ( S430 ). The LKAS system generates a trigger signal that triggers the recording of the EDR system when the parameter D exceeds a preset threshold ('Yes' in S430) (S450).

LKAS system, if the quality signal is less than a preset level (eg, 'High Quality') ('No' in S420), sensor fusion, that is, different sensors (GPS, camera sensor, radar sensor, lidar sensor, etc.) It is possible to determine whether or not to deviate from the lane by using the data obtained from (S440). For example, mapping vehicle location information to a precision map, using data obtained from a vehicle SVM (surround view monitor) camera sensor, or using a radar sensor or lidar sensor It is also possible to determine a lane and determine whether to deviate from the lane based on the virtual lane. When a lane departure is detected based on the sensor fusion ('Yes' in S440), the LKAS system generates a trigger signal that triggers the recording of the EDR system (S450).

(2) Time To Collision (TTC)

Time to Collision (TTC) is one of the most widely used metrics when judging the possibility of collision with an obstacle (vehicle or pedestrian) in front or behind. TTC is used not only for the Forward Collision-Avoidance Assist (FCA) system, but also to measure the risk of collision with obstacles around the vehicle. For example, if there is a short-distance cut-in vehicle, the TTC may be a criterion for determining whether the host vehicle accelerates or decelerates.

TTC is generally defined as 'relative distance/relative speed', and the acceleration of the vehicle is also taken into account to more accurately measure the risk of a collision. The autonomous driving system can calculate the TTC by estimating the relative distance and relative speed of the obstacle ahead through the radar sensor, camera sensor, lidar sensor, ultrasonic sensor, or their data fusion. Alternatively, a more accurate TTC calculation may be made using the position and speed of the surrounding vehicle obtained through V2V communication. If the steering angle or brake state value of the surrounding vehicle is directly acquired through V2V communication, the motion of the surrounding vehicle will be able to be predicted much earlier than the environmental sensor. In addition, if the V2V communication is used, it is possible to obtain information about the surrounding information vehicle even when the field of view is not secured.

FCA or similar collision mitigation systems are primarily aimed at mitigating damage from collisions. That is, even if a collision occurs, the basic concept is to reduce the damage by reducing the collision energy as much as possible through pre-braking. In the case of low-speed driving, emergency braking is performed just before the collision to stop the vehicle immediately before the accident occurs, but at medium speeds or higher, where the inertia is relatively high, it may be difficult to avoid the collision itself even if the FCA operates normally. However, even if a collision occurs, since the emergency braking is already performed immediately before the collision, a significant portion of the speed energy is offset in the actual collision situation, so the damage caused by the accident can be reduced by that much.

According to an embodiment of the present invention, when it is determined that a collision cannot be avoided even if the TTC is very short or emergency braking is performed based on the current speed, the recording of the EDR system may be triggered before the actual collision occurs.

5 is a flowchart illustrating a method of determining occurrence of an event to trigger an EDR system based on TTC, according to an embodiment of the present invention.

A subsystem of the autonomous driving system (eg, FCA or similar collision mitigation system) may calculate a TTC ( S510 ) and determine whether the calculated TTC is extremely short ( S520 ). For example, the subsystem may determine whether the calculated TTC is shorter than a preset threshold (eg, 2sec).

When it is determined that the calculated TTC is extremely short (ie, if the calculated TTC is shorter than a preset threshold value; 'Yes' in S520), the subsystem immediately generates a trigger signal that triggers the recording of the EDR system (S550). At the same time, the subsystem may control the braking system to perform emergency braking.

If the calculated TTC is not shorter than the preset threshold value ('No' in S520), the subsystem determines whether a collision is unavoidable by performing braking based on the current speed of the vehicle (S530). For example, the subsystem may determine that the collision cannot be avoided if the deceleration required to avoid the collision exceeds a predetermined maximum deceleration (limit deceleration) of the vehicle. Alternatively, when it is determined that the current speed of the vehicle exceeds the threshold speed, the subsystem may determine that a collision cannot be avoided. When it is determined that a collision cannot be avoided by performing braking based on the current speed of the vehicle ('Yes' in S530), the subsystem generates a trigger signal for triggering recording of the EDR system (S550).

If the deceleration required to avoid a collision does not exceed the vehicle's predetermined maximum deceleration (limit deceleration) ('No' in S530), the subsystem, until the next TTC is determined, A comparison between the limit decelerations of the vehicle (corresponding to the changed vehicle speed) may be repeatedly performed ( S540 ).

(3) Emergency Maneuver (EM) is driven

Emergency Manoeuvr (EM) is a maneuver performed by an autonomous driving system in the event of a sudden and unexpected event in which a vehicle is in danger of colliding with another object, with the aim of avoiding or mitigating a collision.

The autonomous driving system can detect whether a vehicle is in imminent danger of a collision, eg, with a road user in front or on the side of the vehicle, due to a sudden and unexpected event. If the time is too short for the driver to safely switch back control, Emergency Maneuvering (EM) is automatically initiated. To avoid or mitigate an impending collision, Emergency Maneuvering (EM) may perform protective deceleration up to the vehicle's maximum braking performance or perform automatic avoidance maneuvers. The autonomous driving system may generate a trigger signal that triggers the recording of the EDR system when an emergency maneuver (EM) is initiated.

6A-6D illustrate some example scenarios in which emergency maneuver (EM) occurs in an autonomous driving system.

- Impossible to control inter-vehicle distance maintenance (no longitudinal control; see Fig. 6a): Emergency braking function of the Forward Collision-Avoidance Assist (FCA) system when the risk of collision is expected during lane keeping and inter-vehicle distance maintenance control This can work. However, when the risk of collision between the own vehicle and the preceding vehicle is expected because the preceding vehicle is suddenly stopped due to sudden braking or collision, emergency braking can be performed up to the maximum braking performance of the vehicle in a shorter time than emergency braking of the FCA system.

- Unable to control lane keeping (lateral control impossible; see Fig. 6b): Due to a defect in Motor-Driven Power Steering (MDPS), etc., it is expected that the vehicle will deviate from the current driving lane and collide with a vehicle located in the neighboring lane. In this case, emergency braking may be performed.

- Engine off while driving (longitudinal and lateral direction control impossible; see FIG. 6c): When the engine is turned off during lane maintenance and inter-vehicle distance maintenance control, emergency braking may be performed immediately.

- Dangerous situation after avoidance of a forward obstacle (refer to Fig. 6d): When an abnormality occurs in the vehicle's behavior after avoidance maneuvering for an obstacle in front (eg, when it is impossible to deviate from the lane or return to the driving lane), emergency braking is performed immediately can be

(4) MRM (Minimum Risk Maneuver) is running

When an autonomous vehicle malfunctions, such as a defect in an autonomous driving sensor, an intrusion detection system (IDS) detects an intrusion into the vehicle's internal network, or the driver's manual control is required for other reasons, the autonomous driving system You can ask the driver to transfer control.

Minimal Risk Manoeuvr (MRM) means a procedure aimed at minimizing the traffic risk performed automatically by the autonomous driving system, for example when the driver does not respond to a transition demand .

The autonomous driving system may detect whether the driver has resumed manual control after issuing a switch request. If the driver does not resume manual control within a certain period of time, MRM can be initiated immediately. MRM can be terminated when the vehicle detects that the driver has taken over manual control of the vehicle. MRM may automatically perform one or more of the following actions: (a) slowing the vehicle within its lane, with a deceleration of 4 m/s² or less; (b) Stopping the vehicle in a lane other than the high-speed lane (eg, the low-speed lane, emergency lane, or roadside).

Initiation of MRM may trigger recording of the EDR system. MRM control may be operated in various ways depending on the cause case, and sometimes a situation with a long event duration may occur. Depending on the cause of the initiation of MRM, the recording method of the EDR system may also be different.

In some embodiments, when MRM is initiated because the driver does not respond to a transition demand, the EDR system may continuously store information about the awareness-judgment-control process of the autonomous driving system while the MRM is performed. have.

In some embodiments, when MRM is initiated due to a fault of the autonomous driving sensor, the identifier (ID) of the faulty sensor and the time of failure are stored, while recording of the EDR system until the user receives driving control. ) can be driven.

In some embodiments, from the moment when MRM is initiated, a threshold value of a determination criterion or trigger condition for determining the occurrence of an event triggering recording of the EDR system may be lowered, thereby increasing the sensitivity to the event. For example, if a sudden acceleration/deceleration showing a speed change of 8 km/h or more is detected within 150 ms before MRM starts, the recording of the EDR system is triggered, but during MRM, acceleration/deceleration showing a speed change of 4 km/h or more is detected within 150 ms. This can trigger a recording in the EDR system.

7 is a flowchart illustrating a method of triggering recording of an EDR system according to initiation of MRM, according to an embodiment of the present invention.

When the occurrence of a situation causing MRM initiation is detected (S710), the autonomous driving system may estimate the duration of MRM control based on the situation causing MRM initiation (S720). If the estimated duration is shorter than the preset threshold ('short' in S710), the driving system may immediately trigger the recording of the EDR system (S750).

If the estimated duration is longer than the preset threshold ('long' in S710), the driving system lowers the threshold of the decision criterion or trigger condition for determining the occurrence of an event triggering the recording of the EDR system. It can be (S730). Thereafter, the autonomous driving system may monitor the occurrence of an event triggering the recording of the EDR system by applying the relaxed determination criterion or threshold of the trigger condition ( S740 ). When the occurrence of an event satisfying the relaxed determination criterion is detected ('Yes' in S740), the driving system may trigger recording of the EDR system (S750).

On the other hand, when using various trigger signals for triggering recording of the EDR system, a new trigger signal may be generated before data recording of the EDR system according to any one trigger signal is terminated. In this way, when a plurality of trigger signals are continuously generated, the EDR system can extend the data recording time and store the type of the event (or trigger signal) that has occurred and the time of occurrence of the event. The trigger signal may include an identifier that uniquely identifies the type of event. In this case, the data items recorded during the extended time period may be the same as or less than the data items recorded in the time period corresponding to the initial trigger signal.

(5) When a security threat to the vehicle network is detected

The vehicle may include an Intrusion Detection System (IDS) configured to detect and respond to security threats to the vehicle's internal network. When the intrusion detection system detects an intrusion into the in-vehicle network, the intrusion detection system may generate a trigger signal that triggers the recording of the EDR system.

As such, when the EDR system is configured to receive various trigger signals, the EDR system may apply data items to be recorded and/or recording periods differently to at least some trigger signals.

EDR data items suitable for autonomous vehicles

The self-driving vehicle operates in such a way that the autonomous driving system recognizes the environment, determines (determines) actions in each situation, and controls the actuators based on information collected from inside and outside the vehicle, so recognition-judgment-control If an error occurs in the process, it can lead to an accident. Therefore, in order to properly identify the cause of an accident of an autonomous vehicle, it is desirable to record and store information on the recognition-judgment-control process of the autonomous vehicle.

The present invention proposes EDR data items suitable for identifying the cause of an event occurring in an autonomous vehicle. In particular, these data items may be useful for reconstructing an external environment map (as shown in FIG. 8 ) surrounding the autonomous vehicle at the time of occurrence of the event and reconstructing the internal environment of the vehicle.

EDR data according to an embodiment of the present invention may include software version, autonomous driving data, camera images inside/outside the vehicle, occupant recognition data, V2X message, and the like.

The software version indicates the software version of each electronic control unit (ECU) mounted on the vehicle. The occupant recognition data may indicate the driver's state (eg, distraction, drowsiness, non-responsive state), the presence or absence of a passenger, and the like. In addition, the EDR data may include information about recent V2X messages exchanged with the infrastructure and surrounding vehicles through V2X communication.

As illustrated in FIG. 9 , the autonomous driving data includes <cognition data>, <judgment data>, and <control data>.

<cognitive data> of autonomous driving data may include the following information.

- Detected obstacle classification : Includes information on recognized dynamic obstacles that can cause accidents during vehicle operation (eg, classes of objects such as pedestrians, bicycles, two-wheeled vehicles, passenger cars, and buses).

- Lane recognition information : Includes lane geometry parameters (curvature, curvature rate, heading angle, lateral offset) output from the computer vision system and a quality indicator indicating reliability of lane geometry measurement.

- Vehicle location information and positioning technology used: Information on the positioning technology used includes, for example, whether GPS, V2X, precision map use, and the version of the precision map used.

- Estimated perceived collision time (Time To Collision; TTC)

The <judgment data> of the autonomous driving data may include the following information.

- Whether autonomous driving functions (LKA, FCA, etc.) are activated

- Object data analyzed after recognition of the surroundings: Includes class of recognized objects, x,y coordinates/size/relative speed (using sensor fusion), moving speed and direction of own vehicle. These data can be useful for reconstructing a map of the external environment surrounding the autonomous vehicle at the time of the event.

- Behavioral scenario of the determined driving situation (lateral direction, longitudinal scenario): e.g., lane maintenance, lane change, left/right turn, U-turn, emergency stop, shoulder stop, parking, close vehicle cut-in, front vehicle Cut-out, etc.

The <control data> of the autonomous driving data may include the following information.

- Driver control information : Actuator control information such as steering wheel torque and acceleration/deceleration pedal

- Autonomous driving control information : actuator control information such as steering wheel torque and acceleration/deceleration pedal

It should be understood that the above-described exemplary embodiments may be implemented in many different ways. In some examples, the various methods, apparatuses, systems, and subsystems described in this disclosure may include a processor, memory, disk or other mass storage, communication interface, input/output (I/O) devices, and other peripherals. The branch may be implemented by a general purpose computer. A general purpose computer can function as an apparatus for executing the method described above by loading software instructions into a processor and then executing the instructions to perform the functions described in this disclosure.

It should be understood that the functional components described herein are labeled as ... units or modules to particularly further emphasize their implementation independence. For example, a module may be implemented as a custom very-large-scale integration (VLSI) circuit or hardware circuit including semiconductors such as gate arrays, logic chips, and transistors. A module may also be implemented as a programmable hardware device, such as a field programmable gate array (FPGA), programmable array logic, programmable logic devices, and the like.

Meanwhile, various methods described in the present disclosure may be implemented with instructions stored in a non-transitory recording medium that can be read and executed by one or more processors. The non-transitory recording medium includes, for example, any type of recording device in which data is stored in a form readable by a computer system. For example, the non-transitory recording medium includes a storage medium such as an erasable programmable read only memory (EPROM), a flash drive, an optical drive, a magnetic hard drive, and a solid state drive (SSD).

The above description is merely illustrative of the technical idea of the present embodiment, and those of ordinary skill in the technical field to which the present embodiment pertains will be able to make various modifications and variations without departing from the essential characteristics of the present embodiment. Accordingly, the present embodiments are not intended to limit the technical idea of the present embodiment, but to explain the technical idea, and the scope of the technical idea of the present embodiment is not limited by these embodiments. The scope of protection of this embodiment should be interpreted by the following claims, and all technical ideas within the scope equivalent thereto should be construed as being included in the scope of the present embodiment.

Claims (18)

A method for collecting and recording event data by a vehicle operating fully or partially in an autonomous driving mode, the method comprising:
receiving a trigger signal indicating occurrence of one of a plurality of predefined events from subsystems of the autonomous driving system;
collecting event data including at least data related to a recognition-judgment-control process of the autonomous driving system before and after the occurrence of the event; and
recording the event data in an internal storage and uploading the event data to a remote server communicatively connected to the vehicle
Including, the method. The method of claim 1,
The event data is
The method further comprising at least one of a camera image inside/outside the vehicle, occupant recognition data of the vehicle, software version information of ECUs mounted in the vehicle, and recently used V2X message related information. The method of claim 1,
The event data is
and object recognition information suitable for reconstructing an external environment map surrounding the vehicle before and after the occurrence of the event. The method of claim 1,
and a data item or recording period constituting the event data is different for at least some of the plurality of events. The method of claim 1,
The event data is
A method, characterized in that it is recorded and uploaded together with information indicative of the relevant event or trigger condition. The method of claim 1,
Receiving the trigger signal comprises:
and receiving, from a subsystem that performs an autonomous driving function based at least in part on lane recognition information, a trigger signal indicating that the vehicle has departed from a lane. The method of claim 1,
Receiving the trigger signal comprises:
Receiving a trigger signal indicating that a collision cannot be avoided through braking based on the current speed of the vehicle from a subsystem that performs an autonomous driving function based at least in part on a Time To Collision (TCC) A method comprising: The method of claim 1,
Receiving the trigger signal comprises:
and receiving a trigger signal indicating that Emergency Maneuver (EM) or Minimal Risk Maneuver (MRM) has been initiated from a subsystem of the autonomous driving system. The method of claim 1,
Receiving the trigger signal comprises:
and receiving, from a subsystem of the autonomous driving system, a trigger signal indicating that an intrusion into the in-vehicle network has been detected. The method of claim 1,
When a new trigger signal instructing occurrence of a new event is received before the end of the recording period of the event data, extending the recording period and additionally recording the type of the event and the occurrence time thereof A method, characterized in that A vehicle system comprising:
an autonomous driving system that controls the vehicle to fully or partially operate in an autonomous driving mode;
a wireless communication system that enables communication between the vehicle and an external system; and
EDR system that collects and manages event data
Including,
The EDR system,
(1) receiving a trigger signal instructing occurrence of one of a plurality of predefined events from subsystems of the autonomous driving system, and (2) recognition-judgment-control of the autonomous driving system before and after the occurrence of the event and collecting event data comprising at least data related to the process, (3) recording the event data in an internal storage, and uploading the event data to a remote server via the wireless communication system. . The method of claim 11,
The event data is
A method, characterized in that it is recorded and transmitted together with information indicative of a relevant event or trigger condition. The method of claim 11,
The trigger signal is
and a trigger signal indicating that the vehicle has departed from a lane, received from a subsystem that performs an autonomous driving function based at least in part on lane recognition information. The method of claim 11,
The trigger signal is
and a trigger signal indicating that a collision cannot be avoided through braking based on the current speed of the vehicle, received from a subsystem performing an autonomous driving function based at least in part on a Time To Collision (TCC). A vehicle system, characterized in that The method of claim 11,
The trigger signal is
and a trigger signal indicating that Emergency Maneuver (EM) or Minimal Risk Maneuver (MRM) has been initiated from the subsystem of the autonomous driving system. The method of claim 11,
The trigger signal is
and a trigger signal indicating that an intrusion into the vehicle internal network has been detected from the subsystem of the autonomous driving system. The method of claim 11,
The EDR system,
When a new trigger signal instructing occurrence of a new event is received before the end of the recording period of the event data, extending the recording period and additionally recording the type of the event and the occurrence time thereof A vehicle system, characterized in that The method of claim 11,
The EDR system,
A vehicle system, characterized in that it is embedded as a function module in the main controller or airbag control unit (ACU) of the autonomous driving system, or is connected to the vehicle internal network as a separate electronic control unit (Electronic Control Unit).

Images