Traffic safety assistance system and learning method thereof

Abstract

A traffic safety assistance system is provided to improve the safety, convenience, and smoothness of traffic for multiple traffic participants in a target traffic area. To address the aforementioned problems, the traffic safety assistance system includes: a target traffic area identification unit, which acquires identification information related to traffic participants in the target traffic area; a prediction unit (62), which predicts the risk of the target traffic area based on the identification information; and a coordination assistance information notification unit (65), which sends coordination assistance information to the assisted target. The prediction unit (62) includes: a regional risk prediction unit (620), which extracts high-risk areas from multiple local areas after subdividing the target traffic area based on information obtained through statistical processing of the identification information; and a traffic participant risk prediction unit (625), which predicts the future risk of traffic participants in the high-risk area based on information related to the high-risk area in the identification information.

Description

Technical Field

[0001] This invention relates to a traffic safety assistance system and its learning method. More specifically, it relates to a traffic safety assistance system and its learning method for assisting traffic participants as people or moving bodies in safe movement. Background Technology

[0002] In public transportation, various traffic participants, including four-wheeled vehicles, motorcycles, bicycles, and pedestrians, move at different speeds according to their own wishes. As a technology for improving the safety and convenience of traffic participants in such public transportation, for example, Patent Document 1 discloses a driving assistance device to assist the driver of a vehicle in safe driving.

[0003] The driving assistance device shown in Patent Document 1 includes: a hazard prediction unit that predicts the degree of danger of the vehicle based on information about the vehicle's driving status and the surrounding environment; and a warning control unit that, based on the evaluation result of the predicted hazard, issues a warning to the driver via voice or text display. According to the driving assistance device shown in Patent Document 1, when certain hazards are predicted, the driver can be prompted to perform driving operations to avoid the predicted hazards, thus assisting the driver in safe driving.

[0004] [Previous Technical Documents]

[0005] (Patent Documents)

[0006] Patent Document 1: Japanese Patent Application Publication No. 2021-136001 Summary of the Invention

[0007] [The problem the invention aims to solve]

[0008] However, in the invention shown in Patent Document 1, since the degree of danger is predicted based on information about the surrounding environment obtained by onboard sensors such as cameras or radar mounted on the vehicle, it is impossible to grasp potential risks that exist outside the detection range of the onboard sensors.

[0009] Therefore, in order to provide appropriate assistance in addressing such potential risks, for example, it is possible to aggregate information related to traffic participants in a designated target traffic area onto a server that is communicatively connected to each traffic participant, and use the server to have an overview of the movement of traffic participants in the target traffic area.

[0010] However, when a large number of traffic participants existing in an object traffic area are aggregated on the server, the processing load on the server will become correspondingly high, and therefore it may become impossible to provide appropriate assistance to each traffic participant in the object traffic area in real time.

[0011] The purpose of this invention is to provide a traffic safety assistance system that can improve the safety, convenience, and smoothness of traffic for multiple traffic participants in a target traffic area.

[0012] [Technical means to solve the problem]

[0013] (1) The traffic safety assistance system of the present invention is characterized by comprising: an identification means for identifying traffic participants, including persons or moving bodies, and traffic environments of each traffic participant, within a target traffic area, and acquiring identification information related to these identified objects; a prediction means for predicting risks in the target traffic area based on the aforementioned identification information; and a transmission means for transmitting auxiliary information generated based on the aforementioned identification information and the prediction results of the aforementioned prediction means to an assistance object determined from a plurality of traffic participants in the target traffic area; and the aforementioned prediction means comprising: a regional risk prediction means for extracting at least one of a plurality of local areas after subdividing the target traffic area as a high-risk area based on information obtained by statistical processing of the aforementioned identification information; and a traffic participant risk prediction means for predicting future risks of traffic participants in the aforementioned high-risk area based on information associated with the aforementioned high-risk area in the aforementioned identification information.

[0014] (2) Preferably, the aforementioned traffic participant risk prediction method does not perform the aforementioned prediction process on the local areas among the multiple aforementioned local areas that were not extracted as the aforementioned high-risk areas by the aforementioned regional risk prediction method.

[0015] (3) Preferably, the aforementioned regional risk prediction method estimates the risk level for each of the aforementioned local areas, and the aforementioned sending method sends first auxiliary information generated based on the prediction results of the aforementioned traffic participant risk prediction method to the auxiliary objects that exist in the aforementioned high-risk areas among the multiple aforementioned auxiliary objects, and sends second auxiliary information generated based on the estimation results of the aforementioned regional risk prediction method to the auxiliary objects that exist in the low-risk areas outside the aforementioned high-risk areas.

[0016] (4) The learning method of the traffic safety assistance system of the present invention is the learning method of the traffic safety assistance system according to any one of (1) to (3), characterized in that the aforementioned regional risk prediction means extracts the aforementioned high-risk areas by using a macro risk estimation model, and the macro risk estimation model outputs the risk level of each of the aforementioned local areas when it inputs information obtained by statistical processing of the aforementioned identification information; the aforementioned traffic participant risk prediction means predicts the future risk of traffic participants in the aforementioned high-risk areas by using a micro risk estimation model, and the micro risk estimation model outputs the future risk of traffic participants in the local area when it inputs information related to a specified local area from the aforementioned identification information; the learning method of the traffic safety assistance system includes the following steps: preparing learning data by using input data for the aforementioned macro risk estimation model generated based on the aforementioned identification information and the output of the micro risk estimation model when the aforementioned identification information is input to the aforementioned micro risk estimation model; and using the aforementioned learning data to learn the aforementioned macro risk estimation model.

[0017] (5) The learning method of the traffic safety assistance system of the present invention is the learning method of the traffic safety assistance system according to any one of (1) to (3), characterized in that the aforementioned regional risk prediction means extracts the aforementioned high-risk areas by using a macro risk estimation model, and the macro risk estimation model outputs the risk level of each of the aforementioned local areas when it is input with information obtained by statistical processing of the aforementioned identification information; the aforementioned traffic participant risk prediction means predicts the future risk of traffic participants in the aforementioned high-risk areas by using a micro risk estimation model, and the micro risk estimation model outputs the future risk of traffic participants in the local area when it is input with information related to a specified local area from the aforementioned identification information; the learning method of the traffic safety assistance system includes the following steps: preparing learning data by using input data for the aforementioned macro risk estimation model generated based on first identification information obtained in a specified first period, and positive solution data for the output of the aforementioned micro risk estimation model generated based on second identification information obtained in a second period after the aforementioned first period; and using the aforementioned learning data to learn the overall model that combines the aforementioned macro risk estimation model and the aforementioned micro risk estimation model.

[0018] (The effect of the invention)

[0019] (1) The traffic safety assistance system of the present invention includes: an identification means for identifying traffic participants (including people and moving bodies) and their traffic environments within a target traffic area, and acquiring identification information related to these identified objects; a prediction means for predicting risks in the target traffic area based on the identification information; and a transmission means for transmitting auxiliary information generated based on the identification information and the prediction results of the prediction means to an assistance object determined from multiple traffic participants in the target traffic area. Furthermore, in the prediction means, at least one of multiple subdivided local areas of the target traffic area is extracted as a high-risk area using a regional risk prediction means, and the future risks of traffic participants in the high-risk area are predicted using a traffic participant risk prediction means. Here, in the regional risk prediction means, by using information obtained through statistical processing of the identification information when extracting high-risk areas from multiple local areas, high-risk areas can be extracted with less overhead compared to directly utilizing a large amount of identification information related to the identified objects in the target traffic area. Furthermore, in traffic participant risk prediction methods, by using information associated with the high-risk area from identification information related to the identified objects in the entire target traffic area when predicting the future risk of traffic participants, compared to directly using a large amount of identification information related to the identified objects in the target traffic area, the future risk of traffic participants can be predicted with a less load. Therefore, according to the invention, appropriate auxiliary information generated based on the prediction results can be provided to traffic participants in high-risk areas in real time, thus improving the safety, convenience, and smoothness of traffic in the target traffic area.

[0020] (2) In this invention, the traffic participant risk prediction method does not predict the future risk of traffic participants in other local areas that were not identified as high-risk areas by the regional risk prediction method. Therefore, according to this invention, the computational load can be reduced compared to predicting all local areas. Furthermore, according to this invention, the computational load can be reduced by decreasing the number of local areas to be predicted, thus improving the accuracy of risk prediction for traffic participants in high-risk areas. Therefore, according to this invention, appropriate auxiliary information generated based on highly accurate prediction results can be provided to traffic participants in high-risk areas in real time, further improving the safety, convenience, and smoothness of traffic in the target traffic area.

[0021] (3) In this invention, the regional risk prediction method estimates the risk level for each local area and extracts high-risk areas from multiple local areas based on the estimated risk level. Furthermore, the sending method sends first auxiliary information generated based on a more detailed prediction result from the traffic participant risk prediction method to auxiliary objects existing in high-risk areas among multiple auxiliary objects in the entire target traffic area. This improves the safety, convenience, and smoothness of traffic for traffic participants in high-risk areas. Additionally, the sending method sends second auxiliary information generated based on the estimated result for each local area by the regional risk prediction method to auxiliary objects existing in low-risk areas outside of high-risk areas among multiple auxiliary objects in the entire target traffic area. This also improves the safety, convenience, and smoothness of traffic for traffic participants in low-risk areas. Thus, in this invention, by changing the auxiliary information according to the risk level of each local area, the safety, convenience, and smoothness of traffic for traffic participants in the entire target traffic area can be improved.

[0022] (4) In the learning method of the traffic safety assistance system of the present invention, learning data is prepared by using input data for a macro-risk estimation model generated based on identification information, and the output of the micro-risk estimation model when the identification information is input into the micro-risk estimation model. This learning data is then used to learn the macro-risk estimation model. In general model learning, positive solution data is required to evaluate the correctness of the model's output. In contrast, in the present invention, the output of the micro-risk estimation model can be used as the learning data for learning the macro-risk estimation model, thus allowing for the construction of a high-precision macro-risk estimation model in a relatively simple way. Therefore, according to the present invention, the accuracy of the macro-risk estimation model can be improved by providing assistance information to various traffic participants.

[0023] (5) In the learning method of the traffic safety assistance system of the present invention, learning data is prepared by using input data for a macro-risk estimation model generated based on first identification information obtained in a first period, and positive solution data for the output of a micro-risk estimation model generated based on second identification information obtained in a second period after the first period. This learning data is then used to learn an overall model combining the macro-risk estimation model and the micro-risk estimation model. Therefore, according to the present invention, the second identification information obtained in the second period after the first period can be used as data to evaluate the correctness of the output of the overall model when the first identification information is input, thus improving the accuracy of the overall model combining the macro-risk estimation model and the micro-risk estimation model. Therefore, according to the present invention, the accuracy of the overall model can be improved by providing assistance information to each traffic participant. Attached Figure Description

[0024] Figure 1 This is a diagram illustrating the structure of a traffic safety assistance system according to an embodiment of the present invention and the assistance object of the traffic safety assistance system, namely a part of the object traffic area.

[0025] Figure 2 It is a block diagram illustrating the structure of a coordination aid and multiple regional terminals communicatively connected to the coordination aid.

[0026] Figure 3A It is a block diagram illustrating the structure of a notification device mounted on a four-wheeled vehicle.

[0027] Figure 3B This is a block diagram illustrating the structure of a notification device mounted on a motorcycle.

[0028] Figure 3C This is a block diagram illustrating the structure of a notification device mounted on a portable information processing terminal owned by a pedestrian.

[0029] Figure 4 It is a functional block diagram illustrating the specific structure of the prediction unit.

[0030] Figure 5 This is a diagram that schematically illustrates the concept of risk notification optimization in the risk notification setting unit. Detailed Implementation

[0031] Hereinafter, a traffic safety assistance system according to an embodiment of the present invention will be described with reference to the drawings.

[0032] Figure 1 This is a schematic diagram illustrating the structure of the traffic safety assistance system 1 of this embodiment and a part of the traffic area 9 where the traffic participants are located, which is the assisted object of the traffic safety assistance system 1.

[0033] The traffic safety assistance system 1 identifies pedestrians 4, four-wheeled vehicles 2, and motorcycles 3, which are moving objects, as traffic participants in the target traffic area 9, and notifies each traffic participant of the assistance information generated through this identification. This facilitates communication between traffic participants moving according to their own will (specifically, mutual identification between traffic participants) and the identification of the surrounding traffic environment, thereby assisting in safe and smooth traffic for each traffic participant in the target traffic area 9.

[0034] exist Figure 1 In this section, we will explain the situation where, in an urban area that includes a roadway 51, an intersection 52, a pedestrian walkway 53, and a traffic signal 54 as traffic infrastructure equipment, the area near the intersection 52 of that urban area is designated as a target traffic zone 9. Figure 1 The illustration shows a total of 7 four-wheeled vehicles 2 and 2 motorcycles 3 moving within the roadway 51 and intersection 52. Additionally, 3 groups of pedestrians 4 are moving within the pedestrian walkway 53 and intersection 52. Figure 1 The diagram shows a setup with a total of 3 infrastructure cameras (56).

[0035] The traffic safety assistance system 1 includes: an on-board unit group 20 that moves with each four-wheeled vehicle 2 (in addition to the on-board unit mounted on the four-wheeled vehicle 2, it also includes a portable information processing terminal held or worn by the driver of the four-wheeled vehicle 2); an on-board unit group 30 that moves with each motorcycle 3 (in addition to the on-board unit mounted on the motorcycle 3, it also includes a portable information processing terminal held or worn by the driver of the motorcycle 3); a portable information processing terminal 40 held or worn by each pedestrian 4; multiple infrastructure cameras 56 installed in the target traffic area 9; a signal control device 55 for controlling the signal 54; and a coordination assistance device 6 that is communicatively connected to these on-board unit groups 20, 30, portable information processing terminals 40, infrastructure cameras 56, and signal control devices 55, etc., which exist in the target traffic area 9 (hereinafter also referred to as "area terminals").

[0036] The coordination assistance device 6 consists of one or more computers, which are communicatively connected to the aforementioned multiple regional terminals via base station 57. More specifically, the coordination assistance device 6 consists of a server connected to the multiple regional terminals via base station 57, network core, and the Internet, or an edge server connected to the multiple regional terminals via base station 57 and multi-access edge computing (MEC) core.

[0037] Figure 2This is a block diagram illustrating the structure of the coordination aid 6 and a plurality of regional terminals communicatively connected to the coordination aid 6.

[0038] The vehicle-mounted device group 20 installed on the four-wheeled vehicle 2 in the target traffic area 9 includes, for example: a vehicle-mounted driving assistance device 21 for assisting the driver, a notification device 22 for notifying the driver of various information, a driving subject status sensor 23 for detecting the status of the driver who is driving, a vehicle-mounted communication device 24 for wireless communication between the vehicle and the coordination assistance device 6 or other vehicles in the vicinity of the vehicle, and a portable information processing terminal 25 owned or worn by the driver, etc.

[0039] The vehicle-mounted driver assistance system 21 includes an external sensor unit, a vehicle status sensor, a navigation device, and a driver assistance electronic control unit (ECU). The external sensor unit includes: an external camera unit that captures images of the vehicle's surroundings; multiple onboard external sensors, such as radar units or laser detection and ranging (LIDAR) units that detect objects outside the vehicle using electromagnetic waves; and an external identification device that performs sensor fusion processing on the detection results of these external sensors to obtain information related to the vehicle's surroundings. The vehicle status sensor consists of sensors that acquire information related to the vehicle's driving status, such as a vehicle speed sensor, acceleration sensor, steering angle sensor, yaw rate sensor, position sensor, and orientation sensor. The navigation device includes, for example, a GNSS receiver that determines the vehicle's current position based on signals received from Global Navigation Satellite System (GNSS) satellites, and a storage device for storing map information.

[0040] The driver assistance ECU performs driver assistance controls such as lane departure mitigation control, lane change control, follow-the-leader control, mislaunch mitigation control, collision mitigation braking control, and collision avoidance control based on information obtained from external sensor units, vehicle status sensors, and navigation devices. Additionally, the driver assistance ECU generates driver assistance information to assist the driver in safe driving based on information obtained from external sensor units, vehicle status sensors, and navigation devices, and sends this information to the notification device 22.

[0041] Here, the driver assistance ECU initiates collision mitigation braking control when a moving object that may come into contact with the vehicle exists within a defined collision mitigation braking operating range centered on the vehicle. This collision mitigation braking control automatically operates the vehicle's braking system to mitigate damage caused by contact between the vehicle and other moving objects. Additionally, the driver assistance ECU initiates collision avoidance control when a moving object that may come into contact with the vehicle exists within a defined collision avoidance steering operating range centered on the vehicle. This collision avoidance control automatically operates the vehicle's steering system to avoid contact between the vehicle and other moving objects. Hereinafter, the collision mitigation braking operating range and the collision avoidance steering operating range will be collectively referred to as the "ADAS operating range."

[0042] The driver status sensor 23 is composed of various devices that acquire time-varying data related to the driving ability of the driver. For example, the driver status sensor 23 may include: an in-vehicle camera that detects the direction of the driver's gaze or whether the driver's eyes are open; a seatbelt sensor mounted on the driver's seatbelt that detects the driver's pulse or breathing; a steering wheel sensor mounted on the steering wheel that detects the driver's skin potential; or an in-vehicle microphone that detects whether there is conversation between the driver and passengers.

[0043] The vehicle communication device 24 has the following functions: transmitting information obtained by the driver assistance ECU (including information obtained by external sensor units, vehicle status sensors and navigation devices, and control information related to the driver assistance control being performed) and information related to the driver obtained by the driver status sensor 23 to the coordination assistance device 6; and receiving coordination assistance information sent from the coordination assistance device 6 and sending the received coordination assistance information to the notification device 22.

[0044] The notification device 22 is composed of various devices that enable the human-machine interface (hereinafter sometimes referred to as "HMI") to operate in a manner determined by driving assistance information sent from the vehicle driving assistance device 21 and coordination assistance information sent from the coordination assistance device 6, thereby notifying the driver of various information through the driver's hearing, vision and touch.

[0045] Figure 3A This is a block diagram illustrating the structure of the notification device 22 mounted on a four-wheeled vehicle. Furthermore, in Figure 3A In the illustration, only the module related to control based on coordination assistance information is shown in the notification device 22, which is sent from the coordination assistance device 6.

[0046] The notification device 22 includes: an HMI 220 that operates in a manner that is perceptible to the driver; and an HMI control device 225 that enables the HMI 220 to operate based on coordination assistance information sent from the coordination assistance device 6.

[0047] HMI 220 includes: an audio device 221 that operates in a manner that the driver can perceive by hearing; a head-up display 222 that operates in a manner that the driver can perceive by vision; and a seat belt control device 223 and a seat vibration device 224 that operate in a manner that the driver can perceive by touch.

[0048] The audio system 221 includes: a headrest speaker 221a, disposed on the headrest of the driver's seat, capable of emitting directional dual-channel sound; and a main speaker 221b, disposed near the driver's seat or the front passenger seat. These headrest speakers 221a and the main speaker 221b emit sounds corresponding to commands from the HMI control unit 225. The head-up display 222 displays images corresponding to commands from the HMI control unit 225 within the driver's field of vision (e.g., through the windshield). The seatbelt control unit 223 adjusts the tension of the driver's seatbelt according to commands from the HMI control unit 225. The seat vibration device 224 causes the driver's seat to vibrate with an amplitude and / or frequency corresponding to commands from the HMI control unit 225.

[0049] HMI control device 225 includes: a sanitation control device 226 that provides sanitation notification, the sanitation notification being configured to operate HMI 220 in a manner determined to improve the driver's driving ability (especially cognitive ability); a risk notification control device 227 that provides risk notification, the risk notification being configured to operate HMI 220 in a manner determined to make the driver aware of the presence of an imminent risk; and a risk area notification control device 228 that provides risk area notification, the risk area notification being configured to operate HMI 220 in a manner determined to make the driver aware of information related to the risk in a given area. As will be explained later, the coordination assistance information sent from the coordination assistance device 6 to the four-wheeled vehicle 2 includes: information related to the sanitation notification setting value for setting the sanitation notification to be turned on / off under the sanitation control device 226; information related to the risk notification setting value for setting the risk notification to be turned on / off under the risk notification control device 227 and the type of notification mode described later; information related to risks approaching the driver (hereinafter also referred to as "risk information"); and risk area information used in the risk area notification under the risk area notification control device 228.

[0050] The sanitation notification setting value input to the sanitation control device 226 is set to either "0" or "1", where "0" sets the sanitation notification under the sanitation control device 226 to be off, and "1" sets the sanitation notification under the sanitation control device 226 to be on.

[0051] When the sanitation notification setting value is "0", the sanitation control device 226 sets the sanitation notification to "off". That is, when the sanitation notification setting value is "0", the sanitation control device 226 prevents the HMI 220 from operating. However, this does not prevent the operation of the HMI 220 controlled by the risk notification control device 227.

[0052] When the sanitation notification setting value is "1", the sanitation control device 226 sets the sanitation notification to "on". More specifically, the sanitation control device 226 plays music that the driver is interested in and concerned about, for example, through the headrest speaker 221a or the main speaker 221b, thereby improving the driver's driving ability. In addition, at this time, in order to improve the driver's alertness, the beats per minute (BPM) of the music can be varied, or the bass can be emphasized.

[0053] Thus, since the sanitation control device 226 operates the HMI 220 to improve the driver's driving ability, the sanitation notification can also be turned off when the risk notification under the risk notification control device 227 (described later) is set to "on" (i.e., when the risk notification setting value is "1" or "2"), to avoid disturbing the driver. Furthermore, in this embodiment, the sanitation control device 226 operates the headrest speaker 221a or the main speaker 221b, thereby improving the driver's driving ability primarily through hearing; however, the invention is not limited to this. The sanitation control device 226 can also operate, for example, the seatbelt control device 223 or the seat vibration device 224.

[0054] In the risk notification control device 227, risk notifications can be performed under multiple notification modes that differ from at least one of the operating target devices and operating modes of the HMI 220. More specifically, in the risk notification control device 227, risk notifications can be performed under at least one of the following notification modes: a care notification mode aimed at making the driver aware of the existence of potential risks; a simulation notification mode aimed at making the driver aware of the existence and / or degree of the manifested risks; and a prediction assistance notification mode aimed at informing the driver of information that is beneficial to avoiding the predicted risks. Therefore, the risk notification setting value input to the risk notification control device 227 is set to any one of "0", "1", "2", "3", "4" and "5", where "0" sets the risk notification to off, "1" sets the risk notification to on in care notification mode, "2" sets the risk notification to on in simulation notification mode, "3" sets the risk notification to on in prediction assistance notification mode, "4" sets the risk notification to on in both care notification mode and prediction assistance notification mode, and "5" sets the risk notification to on in both simulation notification mode and prediction assistance notification mode.

[0055] When the risk notification setting value is "0", the risk notification control device 227 sets the risk notification to "off". That is, when the risk notification setting value is "0", the risk notification control device 227 does not enable the HMI 220 to operate. However, this does not prevent the operation of the HMI 220 controlled by the sanitation control device 226.

[0056] When the risk notification setting value is "1", the risk notification control device 227 sets the notification mode to care notification mode and enables risk notification under the set notification mode.

[0057] When the risk notification setting value is "2", the risk notification control device 227 sets the notification mode to the simulation notification mode and enables the risk notification under the set notification mode.

[0058] When the risk notification setting value is "3", the risk notification control device 227 sets the notification mode to the prediction auxiliary notification mode and enables the risk notification under the set notification mode.

[0059] When the risk notification setting value is "4", the risk notification control device 227 sets the notification mode to care notification mode and prediction assistance notification mode, and enables risk notifications under these set notification modes.

[0060] In addition, when the risk notification setting value is "5", the risk notification control device 227 sets the notification mode to simulation notification mode and prediction auxiliary notification mode, and enables risk notifications under these set notification modes.

[0061] Here, when the notification mode is set to predictive assistance notification mode, the risk notification control device 227 generates risk avoidance assistance information that is beneficial for avoiding risks approaching the driver based on the risk information sent from the coordination assistance device 6, and enables the audio device 221 or head-up display 222 of the HMI 220 to operate in a way that allows the driver to perceive the risk avoidance assistance information through hearing or sight. Here, the risk avoidance assistance information includes: information related to the location of traffic participants who may come into contact with the vehicle (hereinafter also referred to as "risk objects"), information related to the location where there is a possibility of contact between the vehicle and the risk object (hereinafter also referred to as "risk occurrence location"), and information that draws the driver's attention to the risk object.

[0062] More specifically, when a motorcycle driven by an unfit rider is in front of a four-wheeled vehicle driven by a driver, the risk notification control device 227 issues a message such as "Caution: Dangerous right turn by a two-wheeled vehicle" via the audio device 221 or displays it on the head-up display 222 as risk avoidance assistance information to avoid contact with the motorcycle. Additionally, at this time, the risk notification control device 227 can also display an arrow image indicating the current or predicted position of the motorcycle via the head-up display 222 as risk avoidance assistance information to avoid contact with the motorcycle.

[0063] Furthermore, when the notification mode is set to the care notification mode, the risk notification control device 227 operates the HMI 220 in a manner that does not annoy the driver, thereby allowing the driver to naturally recognize the existence of the risk object extracted from the risk information sent by the coordination assistance device 6. Thus, in the care notification mode, in order to allow the driver to naturally recognize the existence of the risk object without annoyance, the risk notification control device 227 preferably operates the headrest speaker 221a, which relies on the driver's hearing, among the multiple devices included in the HMI 220. More specifically, when the notification mode is set to the care notification mode, the risk notification control device 227, through the headrest speaker 221a, emits a familiar effect sound at a low volume, with a directional sound pointing towards the location of the risk object or the location where the risk occurs, thereby naturally directing the driver's gaze towards the location of the risk object or the location where the risk occurs.

[0064] Furthermore, when the notification mode is set to simulated notification mode, the risk notification control device 227 causes the HMI 220 to operate in a manner different from the aforementioned care notification mode, thereby making the driver strongly aware of the existence of the risk object extracted from the risk information sent by the coordination assistance device 6 and the degree of risk associated with that risk object. Thus, in simulated notification mode, to make the driver strongly aware of the existence of the risk object, the risk notification control device 227 causes the HMI 220 to operate with a higher notification intensity than determined in care notification mode. Here, notification intensity refers to the strength of attracting the driver's attention and concern. More specifically, when the notification mode is set to simulated notification mode, the risk notification control device 227 emits a buzzer or pulse sound with a louder volume than the effect sound emitted in care notification mode via the headrest speaker 221a or the main speaker 221b. These buzzer or pulse sounds are unfamiliar to the driver and are louder than the effect sound emitted in care notification mode, therefore the notification intensity is higher than the effect sound emitted in care notification mode.

[0065] Furthermore, in this embodiment, the risk notification control device 227 operates the audio device 221 when the notification mode is set to simulated notification mode, but the present invention is not limited thereto. When the notification mode is set to simulated notification mode, the risk notification control device 227 may also operate the seatbelt control device 223 to change the tension of the seatbelt, or operate the seat vibration device 224 to vibrate the seat, instead of operating the audio device 221. In this way, the seatbelt control device 223 and the seat vibration device 224 operate in a manner relying on the driver's tactile sense, so the notification intensity is higher than the effect sound emitted in the care notification mode. Additionally, when the notification mode is set to simulated notification mode, the risk notification control device 227 may also operate the audio device 221, the seatbelt control device 223, and the seat vibration device 224 in combination.

[0066] Furthermore, as described above, in the simulated notification mode, in addition to the presence of the risk object, to ensure the driver is strongly aware of the degree of risk associated with that risk object, the risk notification control device 227 preferably varies the notification intensity based on the degree of risk (e.g., the collision prediction time for the risk object) extracted from the risk information sent by the coordination assistance device 6. Specifically, the risk notification control device 227 may also increase the volume of the beeping sound, the volume of the pulse sound, or the interval between pulse sounds to enhance the notification intensity as the risk level increases (i.e., the collision prediction time shortens). As described above, when the seat belt control device 223 is activated, the risk notification control device 227 may also increase the seat belt tension to enhance the notification intensity as the risk level increases. Additionally, as described above, when the seat vibration device 224 is activated, the risk notification control device 227 may also increase the amplitude of the seat vibration to enhance the notification intensity as the risk level increases.

[0067] Furthermore, when the risk notification control device 227 changes the notification intensity according to the degree of risk, it is preferable to activate the HMI 220 at the moment when the collision mitigation braking control or collision avoidance steering control of the aforementioned driver assistance ECU begins to execute. In other words, when a risky object intrudes into the operating range of the vehicle's ADAS, the HMI 220 is activated in a manner with the greatest notification intensity.

[0068] The risk area notification control device 228 activates the HMI 220 based on risk area information sequentially transmitted from the coordination assistance device 6, thereby notifying the driver of information related to the current risk area. As described later, the risk area information includes information related to the current risk level of each local area after subdividing the target traffic area. Therefore, when the risk area notification control device 228 determines, based on the risk area information, that the vehicle is driving in a local area with a high risk level, it will, for example, issue a message such as "Driving in a high-risk area. Please be aware of your surroundings" via the main speaker 221b or display it on the head-up display 222.

[0069] return Figure 2The portable information processing terminal 25 may be, for example, a wearable terminal worn by the driver of the four-wheeled vehicle 2, or a smartphone held by the driver. The wearable terminal has the following functions: measuring the driver's biometric information such as heart rate, blood pressure, and blood oxygen saturation, and sending the measured data to the coordination assistance device 6; and receiving coordination assistance information sent from the coordination assistance device 6, and notifying the driver of corresponding messages via images, voice, warning sounds, and vibrations. The smartphone has the following functions: sending driver-related information such as the driver's location, acceleration, and schedule to the coordination assistance device 6; and receiving coordination assistance information sent from the coordination assistance device 6, and notifying the driver of corresponding messages via images, voice, warning sounds, melodies, and vibrations.

[0070] The on-board unit group 30 mounted on the motorcycle 3 in the target traffic area 9 includes, for example: an on-board driving assistance device 31 to assist the rider in driving, a notification device 32 to notify the rider of various information, a rider status sensor 33 to detect the status of the rider who is driving, an on-board communication device 34 to conduct wireless communication between the motorcycle and the coordination assistance device 6 or other vehicles in the vicinity of the motorcycle, and a portable information processing terminal 35 owned or worn by the rider, etc.

[0071] The vehicle-mounted driving assistance device 31 includes an external sensor unit, a vehicle status sensor, a navigation device, and a driving assistance ECU. The external sensor unit includes: an external camera unit that captures images of the vehicle's surroundings; multiple onboard external sensors, such as a radar unit or LIDAR unit, that detect objects outside the vehicle using electromagnetic waves; and an external identification device that performs sensor fusion processing on the detection results of these external sensors to obtain information related to the vehicle's surroundings. The vehicle status sensor consists of sensors that acquire information related to the vehicle's driving status, such as a vehicle speed sensor and a 5-axis or 6-axis inertial measurement unit. The navigation device includes, for example, a GNSS receiver that determines the current location based on signals received from GNSS satellites, and a storage device that stores map information.

[0072] The driver assistance ECU performs driver assistance controls such as lane keeping control, lane departure mitigation control, lane change control, follow-the-leader control, mislaunch prevention control, and collision mitigation braking control based on information obtained from external sensor units, vehicle status sensors, and navigation devices. Additionally, the driver assistance ECU generates driver assistance information to assist the rider in safe driving based on information obtained from external sensor units, vehicle status sensors, and navigation devices, and sends this information to the notification device 32.

[0073] Here, the driver assistance ECU initiates collision mitigation braking control on the condition that there is a moving object that may come into contact with the vehicle within a defined collision mitigation braking operating range centered on the vehicle (hereinafter, in conjunction with the terminology defined for four-wheeled vehicles 2, also referred to as the "ADAS operating range"). The collision mitigation braking control automatically operates the vehicle's braking device to mitigate damage caused by contact between the vehicle and other moving objects.

[0074] The rider status sensor 33 is composed of various devices that acquire information related to the rider's driving ability. For example, the rider status sensor 33 may be composed of a seat sensor installed on the rider's seat to detect the presence or absence of the rider's pulse or breathing, or a helmet sensor installed on the rider's helmet to detect the presence or absence of the rider's pulse, breathing, and skin potential.

[0075] The vehicle communication device 34 has the following functions: transmitting information obtained by the driving assistance ECU (including information obtained by external sensor units, vehicle status sensors and navigation devices, and control information related to the driving assistance control being performed) and rider-related information obtained by the rider status sensor 33 to the coordination assistance device 6; and receiving coordination assistance information sent from the coordination assistance device 6 and sending the received coordination assistance information to the notification device 32.

[0076] The notification device 32 consists of various devices that enable the HMI to operate in a manner determined by driving assistance information sent from the onboard driving assistance device 21 and coordination assistance information sent from the coordination assistance device 6, thereby notifying the rider of various information through the rider's hearing, vision, and touch.

[0077] Figure 3B This is a block diagram illustrating the structure of the notification device 32 mounted on a motorcycle. Furthermore, in Figure 3B The diagram only shows modules in the notification device 32, particularly those related to control based on coordination assistance information sent from the coordination assistance device 6.

[0078] The notification device 32 includes: an HMI 320 that operates in a manner that is understandable to the rider; and an HMI control device 325 that enables the HMI 320 to operate based on coordination assistance information sent from the coordination assistance device 6.

[0079] The HMI 320 includes: a head-mounted speaker 321 that operates in a manner that allows the rider to perceive through hearing; and a head-up display 322 that operates in a manner that allows the rider to perceive through vision.

[0080] A head-mounted speaker 321 is mounted on the rider's helmet and is capable of emitting directional, two-channel sound. The head-mounted speaker 321 emits sounds corresponding to commands from the HMI control unit 325. The head-up display 322 displays images corresponding to commands from the HMI control unit 325 within the rider's field of vision (e.g., on the helmet's shield).

[0081] HMI control device 325 includes: a sanitation control device 326 that provides sanitation notifications, the sanitation notifications being configured to operate HMI 320 in a manner determined to improve the rider's driving ability (especially cognitive ability); a risk notification control device 327 that provides risk notifications, the risk notifications being configured to operate HMI 320 in a manner determined to make the rider aware of the presence of an approaching risk; and a risk area notification control device 328 that provides risk area notifications, the risk area notifications being configured to operate HMI 320 in a manner determined to make the rider aware of information related to the risk in a given area. As will be explained later, the coordination assistance information sent from coordination assistance device 6 to motorcycle 3 includes: information related to sanitation notification setting values ​​for setting the on / off state of sanitation notifications under sanitation control device 326; risk notification setting values ​​related to setting the on / off state of risk notifications and the type of notification mode under risk notification control device 327; risk information related to risks approaching the rider; and risk area information used in risk area notifications under risk area notification control device 328, etc.

[0082] The sanitation notification setting value input to the sanitation control device 326 is set to either "0" or "1", where "0" sets the sanitation notification under the sanitation control device 326 to be off, and "1" sets the sanitation notification under the sanitation control device 326 to be on.

[0083] When the sanitation notification setting value is "0", the sanitation control device 326 sets the sanitation notification to "off". That is, when the sanitation notification setting value is "0", the sanitation control device 326 does not enable the HMI 320 to operate. However, this does not prevent the operation of the HMI 320 controlled by the risk notification control device 327.

[0084] When the sanitation notification setting value is "1", the sanitation control device 326 sets the sanitation notification to be enabled. More specifically, the sanitation control device 326, for example, plays music that the rider is interested in and cares about through the headphone speaker 321, thereby improving the rider's riding ability. In addition, at this time, in order to improve the rider's alertness, the BPM of the music can be changed, or the bass can be emphasized.

[0085] Thus, since the sanitation control device 326 enables the HMI 320 to operate in order to improve the rider's driving ability, the sanitation notification can also be turned off when the risk notification under the risk notification control device 327 described later is set to on (i.e., when the risk notification setting value is "1" or "2"), so as not to bother the rider.

[0086] In the risk notification control device 327, risk notifications can be performed under multiple notification modes that differ from at least one of the operating target devices and operating modes of the HMI 320. More specifically, in the risk notification control device 327, risk notifications can be performed under at least one of the following notification modes: a care notification mode aimed at making riders aware of the existence of potential risks; a simulation notification mode aimed at making riders aware of the existence and / or degree of manifested risks; and a prediction assistance notification mode aimed at informing riders of information that is beneficial to avoiding predicted risks. Therefore, the risk notification setting value input to the risk notification control device 327 is set to any one of "0", "1", "2", "3", "4" and "5", where "0" sets the risk notification to off, "1" sets the risk notification to on in care notification mode, "2" sets the risk notification to on in simulation notification mode, "3" sets the risk notification to on in prediction assistance notification mode, "4" sets the risk notification to on in both care notification mode and prediction assistance notification mode, and "5" sets the risk notification to on in both simulation notification mode and prediction assistance notification mode.

[0087] When the risk notification setting value is "0", the risk notification control device 327 sets the risk notification to "off". That is, when the risk notification setting value is "0", the risk notification control device 327 does not enable the HMI 320 to operate. However, this does not prevent the operation of the HMI 320 controlled by the sanitation control device 326.

[0088] When the risk notification setting value is "1", the risk notification control device 327 sets the notification mode to care notification mode and enables risk notification under the set notification mode.

[0089] When the risk notification setting value is "2", the risk notification control device 327 sets the notification mode to the simulation notification mode and enables the risk notification under the set notification mode.

[0090] When the risk notification setting value is "3", the risk notification control device 327 sets the notification mode to the prediction auxiliary notification mode and enables the risk notification under the set notification mode.

[0091] When the risk notification setting value is "4", the risk notification control device 327 sets the notification mode to care notification mode and prediction assistance notification mode, and enables risk notifications under these set notification modes.

[0092] In addition, when the risk notification setting value is "5", the risk notification control device 327 sets the notification mode to simulation notification mode and prediction auxiliary notification mode, and enables risk notifications under these set notification modes.

[0093] Here, when the notification mode is set to predictive assistance notification mode, the risk notification control device 327 generates risk avoidance assistance information beneficial for avoiding risks approaching the rider based on the risk information sent from the coordination assistance device 6, and enables the head-mounted speaker 321 or head-up display 322 of the HMI 320 to operate in a way that allows the rider to perceive the risk avoidance assistance information through hearing or sight. Here, the risk avoidance assistance information includes: information related to the location of the risk object that may come into contact with the vehicle, information related to the location where the risk occurs, and information that draws the rider's attention to the risk object.

[0094] More specifically, when a four-wheeled vehicle driven by an unfit driver is in front of a motorcycle, the risk notification control device 327 emits a message such as "Caution: Dangerous right turn by a four-wheeled vehicle" via a head-mounted speaker 321 or displays it on a head-up display 322 as risk avoidance assistance information to avoid contact with the four-wheeled vehicle. Additionally, the risk notification control device 327 can also display an arrow indicating the current or predicted position of the four-wheeled vehicle via the head-up display 322 as risk avoidance assistance information to avoid contact with the four-wheeled vehicle.

[0095] Furthermore, when the notification mode is set to the care notification mode, the risk notification control device 327 enables the HMI 320 to operate in a way that does not annoy the rider, thereby allowing the rider to naturally recognize the existence of the risk object extracted from the risk information sent by the coordination assistance device 6. Thus, in the care notification mode, in order to enable the rider to naturally recognize the existence of the risk object without annoyance, the risk notification control device 327 preferably operates the headset speaker 321, which relies on the rider's hearing, among the multiple devices included in the HMI 320. More specifically, when the notification mode is set to the care notification mode, the risk notification control device 327 uses the headset speaker 321 to emit a familiar effect sound with a directional dual-channel sound pointing towards the location of the risk object or the location where the risk occurred, at a low volume, thereby naturally directing the rider's gaze towards the location of the risk object or the location where the risk occurred.

[0096] Furthermore, when the notification mode is set to simulated notification mode, the risk notification control device 327 causes the HMI 320 to operate in a manner different from the aforementioned care notification mode, thereby enabling the rider to strongly perceive the existence of the risk object extracted from the risk information sent by the coordination assistance device 6 and the degree of risk associated with that risk object. Thus, in simulated notification mode, to ensure the rider's strong perception of the risk object's existence, the risk notification control device 327 causes the HMI 320 to operate with a higher notification intensity than determined in care notification mode. More specifically, when the notification mode is set to simulated notification mode, the risk notification control device 327 emits a buzzer or pulse sound via the head-mounted speaker 321, with a volume louder than the effect sound emitted in care notification mode. These buzzer or pulse sounds are unfamiliar to the rider and are louder than the effect sound emitted in care notification mode, therefore the notification intensity is higher than the effect sound emitted in care notification mode.

[0097] Furthermore, as described above, in the simulated notification mode, in addition to the presence of the risky object, in order to make the rider strongly aware of the degree of risk for that risky object, the risk notification control device 327 preferably varies the notification intensity based on the degree of risk for the risky object (e.g., the collision prediction time for the risky object) extracted from the risk information sent by the coordination assistance device 6. Specifically, the risk notification control device 327 may also increase the volume of the buzzer sound, or increase the volume of the pulse sound, or shorten the interval of the pulse sound, to increase the notification intensity when the degree of risk is higher (i.e., the shorter the collision prediction time).

[0098] Furthermore, when the risk notification control device 327 changes the notification intensity according to the degree of risk, it is preferable to activate the HMI 320 at the moment when the collision mitigation braking control of the aforementioned driver assistance ECU begins to execute. In other words, when a risky object intrudes into the operating range of the vehicle's ADAS, the HMI 320 is activated in the manner with the greatest notification intensity.

[0099] The risk area notification control device 328 activates the HMI 320 based on risk area information sequentially transmitted from the coordination assistance device 6, thereby notifying the driver of information related to the current risk area. When the risk area notification control device 328 determines, based on the risk area information, that the vehicle is driving in a high-risk local area, it will, for example, issue a message via the head-mounted speaker 321 or display on the head-up display 322 stating, "Driving in a high-risk area. Please be aware of your surroundings."

[0100] return Figure 2The portable information processing terminal 40 owned or worn by pedestrian 4 in the target traffic area 9 is, for example, a wearable terminal worn by pedestrian 4 or a smartphone held by pedestrian 4. The wearable terminal has the following functions: measuring pedestrian 4's biometric information such as heart rate, blood pressure, and blood oxygen saturation, and sending the measured data to the coordination assistance device 6, or receiving coordination assistance information sent from the coordination assistance device 6. Additionally, the smartphone has the following functions: sending pedestrian information related to pedestrian 4, such as location information, movement acceleration, and schedule information, to the coordination assistance device 6, or receiving coordination assistance information sent from the coordination assistance device 6.

[0101] In addition, the portable information processing terminal 40 includes a notification device 42, which notifies pedestrians of various information by enabling the HMI to operate in a manner determined based on the received coordination assistance information, through the pedestrian's hearing, vision and touch.

[0102] Figure 3C This is a block diagram illustrating the structure of the notification device 42 mounted on the portable information processing terminal 40. Furthermore, in Figure 3C In the diagram, only the module related to control based on coordination assistance information is shown in the notification device 42, which is sent from the coordination assistance device 6.

[0103] The notification device 42 includes: an HMI 420 that operates in a manner perceptible to pedestrians; and an HMI control device 425 that enables the HMI 420 to operate based on coordination assistance information sent from the coordination assistance device 6.

[0104] HMI 420 includes: a speaker 421 that operates in a manner that allows a pedestrian to perceive through hearing; and a vibration device 424 that operates in a manner that allows a pedestrian to perceive through touch.

[0105] The speaker 421 emits a sound corresponding to the command from the HMI control device 425. The excitation device 424 causes the main body of the portable information processing terminal 40 to vibrate with an amplitude and / or vibration frequency corresponding to the command from the HMI control device 425.

[0106] As will be explained later, the coordination assistance information sent from the coordination assistance device 6 to the portable information processing terminal 40 owned by the pedestrian includes: information related to risk notification setting values ​​for setting the type of risk notification on / off and notification mode under the HMI control device 425, and risk information related to risks approaching the pedestrian.

[0107] In the HMI control device 425, risk notifications can be performed under multiple notification modes that differ from at least any of the operating objects and operating modes of the HMI 420. More specifically, in the HMI control device 425, risk notifications can be performed under at least any of the following notification modes: a care notification mode aimed at informing pedestrians of the existence of potential risks, and a simulation notification mode aimed at informing pedestrians of the existence and / or degree of risk. Therefore, the risk notification setting value input to the HMI control device 425 is set to any one of "0", "1", and "2", where "0" sets the risk notification under the HMI control device 425 to be off, "1" sets the risk notification under the HMI control device 425 to be on and sets the notification mode to care notification mode, and "2" sets the risk notification under the HMI control device 425 to be on and sets the notification mode to simulation notification mode.

[0108] When the risk notification setting value is "0", the HMI control device 425 sets the risk notification to "off". That is, when the risk notification setting value is "0", the HMI control device 425 does not enable the HMI 420 to operate.

[0109] When the risk notification setting value is "1", the HMI control device 425 sets the notification mode to care notification mode and enables risk notification in the set notification mode.

[0110] In addition, when the risk notification setting value is "2", the HMI control device 425 sets the notification mode to the simulation notification mode and enables the risk notification in the set notification mode.

[0111] Here, when the notification mode is set to the care notification mode, the HMI control device 425 causes the HMI 420 to operate in a way that does not annoy pedestrians, thereby allowing pedestrians to naturally recognize the existence of the risk object extracted from the risk information sent by the coordination assistance device 6. More specifically, when the notification mode is set to the care notification mode, the HMI control device 425 activates the vibration device 424, thereby causing the main body of the portable information processing terminal 40 to vibrate at a specified amplitude and frequency.

[0112] Furthermore, when the notification mode is set to simulated notification mode, the HMI control device 425 causes the HMI 420 to operate in a manner different from the aforementioned care notification mode, thereby making the pedestrian strongly aware of the existence of the risk object extracted from the risk information sent by the coordination assistance device 6 and the degree of risk to that risk object. Thus, in simulated notification mode, in order to make the pedestrian strongly aware of the existence of the risk object, the HMI control device 425 causes the HMI 420 to operate in a manner with a higher notification intensity than determined in care notification mode. More specifically, when the notification mode is set to simulated notification mode, the HMI control device 425 emits a buzzer tone, a pulse tone, and a message indicating the presence of risk via the speaker 421.

[0113] Furthermore, as described above, in the simulated notification mode, in addition to the presence of the risky object, in order to make pedestrians strongly aware of the degree of risk for that risky object, the HMI control device 425 preferably varies the notification intensity based on the degree of risk for the risky object (e.g., the collision prediction time for the risky object) extracted from the risk information sent by the coordination assistance device 6. Specifically, the HMI control device 425 may also increase the notification intensity by increasing the volume of the beep tone, increasing the volume of the pulse tone, shortening the pulse tone interval, increasing the volume of the message, or changing the content of the message, as the degree of risk increases (i.e., the shorter the collision prediction time).

[0114] return Figure 2 Infrastructure camera 56 captures images of traffic infrastructure, including roadways, intersections, and pedestrian walkways within the target traffic area, as well as images of moving bodies or pedestrians moving within these roadways, intersections, and pedestrian walkways, and sends the acquired image information to coordination assistance device 6.

[0115] The signal control device 55 controls the signal and sends signal status information, such as the current illuminated color or the timing of switching the illuminated color, to the coordination auxiliary device 6.

[0116] The coordination assistance device 6 is a computer that, based on information acquired from multiple area terminals existing in the target traffic area as described above, generates coordination assistance information for each traffic participant as an assistance target. This information facilitates communication between traffic participants and the recognition of the surrounding traffic environment, and notifies each traffic participant, thereby assisting in safe and smooth traffic flow for traffic participants in the target traffic area. Furthermore, in this embodiment, traffic participants among the multiple traffic participants existing in the target traffic area who possess means (e.g., vehicle-mounted device groups 20, 30, portable information processing terminal 40, notification devices 22, 32, 42) are the assistance targets of the coordination assistance device 6. These means receive the coordination assistance information generated in the coordination assistance device 6 and cause the HMI to operate in a manner determined based on the received coordination assistance information.

[0117] The coordination assistance device 6 includes: a target traffic area identification unit 60, which identifies people and moving bodies in the target traffic area as various traffic participants; a driver subject information acquisition unit 61, which acquires driver subject status information related to the driving ability of the driver subject of the moving body identified by the target traffic area identification unit 60 as a traffic participant; a prediction unit 62, which predicts the future movements of traffic participants in the target traffic area; a sanitation notification setting unit 63, which sets the activation / deactivation of sanitation notifications for each traffic participant identified by the target traffic area identification unit 60 as an assistance object; a risk notification setting unit 64, which sets the notification mode of risk notifications for each traffic participant identified by the target traffic area identification unit 60 as an assistance object; a coordination assistance information notification unit 65, which sends coordination assistance information generated for each traffic participant identified by the target traffic area identification unit 60 as an assistance object; a traffic environment database 67, which stores information related to the traffic environment of the target traffic area; and a driving history database 68, which stores information related to the past driving history of pre-registered driver subjects.

[0118] The traffic environment database 67 stores pre-registered map information of target traffic areas (e.g., width of carriageways, number of lanes, speed limits, width of pedestrian walkways, presence or absence of guardrails between carriageways and pedestrian walkways, and location of pedestrian crossings), information on hazardous areas, particularly high-risk areas within the target traffic areas, and information related to the traffic environment of traffic participants in the target traffic areas. Hereinafter, the information stored in the traffic environment database 67 will also be referred to as registered traffic environment information. Furthermore, the hazardous area information stored in this traffic environment database 67 is updated on a period of approximately several hours to several days, whereas the risk area information described later differs in that it is updated almost in real time.

[0119] The driving history database 68 stores information related to the past driving history of pre-registered drivers. This information is stored in a state associated with the registration number of the vehicle owned by the driver. Therefore, if the registration number of the vehicle being identified can be determined by the object traffic area identification unit 60 (described later), the past driving history of the driver of the vehicle being identified can be obtained by retrieving the driving history database 68 based on that registration number. Hereinafter, the information stored in the driving history database 68 will also be referred to as registered driving history information.

[0120] The target traffic area identification unit 60 identifies the people or moving bodies contained in the target traffic area, namely each traffic participant, and the traffic environment of each traffic participant in the target traffic area, based on information sent from the aforementioned area terminals (vehicle-mounted device group 20, 30, portable information processing terminal 40, infrastructure camera 56 and signal control device 55) in the target traffic area and registered traffic environment information read from the traffic environment database 67, and obtains identification information related to these identified objects.

[0121] Here, the information sent from the vehicle-mounted driving assistance device 21 and vehicle-mounted communication device 24 included in the vehicle-mounted device group 20 to the target traffic area identification unit 60, and the information sent from the vehicle-mounted driving assistance device 31 and vehicle-mounted communication device 34 included in the vehicle-mounted device group 30 to the target traffic area identification unit 60, includes: information related to the status of traffic participants or the traffic environment around the vehicle obtained by external sensor units, and information related to the status of the vehicle as a traffic participant obtained by vehicle status sensors or navigation devices, etc. Additionally, the information sent from the portable information processing terminal 40 to the target traffic area identification unit 60 includes: information related to the status of pedestrians as traffic participants, such as position and acceleration. Furthermore, the image information sent from the infrastructure camera 56 to the target traffic area identification unit 60 includes: the appearance of traffic infrastructure equipment such as roadways, intersections, and pedestrian walkways in the target traffic area, and the appearance of traffic participants moving in the target traffic area, etc., information related to each traffic participant or their traffic environment. Furthermore, the signal status information sent from the signal control device 55 to the target traffic area identification unit 60 includes information related to the traffic environment of each traffic participant, such as the current illuminated color of the signal and the timing of the color change. Additionally, the registered traffic environment information read by the target traffic area identification unit 60 from the traffic environment database 67 includes map information and danger zone information of the target traffic area, and other information related to the traffic environment of each traffic participant.

[0122] Therefore, in the target traffic area identification unit 60, identification information of each traffic participant in the target traffic area can be obtained based on information sent from these area terminals, including their position, speed, acceleration, direction of movement, vehicle type, vehicle class, registration number, number of pedestrians, and age group (hereinafter also referred to as "traffic participant identification information"). Furthermore, in the target traffic area identification unit 60, identification information of the traffic environment of each traffic participant in the target traffic area can be obtained based on information sent from these area terminals, including the width of the carriageway, number of lanes, speed limit, width of the pedestrian walkway, presence or absence of guardrails between the carriageway and pedestrian walkway, the illuminated color of traffic signals and their switching timing, weather, illumination, road surface condition, and hazardous area information (hereinafter also referred to as "traffic environment identification information").

[0123] Therefore, in this embodiment, the means of identifying traffic participants and the traffic environment in the target traffic area consists of the following devices: a target traffic area identification unit 60; an onboard driving assistance device 21, an onboard communication device 24, and a portable information processing terminal 25 included in the onboard device group 20 of a four-wheeled vehicle 2; an onboard driving assistance device 31, an onboard communication device 34, and a portable information processing terminal 35 included in the onboard device group 30 of a motorcycle 3; a portable information processing terminal 40 for pedestrians 4; an infrastructure camera 56; a signal control device 55; and a traffic environment database 67.

[0124] The traffic area identification unit 60 sends the traffic participant identification information and traffic environment identification information obtained as described above to the driver subject information acquisition unit 61, prediction unit 62, sound notification setting unit 63, risk notification setting unit 64, and coordination assistance information notification unit 65, etc.

[0125] The driver information acquisition unit 61 acquires driver status information and driver characteristic information related to the current driving ability of the driver being identified as a traffic participant by the target traffic area identification unit 60, based on information sent from the aforementioned area terminal (especially the vehicle-mounted device group 20, 30) in the target traffic area and registered driving history information read from the driving history database 68.

[0126] More specifically, when the driver of a four-wheeled vehicle identified as a traffic participant by the object traffic area identification unit 60 is a human, the driver subject information acquisition unit 61 acquires information sent from the on-board unit group 20 mounted on the four-wheeled vehicle as the driver's driving subject status information. Additionally, when the driver of a motorcycle identified as a traffic participant by the object traffic area identification unit 60 is a human, the driver subject information acquisition unit 61 acquires information sent from the on-board unit group 30 mounted on the motorcycle as the rider's driving subject status information.

[0127] Here, the information sent from the driver status sensor 23 and vehicle communication device 24 included in the vehicle-mounted device group 20 to the driver information acquisition unit 61 includes: time-lapse data related to the driver's driving ability, such as the direction of their gaze and whether their eyes are open, biometric information such as pulse, respiration, and skin potential, and voice information such as whether they are talking. Additionally, the information sent from the rider status sensor 33 and vehicle communication device 34 included in the vehicle-mounted device group 30 to the driver information acquisition unit 61 includes: time-lapse data related to the rider's pulse, respiration, and skin potential, and information related to the rider's driving ability. Furthermore, the information sent from the portable information processing terminals 25 and 35 included in the vehicle-mounted device groups 20 and 30 to the driver information acquisition unit 61 includes: the driver's or rider's personal schedule information. For example, when driving a mobile vehicle under a tight schedule, drivers or riders may experience anxiety, leading to a decline in driving ability. Therefore, the schedule information of a driver or rider can be said to be related to their own driving ability.

[0128] The driver subject information acquisition unit 61 acquires driver subject characteristic information related to the current driving ability of the driver subject (e.g., excessive sudden lane changes and excessive sudden acceleration and deceleration) by using either the driver subject status information for the driver subject acquired according to the above process or the registered driving history information read from the driving history database 68.

[0129] The driver information acquisition unit 61 sends the driver status information and driver characteristic information of the driver obtained as described above to the prediction unit 62, the improvement notification setting unit 63, the risk notification setting unit 64, and the coordination and assistance information notification unit 65, etc.

[0130] The prediction unit 62 predicts the risk in the entire target traffic area based on the traffic participant identification information and traffic environment identification information (hereinafter collectively referred to as "identification information") obtained by the target traffic area identification unit 60 and the driver status information and driver characteristic information (hereinafter collectively referred to as "driver information") obtained by the driver subject information acquisition unit 61.

[0131] Figure 4 This is a functional block diagram illustrating the specific structure of the prediction unit 62.

[0132] The prediction unit 62 includes: a regional risk prediction unit 620, which predicts the risk for traffic participants by dividing the target traffic area into multiple local areas; and a traffic participant risk prediction unit 625, which predicts the individual risks of each traffic participant existing in each local area.

[0133] Here, the target traffic area is, for example, a relatively broad traffic area defined by towns and villages. In contrast, the individual local areas obtained by subdividing the target traffic area are, for example, traffic areas near intersections or specific facilities, which a four-wheeled vehicle can pass through in about tens of seconds at legal speeds. That is, each local area is narrower than the target traffic area, but is designed to have a wider operating range than the ADAS (Advanced Driver Assistance Systems) operated by the driver assistance ECUs mounted on each vehicle. Furthermore, the extent of each local area can be fixed or varied depending on the situation. Additionally, a portion of the extent of each local area may overlap with a portion of the extent of another adjacent local area.

[0134] The regional risk prediction unit 620 includes a statistical processing calculation unit 621, a data preprocessing calculation unit 622, a macro risk estimation model 623, and a high-risk area extraction unit 624, which are used to predict the risk of each local area.

[0135] The statistical processing unit 621 extracts information related to the risk level of each local area by performing prescribed statistical processing on the identification information and driver information related to the entire target traffic area.

[0136] The data preprocessing calculation unit 622 generates input data for the macro risk estimation model 623 based on the information that has been statistically processed by the statistical processing calculation unit 621, and inputs it into the macro risk estimation model 623.

[0137] The macro-risk estimation model 623, for example, has a deep neural network (DNN), which is constructed by machine learning in such a way that when the input data has been processed by the data preprocessing unit 622, it outputs the risk level of each local region. Information related to the risk level of each local region calculated by the macro-risk estimation model 623 is sent to the high-risk region extraction unit 624 and the coordination auxiliary information notification unit 65.

[0138] The high-risk area extraction unit 624 extracts at least one high-risk area from multiple local areas constituting the target traffic area based on the risk level of each local area calculated by the macro-risk estimation model 623. More specifically, the high-risk area extraction unit 624 extracts local areas among the multiple local areas whose risk level calculated by the macro-risk estimation model 623 is higher than a predetermined threshold as high-risk areas. Information related to the high-risk areas extracted by the high-risk area extraction unit 624 is sent to the traffic participant risk prediction unit 625.

[0139] The traffic participant risk prediction unit 625 includes a monitoring area information extraction unit 626, a data preprocessing and calculation unit 627, and a micro-risk estimation model 628. By using these, only the high-risk areas extracted by the regional risk prediction unit 620 are designated as monitoring areas, and the future risks of each traffic participant in these monitoring areas are predicted. That is, in the traffic participant risk prediction unit 625, local areas that are not extracted as high-risk areas by the regional risk prediction unit 620 (hereinafter also referred to as "low-risk areas") are not designated as monitoring areas, and the processing shown below is not performed.

[0140] The monitoring area information extraction unit 626 extracts information related to the monitoring area that has been identified as a high-risk area by the area risk prediction unit 620 from the identification information and driver subject information (that is, information related to traffic participants and their traffic environment in the identification information and information related to the driver subject of the moving body in the monitoring area from the driver subject information).

[0141] The data preprocessing and calculation unit 627 generates input data for the micro-risk estimation model 628 based on the identification information and driver information related to the monitoring area extracted by the monitoring area information extraction unit 626, and inputs it into the micro-risk estimation model 628.

[0142] The micro-risk estimation model 628, for example, has a DNN that is constructed by machine learning to output information related to the future risks of each traffic participant in the monitored area (more specifically, information related to the activity routes of each traffic participant, information related to the contact risk of each traffic participant, and the collision prediction time up to the occurrence of the contact risk) when input data related to the monitored area is processed by the data preprocessing unit 622. The information related to the risks of traffic participants in the monitored area calculated by the micro-risk estimation model 628 is sent to the risk notification setting unit 64 and the coordination assistance information notification unit 65.

[0143] In the prediction unit 62, which predicts the risk in the target traffic area by combining macro-risk estimation model 623 and micro-risk estimation model 628 as described above, the prediction accuracy can be improved by using the traffic safety assistance system 1 to provide coordination assistance information to each traffic participant through the server, and by learning these macro-risk estimation models 623 and micro-risk estimation models 628 through the following two processes.

[0144] <First Learning Method>

[0145] The first learning method includes the following steps: preparing learning data by using input data for a macro-risk estimation model 623 generated based on identification information and driver information acquired during service use, and the output of a micro-risk estimation model 628 when input data prepared using the same identification information and driver information is input to the micro-risk estimation model 628; and using the learning data to learn the macro-risk estimation model 623. In this first learning method, since the output of the micro-risk estimation model 628 is used as the forward solution data, it is useful if a highly accurate micro-risk estimation model 628 is obtained in advance.

[0146] <Second Learning Method>

[0147] The second learning method includes the following steps: preparing learning data by using input data for a macro-risk estimation model 623 generated based on first identification information and first driver information acquired during a specified first period in the service application process, and positive solution data for the output of a micro-risk estimation model 628 generated based on second identification information and second driver information acquired during a second period after the first period; and using the learning data to learn an overall model combining the macro-risk estimation model 623 and the micro-risk estimation model 628. In this second learning method, since the positive solution data needs to be prepared manually based on the second identification information, it requires more effort than the first learning method. However, it can learn both the macro-risk estimation model 623 and the micro-risk estimation model 628, thus improving the overall prediction accuracy of the prediction unit 62 compared to the first learning method.

[0148] return Figure 2 The sanitation notification setting unit 63 identifies traffic participants in the target traffic area who are designated as auxiliary objects by the target traffic area identification unit 60 and who are identified as moving bodies, and sets the sanitation notification to be turned on / off for each of these participants. Furthermore, as detailed later, traffic participants who are predicted by the prediction unit 62 to be parties to a potential contact risk become the setting targets for risk notifications under the risk notification setting unit 64. Therefore, it is preferable to remove the setting targets of the risk notification setting unit 64 from the setting targets of the sanitation notification setting unit 63.

[0149] More specifically, firstly, the soundness notification setting unit 63 acquires driver subject status information and driver subject characteristic information associated with the driver subjects of each set object, which are moving bodies, from the driver subject information acquisition unit 61. Secondly, based on the acquired driver subject status information and driver subject characteristic information, the soundness notification setting unit 63 calculates the current soundness of the driver subject for each set object. Thirdly, if the soundness calculated for each set object is less than a predetermined soundness threshold, the soundness notification setting unit 63 determines that the driver subject of that set object is in an unsound state, and sets the soundness notification setting value for that set object to "1" in order to enable the soundness notification. Fourthly, if the soundness calculated for each set object is above the soundness threshold, the soundness notification setting unit 63 determines that the driver subject of that set object is in a sound state, and sets the soundness notification setting value for that set object to "0" in order to disable the soundness notification.

[0150] The sanitation notification setting unit 63, through the process described above, sets the sanitation notification for multiple objects within the target traffic area to be enabled or disabled. Information related to the sanitation notification setting values ​​set for each object by the sanitation notification setting unit 63 is sent to the coordination assistance information notification unit 65.

[0151] The risk notification setting unit 64 sets the risk notification operation mode (i.e., the type of notification mode and the on / off state of risk notification) for each set object. This is based on the prediction results of the traffic participant risk prediction unit 625 of the prediction unit 62, the identification information obtained by the traffic area identification unit 60, and the driver subject information obtained by the driver subject information acquisition unit 61.

[0152] More specifically, the risk notification setting unit 64 sets the operation mode of risk notifications for each set object existing in the monitoring area based on the identification information associated with the monitoring area obtained by the object traffic area identification unit 60, the driver subject information associated with the monitoring area obtained by the driver subject information acquisition unit 61, and the prediction results of the monitoring area by the traffic participant risk prediction unit 625. That is, the risk notification setting unit 64 sets the risk notification setting value for each set object to any one of "0", "1", "2", "3", and "4".

[0153] Thus, in the risk notification setting unit 64, since the operation mode of risk notification is set separately for each set object existing in the monitoring area, for example, when the traffic participant risk prediction unit 625 predicts that a contact risk involving multiple set objects will occur in the monitoring area, risk notifications can be turned on / off at different times for the multiple predicted parties who are predicted to participate in the contact risk, or risk notifications can be sent simultaneously with different notification modes. Hereinafter, the process of setting a suitable risk notification operation mode for each set object in the risk notification setting unit 64 will also be referred to as "risk notification optimization processing".

[0154] Figure 5This diagram schematically illustrates the concept of risk notification optimization processing in the risk notification setting unit 64. Furthermore, the following description of the risk notification optimization processing flow will take, for example, a scenario where the traffic participant risk prediction unit 625 predicts a contact risk between two parties (i.e., the first set object (moving vehicle) and the second set object (moving vehicle)). However, the invention is not limited to this. Since this can be easily generalized to situations where the risk of contact between either party is a pedestrian, or where the risk of contact between all three parties is predicted, the description is omitted.

[0155] in addition, Figure 5 The left side schematically illustrates the change in how risk notifications operate in the first set of objects. Figure 5 The right side schematically illustrates the change in how risk notifications operate in the second setting. Additionally, Figure 5 The two topmost arrows conceptually represent the time elapsed from the initial prediction of a contact risk by the traffic participant risk prediction unit 625 until the first and second pre-defined objects come into contact, i.e., the collision prediction time. However, these two arrows only conceptually represent the collision prediction time, so it does not mean that the risk notification optimization process in the risk notification setting unit 64 cannot be performed if the collision prediction time is not explicitly calculated in the traffic participant risk prediction unit 625. The risk notification optimization process in the risk notification setting unit 64 can be performed starting from the stage before the traffic participant risk prediction unit 625 calculates the explicit collision prediction time. Furthermore, in Figure 5 The diagram illustrates the situation where, at the point when the traffic participant risk prediction unit 625 first predicts that an exposure risk will occur, the risk notifications for the first and second predefined objects are set to off (i.e., the risk notification setting value is "0").

[0156] When the traffic participant risk prediction unit 625 predicts that a contact risk involving multiple auxiliary objects will occur within the monitored area, the risk notification setting unit 64 first, based on the content of the contact risk predicted by the traffic participant risk prediction unit 625, notifies the multiple predicted parties (in the area) related to the contact risk. Figure 5 In the example, priorities are assigned to the first and second target objects. As detailed later, this priority determines the order in which risk notifications (especially risk notifications in the care notification mode) are enabled; risk notifications for higher-priority target objects are enabled earlier than those with lower priority. Furthermore, in Figure 5 The diagram shows a scenario where the priority of the first set object is set higher than that of the second set object.

[0157] Here, the risk notification setting unit 64 sets a priority for each target in a way that avoids the manifestation or occurrence of the predicted contact risk without causing traffic flow disruption among these targets. More specifically, the risk notification setting unit 64 can also identify the risk initiator that will induce the contact risk from among multiple predicted parties related to the contact risk by referring to the prediction results of the traffic participant risk prediction unit 625, the identification information obtained by the target traffic area identification unit 60, and the driver subject information obtained by the driver subject information acquisition unit 61, and set a higher priority for the risk initiator compared to other predicted parties. By setting a higher priority for this risk initiator, the risk notification can be set to open earlier than for other targets, allowing the risk initiator's actions to be corrected before the risk notification is set to open for other targets, thus avoiding the manifestation or occurrence of the initially predicted contact risk.

[0158] Here, the term "risk inducer" can be executors of actions that have a high probability of inducing the aforementioned contact risk (e.g., sudden acceleration, sudden deceleration, sudden lane change, cutting in line, shortening the distance between vehicles, continuously crossing lanes, swerving, driving against traffic, ignoring signals, driving at a speed higher than or exceeding the prescribed speed limit compared to surrounding moving objects, driving at a speed lower than or exceeding the prescribed speed limit compared to surrounding moving objects, driving at a speed higher than or exceeding the prescribed speed limit, driving at a speed lower than or exceeding the prescribed speed limit, and obstructing the movement of surrounding traffic participants).

[0159] Furthermore, the risk notification setting unit 64 can also set priorities based on the traffic environment of each target. More specifically, compared to other target parties, those target parties situated in traffic environments where it is difficult for them to recognize the presence of other target parties are given higher priority, so that the risk notification is set to be activated earlier than other target parties. This improves the cognitive ability of target parties with higher priority settings, thus avoiding the manifestation or occurrence of the initially predicted contact risks.

[0160] The risk notification setting unit 64, based on the traffic participant risk prediction unit 625's prediction of a potential contact risk, sets priorities for each target object through the aforementioned process, and then determines at a predetermined period whether the initially predicted contact risk has materialized. More specifically, the risk notification setting unit 64 determines that the contact risk has not materialized (i.e., the contact risk is potential) if, for example, the traffic participant risk prediction unit 625 predicts a contact risk and the collision prediction time for that contact risk is above a predetermined manifestation threshold (including cases where the traffic participant risk prediction unit 625 has not calculated a specific collision prediction time). Conversely, the risk notification setting unit 64 determines that the contact risk has materialized if, for example, the collision prediction time calculated by the traffic participant risk prediction unit 625 becomes less than the aforementioned manifestation threshold. Here, the threshold for the collision prediction time is the manifestation threshold. Figure 5 As shown, it is set to operate over a wider range than ADAS, in other words, it takes longer than the collision prediction time when the driver assistance ECUs mounted on each moving body begin to execute collision mitigation braking control or collision avoidance steering control.

[0161] Additionally, during the period before the initially predicted exposure risk is determined to have materialized—that is, the period during which the exposure risk is determined to be a potential risk—the risk notification setting unit 64 prioritizes the objects with higher priority (in...). Figure 5 In the example, the risk notification in the care notification mode is initiated first for the first set target. That is, the risk notification setting unit 64 sets the risk notification setting value to "1" or "3" first for the set targets with high priority. Thus, the driver of the set target who receives the risk notification in this care notification mode can recognize the moving objects that may come into contact with the vehicle (in the context of the vehicle). Figure 5 In the example, the existence of a second pre-defined object may sometimes lead to actions to avoid predicted contact risks. When a driver who has received such a risk notification takes actions to avoid contact risks, the traffic participant risk prediction unit 625 may sometimes predict that a contact risk that was initially predicted to occur will not occur before it manifests.

[0162] Additionally, the risk notification setting unit 64 targets settings with low priority (in...) Figure 5In the example (the second set object), after a predetermined time following the initiation of a risk notification in the care notification mode for a set object with a high priority, the risk notification setting unit 64 initiates a risk notification in the care notification mode for a set object with a low priority, respectively. That is, after a predetermined time following the initiation of a risk notification setting value of "1" or "3" for a set object with a high priority, the risk notification setting unit 64 sets the risk notification setting value of "1" or "3" for a set object with a low priority. Furthermore, in order not to cause traffic flow disruption for set objects with low priority, the risk notification setting unit 64 may also refrain from issuing a risk notification in the care notification mode for set objects with low priority until the contact risk manifests. In addition, by issuing a risk notification in the care notification mode to set objects with high priority first, as described above, the occurrence of contact risk may sometimes be avoided. Therefore, the risk notification setting unit 64 may also initiate a risk notification in the care notification mode for set objects with low priority if, after initiating a risk notification in the care notification mode for a set object with a high priority, the driver of that set object has not taken any action to avoid contact risk after a predetermined time.

[0163] Furthermore, after determining that the initially predicted contact risk has materialized, the risk notification setting unit 64 initiates risk notification in simulation mode for all parties involved in the prediction related to the contact risk. That is, after determining that the contact risk has materialized, the risk notification setting unit 64 sets the risk notification setting value for all parties involved in the prediction to either "2" or "4". As described above, in simulation notification mode, the shorter the collision prediction time, the higher the notification intensity. Therefore, all parties involved in the prediction of the contact risk can feel a sense of crisis regarding the approaching contact risk and take actions to avoid it.

[0164] return Figure 2 The coordination assistance information notification unit 65 generates coordination assistance information to promote communication with surrounding traffic participants and the identification of the surrounding traffic environment for each traffic participant identified as an assistance object by the object traffic area identification unit 60, based on the identification information obtained by the object traffic area identification unit 60, the driver subject information obtained by the driver subject information acquisition unit 61, the prediction results related to the monitored area obtained by the traffic participant risk prediction unit 625, the information related to the risk level of each local area obtained by the area risk prediction unit 620 (hereinafter also referred to as "risk area information"), the information related to the sanitation setting value set by the sanitation notification setting unit 63, and the information related to the risk notification setting value set by the risk notification setting unit 64. The generated coordination assistance information is sent to each traffic participant.

[0165] Here, the coordination assistance information notification unit 65 sends coordination assistance information to the auxiliary objects that exist in the monitoring area (i.e., high-risk area) of the traffic participant risk prediction unit 625 among the multiple auxiliary objects existing in the entire traffic area. The information includes information related to the risk notification setting value set based on the prediction result of the traffic participant risk prediction unit 625 and risk area information generated based on the estimation result of the area risk prediction unit 620.

[0166] In addition, the coordination auxiliary information notification unit 65 sends coordination auxiliary information containing risk area information generated based on the estimation results of the regional risk prediction unit 620 to auxiliary objects that exist in low-risk areas among the multiple auxiliary objects in the entire object traffic area and are not extracted as high-risk areas by the regional risk prediction unit 620.

[0167] The traffic safety assistance system 1 and its learning method according to this embodiment achieve the following effects.

[0168] (1) The traffic safety assistance system 1 includes: a target traffic area identification unit 60, which identifies traffic participants (including people and moving bodies) and their traffic environments within the target traffic area 9, and acquires identification information related to these identified objects; a prediction unit 62, which predicts the risks in the target traffic area 9 based on the identification information; and a coordinated assistance information notification unit 65, which sends assistance information generated based on the identification information and the prediction results of the prediction unit 62 to assistance objects determined from multiple traffic participants in the target traffic area 9. Furthermore, in the prediction unit 62, at least one of the multiple local areas after the target traffic area 9 is subdivided is extracted as a high-risk area by the area risk prediction unit 620, and the future risks of traffic participants in the high-risk area are predicted by the traffic participant risk prediction unit 625. Here, in the area risk prediction unit 620, by using information obtained through statistical processing of identification information when extracting high-risk areas from multiple local areas, high-risk areas can be extracted with less overhead compared to directly using a large amount of identification information related to the identified objects in the target traffic area 9. Furthermore, in the traffic participant risk prediction unit 625, by using information associated with the monitoring area from the identification information related to the identified objects in the entire target traffic area 9 when designating high-risk areas as monitoring areas and predicting the future risks of traffic participants in those monitoring areas, the future risks of traffic participants can be predicted with less overhead compared to directly using a large amount of identification information related to the identified objects in the target traffic area 9. Therefore, according to the traffic safety assistance system 1, appropriate assistance information generated based on prediction results can be provided to traffic participants in high-risk areas in real time, thus improving the safety, convenience, and smoothness of traffic in the target traffic area 9.

[0169] (2) In the traffic safety assistance system 1, the traffic participant risk prediction unit 625 does not predict the future risk of traffic participants in low-risk areas that were not extracted as high-risk areas by the area risk prediction unit 620 among multiple local areas. Therefore, according to the traffic safety assistance system 1, the traffic participant risk prediction unit 625 can reduce the computational load compared to performing prediction processing on all local areas. Furthermore, according to the traffic safety assistance system 1, by reducing the number of local areas to be predicted, the computational load can be reduced, thus improving the prediction accuracy of the risk of traffic participants in high-risk areas. Therefore, according to the traffic safety assistance system 1, appropriate auxiliary information generated based on the highly accurate prediction results of the traffic participant risk prediction unit 625 can be provided to traffic participants in high-risk areas in real time, further improving the safety, convenience, and smoothness of traffic in the target traffic area 9.

[0170] (3) In the traffic safety assistance system 1, the regional risk prediction unit 620 estimates the risk level for each local area and extracts high-risk areas from multiple local areas based on the estimation results of the risk level. Furthermore, the coordination assistance information notification unit 65 sends coordination assistance information containing information related to the risk notification setting value generated in the risk notification setting unit 64 based on the more detailed prediction results of the traffic participant risk prediction unit 625 to the auxiliary objects existing in the high-risk areas among the multiple auxiliary objects in the entire target traffic area 9. This improves the safety, convenience, and smoothness of traffic for traffic participants in high-risk areas. Additionally, the coordination assistance information notification unit 65 sends coordination assistance information containing risk area information generated based on the estimation results of the regional risk prediction unit 620 for each local area among the multiple auxiliary objects in the entire target traffic area 9 that exist in low-risk areas outside the high-risk areas. This also improves the safety, convenience, and smoothness of traffic for traffic participants in low-risk areas. Thus, in the traffic safety assistance system 1, by changing the content of the coordination assistance information according to the risk level of each local area, the safety, convenience and smoothness of traffic for traffic participants in the entire target traffic area 9 can be improved.

[0171] (4) In the first learning method, learning data is prepared by using input data for the macro-risk estimation model 623 generated based on identification information, and the output of the micro-risk estimation model 628 when the identification information is input to it. This learning data is then used to learn the macro-risk estimation model 623. In general model learning, positive solution data is needed to evaluate the correctness of the model's output. In contrast, in the first learning method, the output of the micro-risk estimation model 628 can be used as the learning data for learning the macro-risk estimation model 623, thus allowing for the construction of a high-precision macro-risk estimation model 623 in a relatively simple way. Therefore, according to the first learning method, the accuracy of the macro-risk estimation model can be improved by providing coordination assistance information to various traffic participants.

[0172] (5) In the second learning method, learning data is prepared by using input data for the macro-risk estimation model 623 generated based on the first identification information obtained in the first period, and positive solution data for the output of the micro-risk estimation model 628 generated based on the second identification information obtained in the second period after the first period. This learning data is then used to learn the overall model combining the macro-risk estimation model 623 and the micro-risk estimation model 628. Therefore, according to the second learning method, the second identification information obtained in the second period after the first period can be used as data to evaluate the correctness of the output of the overall model when the first identification information is input, thus improving the accuracy of the overall model combining the macro-risk estimation model 623 and the micro-risk estimation model 628. Therefore, according to the traffic safety assistance system 1, the accuracy of the overall model can be improved by providing coordination assistance information to each traffic participant.

[0173] The above description pertains to one embodiment of the present invention, but the invention is not limited thereto. Appropriate modifications to the details may be made within the scope of the spirit of the invention.

[0174] Figure Labels

[0175] 1: Traffic Safety Assistance System

[0176] 9: Target Traffic Area

[0177] 2: Four-wheeled vehicles (mobile vehicles, traffic participants)

[0178] 3: Motorcycles (mobile vehicles, traffic participants)

[0179] 4: Pedestrians (people, traffic participants)

[0180] 6: Coordination Auxiliary Devices

[0181] 60: Object Traffic Area Identification Unit (Identification Method)

[0182] 61: Driver Entity Information Acquisition Unit

[0183] 62: Prediction Unit (Prediction Method)

[0184] 620: Regional Risk Prediction Unit (Regional Risk Prediction Methods)

[0185] 621: Statistical Processing and Computation Department

[0186] 623: Macroeconomic Risk Prediction Model

[0187] 624: High-risk area extraction department

[0188] 625: Traffic Participant Risk Prediction Unit (Traffic Participant Risk Prediction Methods)

[0189] 626: Surveillance Area Information Extraction Department

[0190] 628: Micro-risk estimation model

[0191] 63: Improve the notification setting unit

[0192] 64: Risk Notification Setting Unit

[0193] 65: Coordination and Auxiliary Information Notification Unit (Sending Method)

[0194] 67: Traffic Environment Database

[0195] 68: Driving History Database

Claims

1. A traffic safety assistance system, characterized in that include: Identification methods are used to identify traffic participants, including people or moving bodies, within the target traffic area and the traffic environment of each traffic participant, and to obtain identification information related to these identified objects. The predictive method, based on the aforementioned identification information, predicts the risks in the aforementioned target traffic areas; and, The sending means sends auxiliary information generated based on the aforementioned identification information and the prediction results of the aforementioned prediction means to an auxiliary object determined from multiple traffic participants in the aforementioned target traffic area; and, The aforementioned forecasting methods include: The regional risk prediction method is based on information obtained by statistically processing the aforementioned identification information related to the entire aforementioned target traffic area, and extracting at least one of the multiple local areas after subdividing the aforementioned target traffic area as a high-risk area. and, The traffic participant risk prediction method uses the high-risk area extracted by the aforementioned regional risk prediction method as the monitoring area, extracts the information related to the traffic participants and their traffic environment in the aforementioned monitoring area from the aforementioned identification information related to the entire aforementioned target traffic area, and predicts the future risk of each traffic participant in the aforementioned monitoring area based on the extracted information.

2. The traffic safety assistant system of claim 1, wherein, The aforementioned traffic participant risk prediction method does not perform the aforementioned prediction process on local areas that were not identified as high-risk areas by the aforementioned regional risk prediction method among the multiple aforementioned local areas.

3. The traffic safety auxiliary system according to claim 1, wherein, The aforementioned regional risk prediction methods estimate the risk level for each of the aforementioned local areas separately. The aforementioned sending method sends first auxiliary information generated based on the prediction results of the aforementioned traffic participant risk prediction method to the auxiliary objects that exist in the aforementioned high-risk area among the multiple aforementioned auxiliary objects, and sends second auxiliary information generated based on the inference results of the aforementioned regional risk prediction method to the auxiliary objects that exist in the low-risk area outside the aforementioned high-risk area.

4. The traffic safety assist system of claim 2, wherein, The aforementioned regional risk prediction methods estimate the risk level for each of the aforementioned local areas separately. The aforementioned sending method sends first auxiliary information generated based on the prediction results of the aforementioned traffic participant risk prediction method to the auxiliary objects that exist in the aforementioned high-risk area among the multiple aforementioned auxiliary objects, and sends second auxiliary information generated based on the inference results of the aforementioned regional risk prediction method to the auxiliary objects that exist in the low-risk area outside the aforementioned high-risk area.

5. A learning method for a traffic safety assistance system, wherein the traffic safety assistance system is the traffic safety assistance system according to any one of claims 1 to 4, and the learning method for the traffic safety assistance system is characterized in that... The aforementioned regional risk prediction method uses a macro-risk estimation model to extract the aforementioned high-risk areas. When the macro-risk estimation model is input with information obtained through statistical processing of the aforementioned identification information, it outputs the risk level of each of the aforementioned local areas. The aforementioned traffic participant risk prediction method uses a micro-risk estimation model to predict the future risk of traffic participants in the aforementioned high-risk areas. When the micro-risk estimation model is input with information related to a specified local area from the aforementioned identification information, it will output the future risk of traffic participants in that local area. The learning method of the traffic safety assistance system includes the following steps: Learning data is prepared by using input data for the aforementioned macro-risk presumption model generated based on the aforementioned identification information, and the output of the aforementioned micro-risk presumption model when the aforementioned identification information is input into the aforementioned micro-risk presumption model. and, The aforementioned learning data is used to learn the aforementioned macroeconomic risk estimation model.

6. A learning method for a traffic safety assistance system, wherein the traffic safety assistance system is the traffic safety assistance system according to any one of claims 1 to 4, and the learning method for the traffic safety assistance system is characterized in that... The aforementioned regional risk prediction method uses a macro-risk estimation model to extract the aforementioned high-risk areas. When the macro-risk estimation model is input with information obtained through statistical processing of the aforementioned identification information, it outputs the risk level of each of the aforementioned local areas. The aforementioned traffic participant risk prediction method uses a micro-risk estimation model to predict the future risk of traffic participants in the aforementioned high-risk areas. When the micro-risk estimation model is input with information related to a specified local area from the aforementioned identification information, it will output the future risk of traffic participants in that local area. The learning method of the traffic safety assistance system includes the following steps: Learning data is prepared by using input data for the aforementioned macro-risk presumption model generated based on first identification information obtained in a specified first period, and positive solution data for the output of the aforementioned micro-risk presumption model generated based on second identification information obtained in a second period after the aforementioned first period. and, The aforementioned learning data is used to learn the overall model that combines the aforementioned macro-risk estimation model and the aforementioned micro-risk estimation model.

Images