JP2017097477A - Insurance business support system and insurance business support method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To make it possible to easily introduce infrastructure for telematics insurance and to propose improvement in preferably precise operation to an insurance contractor.SOLUTION: An insurance business support system 10 includes a server device 100 having a storage device 101 for storing a contract information table 125 and an accident information table 126, and an arithmetic device 104 for receiving operation information collected by a portable terminal 200 while a prescribed signal is received from a prescribed device 50 in an insurance contract vehicle 20 from the portable terminal 200, modeling a relationship between an accident having prescribed or more significance with generation of an accident among each combination of attribution of each contractor with each operation manipulation and the present accident based on the operation information, the contract information table 125, and the accident information table 126 and calculating insurance payment risk of each contractor based on the model.SELECTED DRAWING: Figure 2

Description

  The present invention relates to an insurance business support system and an insurance business support method, and more specifically, to a technology that makes it easy to introduce an infrastructure for telematics insurance and proposes a driving improvement with high accuracy for an insurance policyholder. .

  There is a car insurance that estimates the insurance payment risk for each driver based on the driving tendency of the car by the driver, and sets insurance premiums for each. This car insurance is a so-called telematics insurance represented by PHYD (Pay How You Drive).

  As a related art related to such automobile insurance, for example, an automobile insurance premium that acquires a driving state of a vehicle based on a driving operation of the driver and determines a safe driving level of the driver based on the acquired driving state. A setting system, a driver identifying unit that obtains an identifier of the driver, a device that obtains first information relating to a driving state of the vehicle, and obtains second information relating to a traveling environment of the vehicle, and the obtaining A terminal that transmits the driver identifier, the first information, and the second information; and the driver identifier, the first information, and the second information are received, and the first information and the second information are received. And a computer for determining a safe driving level for each of the driver identifiers based on the information of 2, wherein the computer is configured to perform the safe driving based on the safe driving level. An automobile insurance premium setting system including an insurance premium calculation unit that generates an insurance premium calculation formula corresponding to a bell and calculates a driver's automobile insurance premium for each driver identifier using the generated calculation formula (Patent Document 1) Have been proposed).

JP 2009-128486 A

  The telematics insurance mentioned above is rapidly spreading in the West. On the other hand, in Japan, the balance between telematics infrastructure development costs and premium income is difficult to achieve, and the spread is not in full swing. In addition, the lack of telematics infrastructure means that it is difficult to identify driving trends with good accuracy for each driver. Therefore, the so-called MHYD (Management How You Drive) technology that proposes improvement of driving as well as setting insurance premiums according to the driving tendency of each driver has not been established.

  SUMMARY OF THE INVENTION An object of the present invention is to provide a technology that makes it easy to introduce an infrastructure for telematics insurance and makes it possible to propose driving improvements with good accuracy for policyholders.

The insurance business support system of the present invention that solves the above problems includes a contract information table that stores information on each car insurance contract, and an accident information table that stores information on each accident caused by the car insurance contractor. A storage device for storing, and driving information collected by the mobile terminal during reception of the predetermined signal from the predetermined device in the insurance contract vehicle is received from the mobile terminal, the received driving information, the contract information table, and the accident Based on the information table, the relationship between the accident and which of the combinations of the attributes of each contractor who is a vehicle driver and each driving operation is greater than or equal to a predetermined value and the corresponding accident is determined by a predetermined statistical process. A server device is provided that includes an arithmetic device that models and calculates insurance payment risk of each contractor based on the model.

  In addition, the insurance business support method of the present invention stores a contract information table storing information related to each contract of car insurance and an accident information table storing information of each accident caused by the contractor of the car insurance. The server device provided with the device receives the driving information collected by the portable terminal during reception of the predetermined signal from the predetermined device in the insurance contract vehicle from the portable terminal, the received driving information, the contract information table, and Based on the accident information table, the relationship between the corresponding accident and the significance of the occurrence of an accident among the combinations of the attributes of each contractor who is a vehicle driver and each driving operation is determined according to predetermined statistics. It is characterized by modeling by processing, and calculating the insurance payment risk of each policyholder based on the model.

  According to the present invention, it is possible to facilitate the introduction of an infrastructure for telematics insurance and to propose a driving improvement with good accuracy for the policyholder.

It is a network block diagram including the insurance business support system of this embodiment. It is a figure which shows the hardware structural example of the data center server in this embodiment. It is a figure which shows the hardware structural example of the beacon terminal in this embodiment. It is a figure which shows the hardware structural example of the portable terminal in this embodiment. It is a figure which shows the data structural example of the contract information table of this embodiment. It is a figure which shows the example of a data structure of the accident information table of this embodiment. It is a figure which shows the data structural example of the sensor information table of this embodiment. It is a figure which shows the data structural example of the driving | operation information table of this embodiment. It is a figure which shows the example of a data structure of the analysis result table of this embodiment. It is a figure which shows the data structural example of the improvement difference table of this embodiment. It is a figure which shows the example 1 of a flow in the insurance business support method of this embodiment. It is a figure which shows the example 2 of a flow in the insurance business support method of this embodiment. It is a figure which shows the example 3 of a flow in the insurance business support method of this embodiment. It is a figure which shows the example 4 of a flow in the insurance business support method of this embodiment. It is a figure which shows the example 5 of a flow in the insurance business support method of this embodiment.

--- Network configuration ---

  Embodiments of the present invention will be described below in detail with reference to the drawings. FIG. 1 is a network configuration diagram including an insurance business support system 10 of the present embodiment. An insurance business support system 10 shown in FIG. 1 is a computer system that makes it easy to introduce an infrastructure for telematics insurance and proposes a driving improvement with high accuracy to an insurance policyholder.

  As illustrated in FIG. 1, the insurance business support system 10 of the present embodiment includes a data center server 100 (server device) and a mobile terminal 200 that are connected via a network 5 so that data communication is possible. . Among these, the data center server 100 is a server device operated by an automobile insurance company or an insurance business support service providing organization, and is a main execution subject of an insurance business support method.

The mobile terminal 200 is a terminal owned by a driver of the insurance contract vehicle 20 for car insurance, that is, a contractor. As specific examples of the mobile terminal 200, a smartphone, a tablet terminal, a feature phone, and the like can be assumed. However, in any device, an interface for short-range wireless communication executed between the observation device such as its own position and acceleration, and a beacon terminal 50 described later, and a communication interface corresponding to the protocol of the network 5 are provided. Shall be provided. The mobile terminal 200 may be provided to the contractor by an insurance company or an insurance business support service provider. In this case, a mode of transferring or lending to the provider is assumed.

  The above-described beacon terminal 50 is a device installed in the insurance contract vehicle 20, and continues to transmit short-range wireless communication signals at predetermined time intervals at least during vehicle operation. The concept of the beacon terminal 50 includes a vehicle recorder such as a so-called drive recorder or a car navigation system. The drive recorder is a device that continuously captures the surroundings of the vehicle (mainly the forward field of view) while the insurance contract vehicle 20 to be installed is in operation and stores the captured data in an appropriate storage device.

The data center server 100 is connected to the insurance company server 300 via the network 5 so that data communication is possible. This insurance company server 300 is a server of an insurance company that sells and manages the above-mentioned automobile insurance. The insurance company server 300 distributes information about each contract of automobile insurance, that is, contract information, and information on accidents caused by each contractor, that is, accident information, to the data center server 100 via the network 5. Receiving this, the data center server 100 stores the contract information in the contract information table 125 and stores the accident information in the accident information table 126.
--- Hardware configuration ---

  The hardware configuration of each device constituting the insurance business support system 10 is as follows. FIG. 2 is a diagram illustrating a hardware configuration example of the data center server 100 which is a server device of the present embodiment. The data center server 100 is held in a storage device 101 composed of an appropriate nonvolatile storage element such as an SSD (Solid State Drive) or a hard disk drive, a memory 103 composed of a volatile storage element such as a RAM, and the storage device 101. The program 102 is read out to the memory 103 and executed to perform overall control of the apparatus itself, and is connected to the arithmetic device 104 such as a CPU for performing various determinations, calculations and control processing, and to the network 5 and is responsible for communication processing with other devices. A network interface 105.

  The storage device 101 stores at least a contract information table 125, an accident information table 126, a sensor information table 127, an operation information table 128, an analysis result table 129, and an improvement difference table 130 in addition to the program 102 described above. Has been. Details of the data structure of these tables will be described later.

  FIG. 3 is a diagram illustrating a hardware configuration example of the beacon terminal 50 in the present embodiment. The beacon terminal 50 in the present embodiment includes a storage device 51 configured with an appropriate nonvolatile storage element such as a flash memory, a memory 53 configured with a volatile storage element such as a RAM, and a program 52 held in the storage device 51. 53 and the like, and performs overall control of the apparatus itself by executing it by reading to 53, etc., and an arithmetic unit 54 such as a CPU for performing various control processes, and short-range wireless communication for performing wireless communication processing with other apparatuses in accordance with a predetermined short-range wireless communication protocol An interface 56 and an imaging device 57 such as a digital video camera are provided. In this embodiment, since the beacon terminal 50 is assumed as a drive recorder, the configuration includes the imaging device 50. However, in the present invention, the configuration may not include the imaging device 50.

  In addition to the program 52 described above, the storage device 51 stores at least the value of the beacon terminal 50, in other words, the beacon ID 55 that uniquely identifies the policyholder.

  FIG. 4 is a diagram illustrating a hardware configuration example of the mobile terminal 200 in the present embodiment. The mobile terminal 200 in the present embodiment is a terminal possessed by a car insurance policyholder, and assumes a smartphone or the like as described above. The portable terminal 200 reads out a storage device 201 configured by an appropriate nonvolatile storage element such as a flash memory, a memory 203 configured by a volatile storage element such as a RAM, and a program 202 held in the storage device 201 to the memory 203. An arithmetic device 204 such as a CPU that performs various control processing and performs overall control of the device itself, an input / output device 205 such as a touch panel, a network interface 207 that is connected to the network 5 and performs communication processing with other devices, A short-range wireless communication interface 208 that performs wireless communication processing with other devices according to a predetermined short-range wireless communication protocol, and an observation device 210 that observes events corresponding to driving information of the insurance contract vehicle 20 are provided.

  Note that the above-described observation device 210 includes at least a GPS unit 211 that observes the current location of the own device and an acceleration sensor 212 that senses acceleration acting on the own device. Furthermore, a gyro sensor 213, an azimuth angle (magnetism) sensor 214, an atmospheric pressure sensor 215, and a temperature sensor 216 may be included.

  The above-described program 202 held by the storage device 201 includes a driving information collection application 2021 that acquires each observation data of the GPS unit 211 and the acceleration sensor 212 that constitute the above-described observation device 210.

  The driving information collection application 2021 is, for example, a smartphone application provided by an automobile insurance company, and holds a beacon ID 2022 and a user ID 2023. It is assumed that each value of the beacon ID 2022 and the user ID 2023 is notified in advance to each contractor from the car insurance company along with the insurance contract.

The contractor obtains the values of the beacon ID 2022 and the user ID 2023 through the predetermined screen of the driving information collection application 2021 when the driving information collection application 2021 is installed on the mobile terminal 200 or at a predetermined timing after the installation. Set in the collection application 2021. Alternatively, when the car insurance company distributes the driving information collection application 2021 and requests a login operation of the contractor on a member site for an insurance contractor and the login is successful, The user ID and the beacon ID of the beacon terminal 50 lent to the contractor may be set in the driving information collection application 2021, and then downloaded and distributed.
--- Data structure example ---

  Next, tables used by the insurance business support system 10 of this embodiment will be described. FIG. 5 shows an example of the contract information table 125 in the present embodiment. The contract information table 125 is a table that stores information on each contract of car insurance. The data structure is based on the use car type, class, region, new car discount, age condition, insured person age classification, which is the contract contents of the corresponding insurance contract, using the user ID uniquely identifying the user who is the contractor and the ticket number as a key. , Purpose of use, license coefficient, non-fleet (NF) grade, driver limited, underwriter insurance company, contractor's e-mail address, ID of mobile terminal 200, and UUID which is beacon ID of beacon terminal 50 Is the body.

FIG. 6 shows an example of the accident information table 126 in the present embodiment. The accident information table 126 is a table that stores information on each accident caused by a car insurance contractor. The data structure includes an accident ID that uniquely identifies each accident as a key, a user ID indicating the contractor who has caused the accident, accident date and time, collateral, accident type, accident cause, driver age, driver classification, Victim injury classification, negligence rate, accident location, number of accidents, total vehicle damage division classification, sequelae disability classification, insured classification, injured classification, damage classification, accident status, incidental expenses, and payment It is a collection of records consisting of data such as insurance money.

  FIG. 7 shows an example of the sensor information table 127 in the present embodiment. The sensor information table 127 is a table in which the data center server 100 obtains and stores observation data obtained by the driving information collection application 2021 from the observation device 210 of the portable terminal 200 from the portable terminal 200. The data structure is such that a user ID that uniquely identifies a contractor is used as a key, a terminal ID that uniquely identifies the mobile terminal 200 held by the contractor, a data acquisition date and time, a data reception date and time, a trip ID (once ID that uniquely identifies the drive itinerary), terminal coordinate system triaxial acceleration, GPS coordinates, terminal coordinate system triaxial angular velocity, terminal coordinate system triaxial magnetic field, barometric pressure, and temperature. .

  FIG. 8 shows an example of the driving information table 128 in the present embodiment. The driving information table 128 is a table that stores driving information obtained by the data center server 100 performing appropriate preprocessing (described later) on the observation data stored in the sensor information table 127 described above. The data structure includes a user ID that uniquely identifies the contractor as a key, a terminal ID that uniquely identifies the mobile terminal 200 held by the contractor, data acquisition date and time, data reception date and time, trip ID, and terminal coordinates. System triaxial acceleration, GPS coordinates, terminal coordinate system triaxial differential acceleration, terminal coordinate system triaxial angular velocity, terminal coordinate system triaxial magnetic field, barometric pressure, temperature, road type, road gradient, installation angle, and automobile coordinate system triaxial acceleration This is a set of records composed of data such as the car coordinate system three-axis magnetism and the vehicle speed (average).

  FIG. 9 shows an example of the analysis result table 129 in the present embodiment. The analysis result table 129 is a table that stores insurance payment risk information of each contractor in the most recent future obtained by the data center server 100 performing a predetermined process. The data structure is a collection of records composed of data such as start date and time, end date and time, risk level, predicted loss amount for the next year, evaluation axis 1 to evaluation axis N, using trip ID and user ID as keys. .

FIG. 10 shows an example of the improvement difference table 130 in the present embodiment. The improvement difference table 130 is a table that stores information related to risk reduction in accordance with the improvement of driving operation in the contractor after the driving improvement guidance information is presented from the data center server 100. The data structure consists of a trip ID and user ID as keys, and a comparison target trip ID (trip of a similar itinerary of the same contractor, that is, a trip of a similar itinerary before driving improvement guidance information is presented), recording date and time, risk improvement This is a set of records composed of data such as degree, predicted loss improvement degree, evaluation axis 1 to evaluation axis N. In the present embodiment, the “similar itinerary” is a comparison target, but the comparison target is not limited to this.
--- Flow example 1 ---

  Hereinafter, the actual procedure of the insurance business support method in the present embodiment will be described with reference to the drawings. Various operations corresponding to the insurance business support method described below are realized by a program that each device in the insurance business support system 10 reads into a memory or the like and executes. And this program is comprised from the code | cord | chord for performing the various operation | movement demonstrated below.

  FIG. 11 is a diagram showing a flow example 1 of the insurance business support method in the present embodiment. Here, first, processing related to collection of driving information by the mobile terminal 200 and transmission to the data center server 100 will be described. Therefore, the execution subject of each step in the flow is the mobile terminal 200. Hereinafter, the mobile terminal 200 is assumed to be a smartphone and is referred to as a smartphone 200.

  Here, it is assumed that a car insurance contractor carries the smartphone 200 and enters the insurance contract vehicle 20. The beacon terminal 50 installed in the insurance contract vehicle 20 is supplied with operating power from the insurance contract vehicle 20 together with a vehicle starting operation such as engine start of the insurance contract vehicle 20 or power-on at the time of key-in. A communication signal is transmitted. It is assumed that the intensity of the signal transmitted by the beacon terminal 50 is controlled to a predetermined intensity with the distance from the beacon terminal 50 to the vicinity of the driver's seat as a communication limit, for example. Therefore, it is avoided that the above signals reach the passenger's smartphone riding in the passenger seat or the rear seat in a communicable manner. In addition, about the operating power supply of the beacon terminal 50, there also exists a form supplied with the battery of the insurance contract vehicle 20, or a built-in battery. The short-range wireless communication signal transmission may be configured to always transmit.

  In this flow, the smartphone 200 receives the above-mentioned signal transmitted from the beacon terminal 50 by the short-range wireless communication interface 208 (s110), and starts the driving information collection application 2021 using this as a trigger (s111).

  Next, the smartphone 200 acquires a beacon ID from the signal received from the beacon terminal 50, and performs an authentication process based on a match determination between the acquired beacon ID and the beacon ID 2022 held in advance by the driving information collection application 2021 (s112). .

  When the above-mentioned authentication fails (s112: NO), the smartphone 200 ends the driving information collection application 2021 (s118), and the flow is also ended.

  On the other hand, when the above-described authentication is successful (s112: YES), in the smartphone 200, the driving information collection application 2021 collects respective observation data as sensor information from the GPS unit 211 and the acceleration sensor 212 of the observation device 210, Once stored in the memory 203 (s113).

  Next, the smartphone 200 determines whether the latest observation data obtained by the driving information collection application 2021 from the acceleration sensor 212 of the observation device 210 indicates an excessive acceleration that is equal to or higher than a predetermined reference, thereby causing an accident in the insurance contract vehicle 20. It is determined whether or not a specific event such as the occurrence has occurred (s114).

  As a result of the determination, when the occurrence of a specific event such as an accident in the insurance contract vehicle 20 is detected (s114: YES), the smartphone 200 stores the driving information stored in the memory 203 by the driving information collection application 2021 up to this point. Then, after associating with the user ID held by the driving information collection application 2021, it is immediately transmitted to the data center server 100 (s116). At this time, it is preferable that the smartphone 200 rejects a user operation to the driving information collection application 2021 or the memory 203 in which the driving information collection application 2021 stores the driving information, and transitions its own interface to a locked state. .

  On the other hand, as a result of the above-described determination, if the occurrence of a specific event such as an accident in the insurance contract vehicle 20 is not detected (s114: NO), the smartphone 200 determines whether a predetermined time has elapsed since the execution of step s113, for example. (S115). If it is determined as a result of this determination that a predetermined time has elapsed since the execution of step s113 (s115: YES), the smartphone 200 stores the driving information stored in the memory 203 by the driving information collection application 2021 during this predetermined time. After associating with the user ID held by the driving information collection application 2021, the data is transmitted to the data center server 100 via the network interface 207 (s116). In this case, the data center server 100 receives the operation information and stores it in the sensor information table 127.

  The smartphone 200 determines whether or not the signal from the above-described beacon terminal 50 is continuously received every time the driving information is transmitted in the above-described step s116 (s117). This determination corresponds to the smartphone 200 determining whether or not a signal is received from the beacon terminal 50 at the short-range wireless communication interface 208 at regular intervals.

  As a result of the above determination, when it is determined that the signal from the beacon terminal 50 is continuously received (s117: YES), the smartphone 200 performs the above-described step s113, that is, collecting driving information by the driving information collection application 2021. Re-execute the process.

  On the other hand, if it is determined as a result of the above determination that the signal from the beacon terminal 50 has not been continuously received (s117: NO), the smartphone 200 ends the driving information collection application 2021 (s118), and the flow is also ended. To do.

  The applicant has already obtained a technique for obtaining observation data from sensors such as the GPS unit 211 and the acceleration sensor 212 of the smartphone 200 and analyzing the behavior of the insurance contract vehicle 20 with high accuracy based on the driving information as the observation data. We are developing. By using the smartphone 200 in which such technology is implemented as a driving information collection subject in the insurance contract vehicle 20, it is not necessary to introduce a dedicated sensor device or communication device other than the smartphone 200.

Further, by performing authentication using the beacon ID as a key between the beacon terminal 50 and the smartphone 200, the driver of the insurance contract vehicle 20 is a car insurance contractor or a family specified in the contract. "Who" or "Which car is driven" can be accurately and efficiently specified only by the resources of the smartphone 200 and the beacon terminal 50. In addition, the observation data obtained from the GPS unit 211 and the acceleration sensor 212 in the smartphone 200 and the analysis result thereof have good accuracy, and the reliability of data sufficient for actual operation of the automobile insurance product can be ensured.
--- Flow example 2 ---

  Next, processing after the data center server 100 receives the driving information from the smartphone 200 will be described with reference to the drawings. FIG. 12 is a diagram showing a flow example 2 of the insurance business support method in the present embodiment.

In this case, the data center server 100 receives the operation information transmitted from the smartphone 200 via the network 5 (s120), and stores it in the sensor information table 127.
Next, the data center server 100 sorts the operation information stored in the sensor information table 127 as described above (s121). In this data journal, the observation data included in the driving information is classified into GPS coordinate values obtained by the GPS unit 211 or acceleration data obtained by the acceleration sensor 212 or the like (the gyro sensor 213 or the azimuth angle sensor 214 may be further used). It becomes.

  Subsequently, the data center server 100 identifies the vehicle type classification based on the record of the user ID (included in the driving information received from the smartphone 200) in the contract information table 125 with respect to the acceleration data identified in the above data journal. Thus, the vehicle acceleration is converted by a predetermined analysis algorithm (s123). The existing technology may be adopted for the process of calculating the vehicle acceleration based on the observation data of the acceleration sensor 212.

Further, the data center server 100 executes a filtering process such as removing statistically excessively large or small abnormal values from the vehicle acceleration values obtained in step s123 described above (s124), and the flow Exit. The data center server 100 stores the value of the vehicle acceleration that has undergone this filtering process in the driving information table 128 after associating it with the user ID.

  Further, the data center server 100 collates each GPS coordinate value specified by the above data journal with map information (not shown) held in the storage device 101 in advance, and the corresponding position on the map and the road type of the position. Etc. are specified (s125). The map information is general electronic map data, in which a correspondence between a GPS coordinate value and an actual map position is defined.

Next, the data center server 100 specifies the travel route based on the position information group on the map of the insurance contract vehicle 20 specified in step s125 described above (s126), and ends the flow. Note that the data center server 100 stores the travel route information obtained by this specifying process in the driving information table 128 after associating it with the user ID.
--- Flow example 3 ---

  Next, the relationship between the accident and which of the combinations of the attribute of each contractor who is the vehicle driver and each driving operation is more than a predetermined significance and the corresponding accident is modeled by a predetermined statistical process. Processing will be described with reference to the drawings. FIG. 13 is a diagram showing a flow example 3 of the insurance business support method in the present embodiment.

  In this case, the data center server 100 fetches each storage information of the contract information table 125, the accident information table 126, and the driving information table 128 of the storage device 101 into the memory 103 (s130).

  Next, the data center server 100 executes association of each piece of information fetched into the memory 103 at the above-described step s130, that is, contract information, accident information, and driving information, using the user ID as a key (s131). In addition, the data center server 100 executes appropriate data cleansing and filtering processes for eliminating statistically abnormal values and the like for each piece of information that has been linked as described above (s132). For this data cleansing and filtering, existing technology may be employed. In addition, information that has undergone step s132 is pre-processed information.

  Subsequently, the data center server 100 searches for a feature amount (evaluation axis) (s133). Inputs for this process include pre-processed driving information, contract information (age, gender, vehicle type, etc.), accident information (amount of damage for each accident, accident status: location, time, self-loss, other losses, etc.) Ratio) and external information (weather, traffic jam information, etc.). This external information is acquired by the data center server 100 from a server of a weather forecasting company or a road management organization when executing this step.

  The data center server 100 also performs total extraction of evaluation axis candidates. As one of the extraction methods, a method of executing using the evaluation axis and category items, which have been conventionally used as indices in PHYD insurance, can be assumed. Examples of the evaluation axis include travel characteristics (sudden accelerator frequency, sudden brake frequency, sudden handle frequency, front and rear wobble, left and right wobble), travel time, travel distance, maximum speed, and the like. Examples of category items include travel time zones (early morning: sunrise-1 hour to +1 hour after sunrise, morning: +1 hour to 11:00 after sunrise, noon: 11:00 to 14:00, and noon: 14:00 to 16:00. , Evening: 16:00 to sunset + 1 hour, night: sunset + 1 hour to 10:00, midnight: 10:00 to sunrise-1 hour), driving location (downtown, suburbs, residential area, countryside, face-to-face with no median) General roads, one-way one-lane general roads, one-way two-lane general roads, highways, accident-prone locations), age groups (20s, 30s, etc.), gender, vehicle type / vehicle weight (sedans, minivans, light / Less than 0.5t, 0.5t to less than 1.0t, etc.), past insurance claims, and the like.

  Further, as another method for extracting evaluation axis candidates, it can be assumed that an evaluation axis is automatically extracted by a machine learning method such as so-called deep learning. Specific examples include feature amount extraction by unsupervised learning (eg, feature amount extraction by prior learning with an auto encoder or restricted Boltzmann machine, existing technology).

  In addition, the data center server 100 calculates each accident risk (for example, the number of accidents occurring during a certain period × total damage amount × a negligence ratio / travel distance during a certain period) for each evaluation axis candidate extracted as described above. . In this case, the data center server 100 creates, for example, all combinations of each travel feature and each category item of the evaluation axis candidate according to a preset number of combinations.

  Further, the data center server 100, for each combination created above, performs correlation analysis, search for a combination whose appearance frequency is statistically different between the safety driver and the dangerous driver, and regression analysis (reference: JP 2009-128486 A). Gazette) is applied to analyze the relationship between each combination and accident risk. For example, in the correlation analysis, for the combination of sudden accelerator frequency + travel time zone + age group, analyze the correlation between the sudden accelerator frequency in the evening + 20s and accident risk, and the correlation between accident risk and a certain value (ex. 0.4 or more) ) Is specified as an evaluation axis. In addition, in the search method of a combination of statistically different appearance frequencies compared with a safe driver and a dangerous driver, a safety driver (eg: accident history 0 times) and a dangerous driver (eg: accident history more than once in a certain period) Compare with the driver and search for an evaluation axis having a statistically different appearance frequency. Further, in regression analysis, assuming a regression line indicating the appearance frequency of a certain feature value, it is determined whether or not the feature value to be evaluated is valid when the statistic is equal to or greater than a certain value by t-test with actual data. In this case, a parameter whose t-test statistic is a certain value or more is extracted as an evaluation axis. The data center server 100 specifies an evaluation axis by combining any of the above analysis methods or a plurality of analysis methods.

  In this way, the data center server 100 identifies a feature quantity (evaluation axis) that is significant between the accident risk and the like as a sudden accelerator frequency during traveling on a general road in the early 20s.

  Subsequently, the data center server 100 creates a prediction model for insurance payment risk (s134). The inputs in this case are pre-processed driving information, contract information, external information, accident information, and information on the evaluation axis specified above.

  In this case, the data center server 100 first causes an accident caused by the contractor in the past regarding a combination of preprocessed driving information, contract information, and external information having information that matches the evaluation axis specified in step s133. By identifying the presence or absence of the accident based on the accident information, the significance of the above-mentioned evaluation axis, that is, the relationship with the accident risk is evaluated.

  Subsequently, the data center server 100 creates a prediction model of accident risk (insurance payment risk) by machine learning using the evaluation result obtained above as an input and the accident risk as an output. As examples of machine learning, a support vector machine (SVM), a multilayer perceptron, a Bayesian network, and the like can be assumed.

  Further, the data center server 100 evaluates and verifies the prediction model created so far (s135). In this case, the data center server 100 calculates an accident risk by using, as input, driving information that is not used for learning in the created prediction model. In addition, the data center server 100 determines whether the estimated accuracy of the accident risk calculated here is equal to or higher than a certain value (eg, 95%), and if it is equal to or larger than a certain value (s135: Yes), the target accuracy is predicted to be achieved. Finish creating the model.

  On the other hand, when the above-described estimation accuracy is less than a certain value, the data center server 100 adjusts machine learning parameters. As an example of this parameter adjustment, automatic adjustment combining grid search and cross validation can be adopted. After such adjustment, the data center server 100 evaluates the prediction model again.

  As a result of this evaluation, if the estimation accuracy is equal to or greater than a certain value (s135: Yes), the data center server 100 finishes creating the prediction model, and if the estimation accuracy is less than the certain value (s135: No), step Returning to s133, search / extraction of a new evaluation axis is executed. In this case, the already described evaluation axis search and extraction is executed again.

  The data center server 100 repeats similar processes such as evaluation and model creation using the evaluation axis extracted first and the evaluation axis newly extracted in this way. In this way, the data center server 100 additionally specifies a new evaluation axis. The data center server 100 obtains a learned accident risk prediction model by repeatedly searching / extracting a new evaluation axis and evaluating a prediction model until the accident risk estimation accuracy becomes a certain value or more.

  Note that the data center server 100 updates the accident risk prediction model described above when a certain period (eg, one year) arrives, or updates it according to the necessity of updating. Of these, for example, the number of accidents (or insurance payments) predicted by the corresponding model was compared with the actual number of accidents and the specified error exceeded the specified standard. The state is determined to be necessary and executed. The update process can be handled by executing the above steps s130 to s135 again.

Note that, when the driving information obtained from the smartphone 200 is from a plurality of vehicle drivers regarding the same vehicle, the data center server 100 has the longest driving time indicated by the driving information of each vehicle driver during the predetermined period. A thing is identified as driving information related to the main driver, and is set as a processing target in the step using the driving information among the above steps.
--- Flow example 4 ---

  Next, accident risk (insurance payment risk) estimation processing using the above-described prediction model will be described with reference to the drawings. FIG. 14 is a diagram showing a flow example 4 of the insurance business support method in the present embodiment. The data center server 100 here uses the driving information relating to each contractor (the driver of the insurance contract vehicle 20) and the attribute information (contractor information) of the contractor as inputs, and assesses the risk of the driver (future future). Accident rate and damage cost forecast).

  In the flow of FIG. 14, the data center server 100 reads records relating to each contractor from the contract information table 125 and the operation information table 128 into the memory 103 (s140), and associates the contract information and the operation information for each contractor. (S141).

  Subsequently, the data center server 100 excludes data that is associated with the contractor information and the driving information of each contractor and whose information included in the set is a statistically abnormal value or the like associated with each other in step s141. Cleansing and filtering processing is executed (s142), and a set of pre-processed contract information and driving information is obtained. For such processing, existing technology may be adopted as appropriate.

  Subsequently, the data center server 100 gives the accident risk prediction model as input to the preprocessed contract information and the operation information described above, calculates the accident risk (s143), and stores the calculation result in the analysis result table 129.

Such accident risk calculation processing is executed, for example, at regular intervals (e.g., every trip, every 1 hour of travel time accumulated, 1 day, 1 month, etc.). Of course, the operation of such calculation processing may be changed according to the policy of the car insurance company.

Further, the data center server 100 transmits the accident risk value obtained in step s143 to the insurance company server 300 (s144), and ends the flow.
--- Flow example 5 ---

  Next, a process for transmitting predetermined driving improvement guidance information related to driving operation to the smartphone 200 of each contractor will be described with reference to the drawings. FIG. 15 is a diagram showing a flow example 5 of the insurance business support method in the present embodiment.

  In this case, the data center server 100 first performs a process of identifying a contractor who is a target of information transmission for driving improvement guidance (s150). In this specific process, the data center server 100 determines each policy in the car insurance based on the insurance payment risk of each contractor held in the analysis result table 129, the preprocessed driving information and contract information of each driver. A driver group at an arbitrary risk level that occupies a predetermined percentage or more of the total amount of insurance payment indicated by the accident risk of the driver is specified. In the present embodiment, a person who occupies a predetermined amount or more of the total amount of insurance payment indicated by the accident risk is extracted, but a configuration for identifying a driver group at a predetermined risk level or a driver group exhibiting a predetermined driving characteristic It is good. This also makes it possible to send a safe driving promotion message according to the driver group. In addition, the data center server 100 in step s150 identifies a driving operation whose significance with the occurrence of an accident is defined in the above-described prediction model among the driving operations of the respective contractors forming the contractor group.

  Subsequently, the data center server 100 holds driving improvement guidance information (stored in the storage device 101) that is defined in advance with respect to the driving operation specified above for each contractor's smartphone 200 included in the group of contractors described above. Suppose) is transmitted (s151).

  Subsequently, the data center server 100 detects that the predetermined period has elapsed after transmitting the above driving improvement guidance information (s152), and for the contractor, the driving information used in the last executed step s150 and With respect to the latest driving information obtained most recently, a difference in value between predetermined items (eg, acceleration) is calculated, and a predetermined reference (eg, acceleration is below a predetermined value) with a predetermined driving tendency is calculated. It is determined whether or not the improvement has been made (s153). The data center server 100 stores this determination result in the improvement difference table 130.

  Next, when it is determined as a result of the determination in step s153 that no improvement is observed in the driving tendency even after transmission of the driving improvement guidance information (s154: improvement is required), the data center server 100 provides it at a driving school, etc. A notification of guidance to the learning program that is being sent is transmitted to the corresponding contractor's smartphone 200 (s156), and the process returns to step s153.

  On the other hand, as a result of the determination in step s153, if it is determined that the driving tendency has improved after the transmission of driving improvement guidance information (s154: sufficient improvement), the transmission of a message that praises the improvement effort, insurance Incentive information such as point granting to be applied to the fee is transmitted to the corresponding contractor's smartphone 200 (s155), and the flow ends.

  Although the best mode for carrying out the present invention has been specifically described above, the present invention is not limited to this, and various modifications can be made without departing from the scope of the invention.

  According to this embodiment, it is possible to facilitate the introduction of the infrastructure for telematics insurance, and to propose driving improvement with high accuracy for the policyholder.

At least the following will be clarified by the description of the present specification. That is, in the insurance business support system of the present embodiment, the computing device of the server device is based on the calculated insurance payment risk of each contractor, the driving information of each contractor, and the contract information table. A driver group having a predetermined risk with respect to the total amount of insurance payment according to the insurance claim payment risk of each contractor in the car insurance, and the occurrence of an accident in the model among the driving operations of each contractor constituting the driver group A process for identifying a driving operation for which significance is defined, and a process for transmitting information on predetermined driving improvement guidance regarding the identified driving operation to the terminals of each contractor included in the driver group. Further, it may be executed.

  According to this policy, those who are relatively high in insurance payment risk among policyholders are encouraged to improve driving operations, or those who are low in insurance payment risk are encouraged to maintain safe driving. As a result, it is possible to reduce the total amount of insurance claims paid in the car insurance. This also leads to a reduction in processing load corresponding to the insurance payment business in the insurance company server.

  In the insurance business support system according to the present embodiment, when the computing device of the server device receives driving information from a plurality of vehicle drivers regarding the same vehicle from the mobile terminal during the modeling, the operation device of each vehicle driver The driving time indicated by the information is identified as driving information relating to one or more drivers to be diagnosed during a predetermined period of time, and the identified driving information, the contract information table, and the accident information table, Based on the above, the relationship between the accident and which of the combinations of the attribute of the contractor (one person) and each driving operation is more than a predetermined significance and the corresponding accident is modeled by a predetermined statistical process. There may be. In addition, about the driver | operator in which the driving | running | working data for more than fixed time were collected, the longest thing may be specified as a main driver | operator and you may make it the processing object of driving | operation diagnosis.

  According to this, it corresponds to the situation where a single insurance contract vehicle is used by multiple people such as family members, and for example, the person who has the longest driving time in the most recent year, that is, the insurance premium payment risk can be most affected Driving information related to the person can be used for the above-mentioned model creation. This leads to the effect of calculating the insurance payment risk for the insurance contract vehicle with higher accuracy.

  In the insurance business support system of the present embodiment, a storage device storing a predetermined application for acquiring observation data from at least one of the vehicle position, behavior, and surrounding image, and the predetermined device from the predetermined device When the predetermined signal including the identification ID is received, the predetermined application is activated, and the identification ID included in the received predetermined signal is compared with the identification ID of the predetermined device held in advance by the predetermined application. , Processing for authenticating whether a pair of a predetermined device and a portable terminal for which cooperation is defined in advance with respect to the contractor, and if the authentication is successful, the predetermined signal is not received from the predetermined device. The application collects driving information including the observation data related to the vehicle, and the server collects the collected driving information according to a predetermined opportunity. It may further include a mobile terminal provided with an arithmetic unit for executing processing for transmitting to the location.

  According to this, in conjunction with a predetermined device such as a beacon terminal or a drive recorder installed in an insurance contract vehicle, activation and collection of driving information are performed after confirming the contractor of the car insurance. The collected driving information is distributed to the server device efficiently and quickly via a network via a mobile phone network or the like. In this way, in the server device that receives the distribution of driving information, the driving information related to the genuine contractor is efficiently and quickly acquired, and the above-described model generation and insurance payment risk calculation accuracy are also good. .

In the insurance business support system of the present embodiment, the computing device of the mobile terminal determines whether the predetermined observation data by the observation device matches a predetermined criterion corresponding to occurrence of a predetermined event in the vehicle when transmitting the driving information. As a result of the determination, when the occurrence of a predetermined event in the vehicle is detected, the collected driving information may be transmitted to the server device.

  According to this, for example, when an accident has occurred in an insurance contract vehicle, the mobile terminal can distribute the driving information acquired from its own application or drive recorder to the server device. In the server device, it is possible to reliably acquire driving information especially for an accident that greatly affects the insurance payment risk. In addition, since it is possible to avoid high-frequency driving information distribution from the mobile terminal to the server device, the traffic on the network connecting the mobile terminal and the server device is reduced, and the processing speed associated with communication is improved. Note that the mobile terminal may acquire appropriate observation data such as the latest photographing data from the drive recorder when a predetermined event such as an accident occurs.

  In the insurance business support method of the present embodiment, the server device calculates each insurance policy payment risk of each contractor, the driving information of each contractor, and the contract information table. Significance of accident occurrence in the above model among the policyholders that account for a predetermined percentage or more of the total insurance payment according to the policyholders' insurance risk and the driving operations of each policyholder that constitutes the policyholders. It is further assumed that a process of specifying a prescribed driving operation and a process of transmitting information on predetermined driving improvement guidance related to the specified driving operation to the terminals of each contractor included in the contractor group are further executed. Also good.

  In the insurance business support method of the present embodiment, when the server device receives driving information from a plurality of vehicle drivers regarding the same vehicle from the portable terminal during the modeling, the driving information of each vehicle driver indicates The longest driving time in a predetermined period is identified as driving information related to the main driver, and the attributes of each contractor who is a vehicle driver based on the identified driving information, the contract information table and the accident information table It is also possible to model the relationship between an accident that has a significance greater than or equal to the occurrence of an accident among the combinations of driving operations and the corresponding accident by a predetermined statistical process.

  In the insurance business support method according to the present embodiment, a portable terminal including a storage device storing a predetermined application that acquires observation data from at least one of a vehicle position, behavior, and surrounding video is provided as the predetermined device. When the predetermined signal including the identification ID of the predetermined device is received, the predetermined application is activated, the identification ID included in the received predetermined signal, and the identification ID of the predetermined device held in advance by the predetermined application , And a process for authenticating whether or not a pair of a predetermined device and a portable terminal for which cooperation is prescribed in advance for the contractor is authenticated, and if the authentication is successful, the predetermined signal can be received from the predetermined device. Until there is no longer, collection of driving information including the observation data regarding the vehicle by the predetermined application is executed, and the collected driving information according to a predetermined opportunity The may further execute a process of transmitting to the server device.

  In the insurance business support method of the present embodiment, when the mobile terminal transmits the driving information, the mobile terminal determines whether the predetermined observation data by the observation device matches a predetermined criterion corresponding to the occurrence of a predetermined event in the vehicle, As a result of the determination, when the occurrence of a predetermined event in the vehicle is detected, the collected driving information may be transmitted to the server device.

5 Network 10 Insurance Business Support System 20 Insurance Contract Vehicle 50 Beacon Terminal (Predetermined Device)
51 Storage Device 52 Program 53 Memory 54 Computing Device 55 Beacon ID
56 Near Field Communication Interface 57 Imaging Device 100 Data Center Server (Server Device)
101 storage device 102 program 103 memory 104 arithmetic device 105 communication device 125 contract information table 126 accident information table 127 sensor information table 128 operation information table 129 analysis result table 130 improvement difference table 200 portable terminal 201 storage device 202 program 2021 operation information collection application 2022 Beacon ID
2023 User ID
203 Memory 204 Computing Device 205 Input / Output Device 207 Network Interface 208 Short-range Wireless Communication Interface 210 Observation Device 211 GPS Unit 212 Acceleration Sensor 213 Gyro Sensor 214 Azimuth Sensor 215 Barometric Pressure Sensor 216 Temperature Sensor 300 Insurance Company Server

Claims (10)

A storage device for storing a contract information table storing information about each contract of car insurance, and an accident information table storing information of each accident caused by the contractor of the car insurance;
Driving information collected by the mobile terminal during reception of the predetermined signal from the predetermined device in the insurance contract vehicle is received from the mobile terminal, and based on the received driving information, the contract information table, and the accident information table, The relationship between the corresponding accident and the significance of the occurrence of an accident among the combinations of the attributes of each contractor who is a vehicle driver and each driving operation is modeled by a predetermined statistical process. A computing device that calculates insurance payment risk for each policyholder based on
An insurance business support system including a server device provided with The arithmetic unit of the server device is
Based on the calculated insurance payment risk of each policyholder, the driving information of each policyholder, and the contract information table, the total amount of insurance payment corresponding to the insurance policy payment risk of each policyholder in the car insurance is predetermined. A process for identifying a group of drivers having a risk of driving and a driving operation in which significance of an accident occurrence is defined in the model among driving operations of each contractor constituting the driver group,
A process of transmitting information on predetermined driving improvement guidance related to the specified driving operation to each contractor's terminal included in the driver group;
The insurance business support system according to claim 1, further comprising: The arithmetic unit of the server device is
In the case of modeling, when driving information from a plurality of vehicle drivers regarding the same vehicle is received from the portable terminal, driving related to the driver whose driving time indicated by the driving information of each vehicle driver is a predetermined time or more during a predetermined period. Information is specified, and based on the specified driving information, the contract information table, and the accident information table, the significance of the occurrence of an accident in each combination of the attribute of each contractor and each driving operation is not less than a predetermined value. Model the relationship between the accident and the accident with a predetermined statistical process.
The insurance business support system according to claim 1, wherein: A storage device storing a predetermined application for acquiring observation data from at least one of the vehicle position, behavior, and surrounding video;
When the predetermined signal including the identification ID of the predetermined device is received from the predetermined device, the predetermined application is started, and the predetermined ID that is included in the received predetermined signal, and the predetermined device that the predetermined application holds in advance A process for verifying whether the pair is a pair of a predetermined device and a portable terminal for which cooperation is defined in advance for the contractor,
When the authentication is successful, the predetermined app collects driving information including the observation data about the vehicle until the predetermined signal cannot be received from the predetermined device, and the collected driving information is used as a predetermined opportunity. In response, an arithmetic device that executes processing to be transmitted to the server device;
The insurance business support system according to claim 1, further comprising: a portable terminal comprising: The computing device of the mobile terminal is:
When transmitting the driving information, it is determined whether the predetermined observation data by the observation device matches a predetermined standard corresponding to the predetermined event occurrence in the vehicle, and if the predetermined event occurrence in the vehicle is detected as a result of the determination, The collected operation information is transmitted to the server device.
The insurance business support system according to claim 4. A server device comprising a storage device for storing a contract information table storing information related to each contract of car insurance and an accident information table storing information of each accident caused by the contractor of the car insurance,
Driving information collected by the mobile terminal during reception of the predetermined signal from the predetermined device in the insurance contract vehicle is received from the mobile terminal, and based on the received driving information, the contract information table, and the accident information table, The relationship between the corresponding accident and the significance of the occurrence of an accident among the combinations of the attributes of each contractor who is a vehicle driver and each driving operation is modeled by a predetermined statistical process. Based on each policyholder's insurance payment risk,
An insurance business support method characterized by that. The server device is
Based on the calculated insurance payment risk of each policyholder, the driving information of each policyholder, and the contract information table, the total amount of insurance payment corresponding to the insurance policy payment risk of each policyholder in the car insurance is predetermined. A process for identifying a group of drivers having a risk of driving and a driving operation in which significance of an accident occurrence is defined in the model among driving operations of each contractor constituting the driver group,
A process of transmitting information on predetermined driving improvement guidance related to the specified driving operation to each contractor's terminal included in the driver group;
The insurance business support method according to claim 6, further comprising: The server device is
In the case of modeling, when driving information from a plurality of vehicle drivers regarding the same vehicle is received from the portable terminal, driving related to the driver whose driving time indicated by the driving information of each vehicle driver is a predetermined time or more during a predetermined period. Information is specified, and based on the specified driving information, the contract information table, and the accident information table, the significance of the occurrence of an accident in each combination of the attribute of each contractor and each driving operation is not less than a predetermined value. Modeling the relationship between the accident and the accident with a predetermined statistical process,
The insurance business support method according to claim 6. A mobile terminal equipped with a storage device storing a predetermined application for acquiring observation data from at least one of the observation device of the position, behavior, and surrounding video of the vehicle,
When the predetermined signal including the identification ID of the predetermined device is received from the predetermined device, the predetermined application is started, and the predetermined ID that is included in the received predetermined signal, and the predetermined device that the predetermined application holds in advance A process for verifying whether the pair is a pair of a predetermined device and a portable terminal for which cooperation is defined in advance for the contractor,
When the authentication is successful, the predetermined app collects driving information including the observation data about the vehicle until the predetermined signal cannot be received from the predetermined device, and the collected driving information is used as a predetermined opportunity. In response, processing to send to the server device;
The insurance business support method according to claim 6, further comprising: The mobile terminal is
When transmitting the driving information, it is determined whether the predetermined observation data by the observation device matches a predetermined standard corresponding to the predetermined event occurrence in the vehicle, and if the predetermined event occurrence in the vehicle is detected as a result of the determination, Transmitting the collected driving information to the server device;
The insurance business support method according to claim 9.

Images