Information processing method, method for generating model, computer program, and information processing device

Abstract

To provide an information processing method and the like that can provide an estimate of vehicle inspection in a short time.SOLUTION: An information processing method according to one embodiment of the present disclosure acquires vehicle information related to a vehicle subject to vehicle inspection, inputs the acquired vehicle information into a learning model that has been trained to output necessity information for items that require part replacements or inspection when the vehicle information is input, and outputs the necessity information. In addition, the vehicle information includes maintenance information related to maintenance of the vehicle subject to vehicle inspection. The maintenance information includes the number of days since the last maintenance of the vehicle subject to vehicle inspection, mileage, the number of days since first registration, the number of inspections, or information regarding part replacements.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present technology relates to an information processing method, a model generating method, a computer program, and an information processing device. [Background technology]

[0002] Conventionally, vehicle maintenance or repairs are performed by vehicle inspection companies during vehicle inspections (vehicle inspections) conducted based on the Road Transport Vehicle Law. The content of the vehicle maintenance or repairs is estimated by the vehicle inspection company based on an inspection conducted before the vehicle inspection and proposed to the customer.

[0003] For example, when the system device described in Patent Document 1 receives the judgment results for each inspection item performed by an inspector, the system device displays the judgment results to a customer and calculates an estimated cost. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2005-190049 A Summary of the Invention [Problem to be solved by the invention]

[0005] In the system device described in Patent Document 1, an inspector must actually inspect the customer's vehicle, and it takes a lot of time to make an estimate for the vehicle inspection.

[0006] The present disclosure has been made in consideration of the above circumstances, and aims to provide an information processing method etc. that enables estimates to be made at the time of vehicle inspection in a short period of time. [Means for solving the problem]

[0007] An information processing method according to one embodiment of the present disclosure acquires vehicle information regarding a vehicle subject to inspection, inputs the acquired vehicle information into a learning model trained to output necessity information for items that require part replacement or inspection when the vehicle information is input, and outputs necessity information.

[0008] In an information processing method according to an embodiment of the present disclosure, the vehicle information includes maintenance information related to maintenance of the vehicle subject to vehicle inspection.

[0009] In an information processing method according to one embodiment of the present disclosure, the maintenance information includes information regarding the number of days since the last maintenance of the vehicle subject to vehicle inspection, mileage, the number of days since the first year of registration, the number of inspections, or part replacement.

[0010] In an information processing method according to an embodiment of the present disclosure, the necessity information includes whether or not replacement or inspection of a part is necessary, the likelihood of the necessity, or a classification of the likelihood of the necessity.

[0011] An information processing method according to an embodiment of the present disclosure lists and outputs the items, the necessity information, and major factors of the vehicle information that had the greatest influence on output of the necessity information.

[0012] An information processing method according to one embodiment of the present disclosure outputs a command to create a first proposal for a vehicle inspection company, in which the necessity information is described, and a command to create a second proposal for a customer, which is different from the first proposal.

[0013] An information processing method according to one embodiment of the present disclosure outputs, in the second proposal, the items, the necessity information, the major factors of the vehicle information that had the greatest impact on the output of the necessity information, and a message corresponding to the items and the major factors.

[0014] An information processing method according to an embodiment of the present disclosure outputs the items for which the likelihood of the need for part replacement or inspection is equal to or greater than a predetermined value.

[0015] An information processing method according to one embodiment of the present disclosure refers to a table that stores required items that must be included in a proposal regardless of the necessity information, and includes the required items in the proposal regardless of the output of the necessity information.

[0016] An information processing method according to one embodiment of the present disclosure refers to a table that stores free items for which no fee is charged for replacement or inspection of the part, and among the items for which necessity information has been output, those items that correspond to the free items are written in a proposal.

[0017] A model generation method according to one embodiment of the present disclosure acquires training data including vehicle information on a vehicle subject to inspection and necessity information for items that require replacement or inspection of parts of the vehicle subject to inspection, and generates a learning model that outputs the necessity information when the vehicle information is input based on the acquired training data.

[0018] In one embodiment of the present disclosure, a model generation method is provided in which original data for the training data is collected from store devices installed in the stores of multiple vehicle inspection companies, the training data is generated by converting the store management numbers included in the original data into common management numbers by referring to a table that stores correspondence between store management numbers assigned to each item, which differ for each store device, and common management numbers that are common to all stores, and the vehicle information and necessity information included in the training data are associated with each other by the common management number.

[0019] A computer program according to one embodiment of the present disclosure causes a computer to execute a process of acquiring vehicle information regarding a vehicle subject to inspection, inputting the acquired vehicle information into a learning model that has been trained to output necessity information for items that require part replacement or inspection when the vehicle information is input, and outputting necessity information.

[0020] An information processing device according to one embodiment of the present disclosure includes a processing unit that acquires vehicle information regarding a vehicle subject to inspection, inputs the acquired vehicle information into a learning model that has been trained to output necessity information for items that require part replacement or inspection when the vehicle information is input, and outputs necessity information. Effect of the Invention

[0021] In the information processing method according to an embodiment of the present disclosure, it is possible to perform an estimate at the time of vehicle inspection in a short period of time. [Brief description of the drawings]

[0022] [Figure 1] FIG. 1 is an explanatory diagram showing a configuration example of a vehicle inspection proposal system. [Diagram 2] 1 is a block diagram showing an example of the configuration of an information processing device; [Diagram 3] FIG. 2 is a block diagram showing a configuration example of a store device. [Figure 4] FIG. 2 is an explanatory diagram illustrating an example of a store vehicle information table. [Diagram 5] FIG. 4 is an explanatory diagram illustrating an example of a vehicle information table. [Figure 6] FIG. 11 is an explanatory diagram illustrating an example of an item table. [Figure 7] FIG. 1 is an explanatory diagram of a learning model. [Figure 8] FIG. 11 is an explanatory diagram of a learning model according to a first modified example of the first embodiment. [Figure 9] FIG. 13 is an explanatory diagram of a learning model according to a second modified example of the first embodiment. [Figure 10] FIG. 13 is an explanatory diagram showing an example of an output result display screen. [Figure 11] FIG. 11 is an explanatory diagram showing an example of a first proposal. [Figure 12] FIG. 11 is an explanatory diagram showing an example of a second proposal. [Figure 13] FIG. 11 is an explanatory diagram illustrating an example of a message table. [Figure 14] 11 is a flowchart illustrating an example of processing by an information processing device. [Figure 15] 10 is a flowchart illustrating an example of a process of a store device. [Figure 16] FIG. 11 is a block diagram showing an example of the configuration of a store device according to a second embodiment. [Figure 17] FIG. 13 is an explanatory diagram illustrating an example of a required item table. [Figure 18] FIG. 11 is an explanatory diagram showing an example of a store vehicle information table according to the second embodiment. [Figure 19] FIG. 11 is an explanatory diagram illustrating an example of a maintenance pack table. [Figure 20] FIG. 11 is an explanatory diagram showing an example of a second proposal according to the second embodiment. [Figure 21] 10 is a flowchart showing an example of a process performed by a store device according to the second embodiment. [Figure 22] FIG. 11 is a block diagram showing an example of the configuration of an information processing device according to a third embodiment. [Figure 23] FIG. 11 is a block diagram showing a configuration example of a store device according to a third embodiment. [Figure 24] FIG. 11 is an explanatory diagram showing an example of a vehicle inspection history table according to the third embodiment. [Diagram 25] FIG. 11 is an explanatory diagram illustrating an example of an item master table. [Figure 26] 13 is a flowchart illustrating an example of a learning model generation processing procedure. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0023] (Embodiment 1) Hereinafter, the present invention according to the first embodiment will be described with reference to the drawings. FIG. 1 is an explanatory diagram showing an example of the configuration of a vehicle inspection proposal system, FIG. 2 is a block diagram showing an example of the configuration of an information processing device, and FIG. 3 is a block diagram showing an example of the configuration of a store device. The vehicle inspection proposal system S includes an information processing device 1 and a plurality of store devices 2. The store devices 2 are devices provided in the stores of vehicle inspection companies. The information processing device 1 and the store devices 2 communicate with each other via a network N. Each store device 2 accepts input of vehicle information related to vehicles to be inspected owned by customers at each store and stores it. The information processing device 1 acquires vehicle information from the store devices 2, and based on the acquired vehicle information, outputs necessity information of items that require replacement or inspection (maintenance) in the vehicle to be inspected. The store device 2 acquires the necessity information output by the information processing device 1, creates a proposal for vehicle inspection (vehicle inspection proposal) containing the necessity information, and presents it to the vehicle inspection company and the customer.

[0024] The information processing device 1 is, for example, a server computer, and includes a processing unit 11, a storage unit 12, and a communication unit 13. The processing unit 11 is configured with a CPU (Central Processing Unit), an MPU (Micro Processing Unit), a GPU (Graphical Processing Unit), a quantum processor, or the like, and performs various control processing, arithmetic processing, and the like by reading and executing a program P (program product) and a database stored in advance in the storage unit 12. Note that a database server or the like may be provided outside the information processing device 1, and the database may be read from the database server or the like. Furthermore, the information processing device 1 may be one whose functions are realized by a plurality of server devices or computers. Furthermore, the information processing device 1 may correspond to a node on a blockchain.

[0025] The storage unit 12 of the information processing device 1 is, for example, a volatile memory and a non-volatile memory. The storage unit 12 stores a program P, a plurality of learning models M, a vehicle information table 121, and an item table 122. The program P may be provided to the information processing device 1 using a storage medium 12a in which the program P is stored so as to be readable by a computer. The storage medium 12a is, for example, a portable memory. Examples of the portable memory include a CD-ROM, a USB (Universal Serial Bus) memory, an SD card, a micro SD card, and a compact flash memory (registered trademark). When the storage medium 12a is a portable memory, the processing element of the processing unit 11 may read the program P from the storage medium 12a using a reading device (not shown). The read program P is written to the storage unit 12. Furthermore, the program P may be provided to the information processing device 1 by the communication unit 13 communicating with an external device. Details of the learning model M, the vehicle information table 121, and the item table 122 will be described later.

[0026] The communication unit 13 of the information processing device 1 is a communication module or communication interface for communicating with the store device 2 by wire or wirelessly, for example, a wide-area wireless communication module. The processing unit 11 communicates with the store device 2 via the communication unit 13, for example, through an external network N such as the Internet. The processing unit 11 acquires information (vehicle information) related to the customer's vehicle from the store device 2 through the communication unit 13, and transmits an output result by the learning model M to the store device 2. The learning model M may be executed by another computer (not shown). In addition, in the present embodiment, an example in which the learning model M is executed in the information processing device 1 is shown, but this is not limited thereto. For example, the learned learning model M may be deployed to the store device 2, and the store device 2 may output necessity information of items that require replacement or inspection (maintenance).

[0027] The in-store device 2 is, for example, a server computer, a personal computer, a smartphone, or a tablet terminal installed in a store of a vehicle inspection dealer. The in-store device 2 includes a processing unit 21, a storage unit 22, a communication unit 23, and an input / output I / F 24. The processing unit 21 and the communication unit 13 have the same configuration as the processing unit 11 and the communication unit 13 of the information processing device 1.

[0028] The storage unit 22 of the store device 2 stores a store vehicle information table 221 and a message table 222. The store vehicle information table 221 stores vehicle information on vehicles owned by customers in the store in which the store device 2 is installed. The message table 222 stores messages to be described in a second proposal 42, which will be described later. The details of the store vehicle information table 221 and the message table 222 will be described later.

[0029] The input / output I / F 24 is compliant with a communication standard such as USB or DSUB, and is a communication interface for serial communication with an external device connected to the input / output I / F 24. A display unit 241 such as a display, and an input unit 242 such as a mouse and a keyboard are connected to the input / output I / F 24, and the processing unit 21 outputs to the display unit 241 a result of information processing performed based on an execution command or an event input from the input unit 242. Also connected to the input / output I / F 24 is a printer 243 that prints a proposal, which will be described later.

[0030] FIG. 4 is an explanatory diagram showing an example of a store vehicle information table. In the store vehicle information table 221, vehicle information on vehicles subject to vehicle inspection of customers of each store is registered. The store vehicle information table 221 includes, for example, a customer name field, a customer number field, a registration number field, a vehicle type field, a number of days since initial registration field, a total mileage field, a number of vehicle inspections field, a number of inspections field, a number of days since inspection field, a mileage after inspection field, a maintenance information field, and a maintenance date field. The customer name field stores the name of the customer who is the owner of the vehicle subject to vehicle inspection. The customer number field stores a number assigned to the customer. The registration number field stores the registration number written on the license plate of the vehicle registered by the local government. The vehicle type field stores the vehicle type of the vehicle subject to vehicle inspection. The number of days since initial registration field stores the number of days that have passed since the initial registration of the vehicle subject to vehicle inspection to the present. The total mileage field stores the distance that the vehicle subject to vehicle inspection has traveled since its initial registration to the present. The number of vehicle inspections field stores the number of inspections that have been performed on the vehicle subject to inspection from the initial registration to the present. The number of inspections field stores the number of inspections that have been performed on the vehicle subject to inspection from the initial registration to the present. Inspections are not inspections based on the Road Transport Vehicle Act, but are inspections performed at the request of the customer. Note that inspections may also be performed at the same time as vehicle inspections. The mileage after inspection field stores the distance traveled by the vehicle subject to inspection from the last inspection to the present. The maintenance information field stores parts that have been previously maintained (parts replaced or inspected) on the vehicle subject to inspection. The maintenance date field stores the date on which the maintenance stored in the maintenance information field was performed. Note that the store vehicle information table 221 may contain multiple maintenance information fields and maintenance date fields.

[0031] For example, when the processing unit 21 of the store device 2 receives an input of a proposal creation instruction, the processing unit 21 inserts a new record into the store vehicle information table 221. At this time, the processing unit 21 creates a new record to be inserted into the store vehicle information table 221 by referring to a record already stored in the store vehicle information table 221 for the corresponding vehicle to be inspected. For example, in the number of days since initial registration field and the number of days since inspection field, if a record with the same registration number exists in the past, the number of days stored in the corresponding record, the date on which the corresponding record was created, and the number of days calculated based on the current date are stored. In addition, the number of times stored in the number of inspections field is the number of times incremented by 1 from the record with the same registration number created last time. The number of times stored in the number of inspections field is the number of inspections after the initial registration of the vehicle to be inspected, which is acquired by referring to an inspection history table (not shown). In addition, the total mileage field and the mileage after inspection field may store a newly input distance based on the display of the meter of the vehicle to be inspected.

[0032] FIG. 5 is an explanatory diagram showing an example of a vehicle information table. The vehicle information table 121 stores information about vehicles that are subject to vehicle inspection. The vehicle information table 121 includes, for example, a store field, a customer name field, a customer number field, a registration number field, a vehicle type field, a number of days since initial registration field, a total mileage field, a number of vehicle inspections field, a number of inspections field, a number of days since inspection field, a mileage since inspection field, and a replacement part field. The store field stores the name of the store that performs the vehicle inspection. The other fields are the same as those included in the store vehicle information table 221, and store vehicle information acquired by the processing unit 11 of the information processing device 1 from the storage unit 22 of the store device 2.

[0033] FIG. 6 is an explanatory diagram showing an example of an item table. The item table 122 stores items for which necessity information is output. The item table 122 includes, for example, an item number field, a type field, and an item field. The item number field stores a number assigned to each item. The type field stores the type to which each item belongs. The item field stores items that may require maintenance at the time of vehicle inspection. In this embodiment, the items are classified into three types: regularly replaced parts, functional parts, and value-added products.

[0034] 7 is an explanatory diagram of the learning model M. The learning model M is a machine learning model that uses vehicle information of a vehicle to be inspected as input data and outputs necessity information for each item stored in an item table. In this embodiment, the necessity information is output as the likelihood that each item is estimated to require maintenance at the time of inspection.

[0035] A plurality of learning models M are stored for each item. That is, a learning model M that outputs necessity information for each item is stored, such as a first learning model M that uses vehicle information as input data and outputs information on the necessity of battery replacement, a second learning model M that outputs information on the necessity of tire replacement, and a third learning model M that outputs information on the necessity of engine oil replacement. Each learning model M has the same configuration.

[0036] The learning model M is constructed, for example, by LGBoost (Light Gradient Boosting). LGBoost is an ensemble learning method that combines gradient boosting and random forests. The learning model M is configured to construct multiple decision trees (weak learners) and perform boosting to construct a new decision tree using information from the previous decision tree. Specifically, a new decision tree is constructed with the error (gradient of the loss function) that could not be predicted by the previous decision tree as the objective variable. In each decision tree, the input data is classified according to conditions on the way from the root to the branches, and when it reaches the terminal leaf node, the value given to that terminal leaf node is output as the predicted value.

[0037] The input data of the learning model M is vehicle information. The vehicle information is information stored in each field of the vehicle information table 121. In this embodiment, the information stored in the vehicle type field, the number of days since initial registration field, the total mileage field, the number of vehicle inspections field, the number of inspections field, the number of days since inspection field, the mileage after inspection field, the maintenance information field, and the maintenance date field among the fields of the vehicle information table 121 is used as the input data. The vehicle information that is the input data of the learning model M may include customer information including the age, sex, residential area, license category, or information on the insurance already purchased of the customer who is the owner of the vehicle to be inspected. In addition, only vehicle usage information including the number of days since initial registration, the total mileage, the number of days since inspection, and the mileage after inspection may be used as the input data of the learning model M. The input data of the learning model M is not limited to this, and only a part of the information included in the vehicle information may be used as the input data.

[0038] When vehicle information is input, the learning model M outputs the likelihood that maintenance will be required at the time of vehicle inspection for each item. The likelihood value ranges from 0 to 1.

[0039] The output of the learning model M may be a classification result of the necessity (necessary or unnecessary) of each item. FIG. 8 is an explanatory diagram of the learning model M according to a first modified example of the first embodiment. In this example, when vehicle information is input, the learning model M outputs the accuracy for each class of necessary or unnecessary. The learning model M can output the class whose accuracy is equal to or greater than a threshold value as the output value.

[0040] The learning model M may be a multi-class classification model that classifies the level of likelihood that maintenance is required at the time of vehicle inspection for each item. FIG. 9 is an explanatory diagram of the learning model M according to the second modified example of the first embodiment. In this example, the learning model M outputs a level according to the likelihood, such as a first level when the likelihood is 0 or more and less than 0.25, a second level when the likelihood is 0.25 or more and less than 0.5, a third level when the likelihood is 0.5 or more and less than 0.75, and a fourth level when the likelihood is 0.75 or more and less than 1. The learning model M in the second modified example includes a feature extraction layer m1 and a plurality of output layers m2. Each output layer m2 outputs the level of likelihood for each item.

[0041] In the above, an example in which the learning model M is LGBoost has been described, but the configuration of the learning model M is not limited, and it is sufficient if it can output the likelihood that maintenance is required for each item at the time of vehicle inspection. The learning model M may be, for example, a neural network such as Transformer, CNN (Convolution Neural Network), RNN (Recurrent Neural Network), or LSTM (Long Short Term Memory), or may use other learning algorithms such as a support vector machine, logistics regression, random forest, or XGBoost (eXtreme Gradient Boosting). In addition, the input data to the learning model M may be time-series data including vehicle information previously registered in the vehicle information table 121.

[0042] The processing unit 11 calculates the contribution of each piece of information included in the vehicle information input to the learning model M, thereby identifying the information (main factor) that contributes most to the output likelihood. The method of calculating the contribution is not limited, but for example, SHAP (SHapley Additive exPlanation), LIME (Local Interpretable Model-Agnostic Explanations), Attention mechanism, etc. can be used. The greater the contribution of the information, the greater the influence on the output of the likelihood.

[0043] 10 is an explanatory diagram showing an example of an output result display screen. The processing unit 11 of the information processing device 1 transmits the necessity information of each item and the identified main factors output by the learning model M to the store device 2. The store device 2 displays the necessity information of each item and the identified main factors acquired from the information processing device 1 on the display unit 241. The output result display screen 3 includes a vehicle information field 31, a previous vehicle inspection and maintenance field 32, an upcoming maintenance field 33, a recommended vehicle inspection and maintenance field 34, and a command field 35.

[0044] The vehicle information field 31 stores vehicle information of the vehicle subject to inspection that is input data for the learning model M. The previous inspection maintenance field 32 displays the details, date, and fee of the maintenance performed at the previous inspection of the vehicle subject to inspection. The details entered in the previous inspection maintenance field 32 are read from a vehicle inspection history table (not shown) that stores the history of vehicle inspections of the vehicle and is stored in the memory unit 22 of the store device 2. The most recent maintenance field 33 displays the maintenance information closest to the current date among the maintenance information stored in the maintenance information field of the store vehicle information table 221.

[0045] The recommended vehicle inspection maintenance field 34 displays the items for which the necessity information has been output by the information processing device 1, the item numbers of the items, the main causes identified by the information processing device 1, and the maintenance recommendation level. In the maintenance recommendation level, for example, the maintenance recommendation level based on the likelihood output by the learning model M is indicated by the number of stars. In this example, when the likelihood is 0 or more and less than 0.25, 0 stars are displayed, when the likelihood is 0.25 or more and less than 0.5, 1 star is displayed, when the likelihood is 0.5 or more and less than 0.75, 2 stars are displayed, and when the likelihood is 0.75 or more and 1 or less, 3 stars are displayed. Each item is displayed, for example, in descending order of recommendation level for each type of item.

[0046] The command field 35 includes a first proposal creation command 351 and a second proposal creation command 352. Details of the first proposal 41 and the second proposal 42 will be described later, but the first proposal 41 is a proposal to be presented to the vehicle inspection company, and the second proposal 42 is a proposal to be presented to the customer. When the first proposal creation command 351 is selected, the processing unit 21 of the in-store device 2 creates the first proposal 41 and causes the printing device 243 to print it. The same applies when the second proposal creation command 352 is selected. Note that the first proposal 41 and the second proposal 42 may be displayed on the display unit 241 of the in-store device 2 or on a smartphone or the like owned by the customer.

[0047] FIG. 11 is an explanatory diagram showing an example of the first proposal. The first proposal 41 includes a vehicle information field 411, a previous vehicle inspection and maintenance field 412, a most recent maintenance field 413, and a recommended vehicle inspection and maintenance field 414. That is, the first proposal 41 includes the same contents as those displayed on the output result display screen 3 except for the command field 35. The recommended vehicle inspection and maintenance field 414 may include the previous maintenance date of each item. In this case, for the previous maintenance date of an item, if there is a corresponding item in the maintenance information field of the store vehicle information table 221, the maintenance date of the corresponding item is included. Also, items with a low degree of recommendation may not be included in the recommended vehicle inspection and maintenance field 414. For example, an item with zero stars displayed on the output result display screen 3 may not be included in the first proposal 41. The first proposal 41 is printed as a proposal for the vehicle inspection company.

[0048] FIG. 12 is an explanatory diagram showing an example of the second proposal, and FIG. 13 is an explanatory diagram showing an example of a message table. The second proposal 42 includes a vehicle information field 421, a recommended vehicle inspection and maintenance field 424, and a message field 425. The contents described in the vehicle information field 421 and the recommended vehicle inspection and maintenance field 424 are the same as the contents described in the vehicle information field 411 and the recommended vehicle inspection and maintenance field 414 of the first proposal. Note that, like the recommended vehicle inspection and maintenance field 414 of the first proposal 41, the recommended vehicle inspection and maintenance field 424 of the second proposal 42 may describe the last maintenance date of each item. Also, in the recommended vehicle inspection and maintenance field 424, items with a low degree of recommendation may not be described. For example, an item with zero stars displayed on the output result display screen 3 may not be described in the second proposal 42. The second proposal 42 is printed as a proposal for the customer.

[0049] In the message field 425, messages for encouraging the customer to replace or inspect parts are written for a plurality of items in descending order of the maintenance recommendation level. The messages written in the message field 425 are selected with reference to the message table 222 based on the items for which the necessity information has been output and the identified major factors. As shown in FIG. 13, the message table 222 includes, for example, an item field, a major factor field, and a message field. The item field stores the maintenance items. The major factor field stores vehicle information that may be a major factor. The message field stores the messages written in the message field 425 of the second proposal for the items and major factors.

[0050] When the second proposal creation command 352 is selected, the processing unit 21 of the in-store device 2 determines the item for which a message is to be written in the second proposal 42. In the example shown in FIG. 12, it is determined that messages for three items with high maintenance recommendation levels (wiper replacement, air element replacement, and V-belt replacement) are to be written. Next, the processing unit 21 of the in-store device 2 refers to the message table and specifies the item for which a message is to be written and a message based on the main factor. The processing unit 21 creates the second proposal 42 in which the determined item and the specified message are written in the message field 425, and causes the printing device 243 to print it. In the example shown in FIG. 12, since the main factor for the high maintenance recommendation level (likelihood level) for the item "wiper replacement" is specified as "number of days since inspection", the processing unit 21 refers to the message table 222 and determines the message to be written in the second proposal 42 as "Many days have passed since inspection. If wiping is uneven, eye fatigue will also increase. Would you like to replace it this time?". This makes it possible to increase the motivation of customers to carry out maintenance. The message written in the second proposal 42 may be specified based on the season in which the vehicle inspection is performed or on customer information. The processing unit 21 may also refer to a fee table (not shown) that stores the amount required for maintenance for each item, and write in the second proposal 42 an estimated amount calculated by accumulating the amount required for maintenance of items whose maintenance recommendation level is equal to or higher than a certain level.

[0051] FIG. 14 is a flowchart for explaining an example of the processing of the information processing device 1. For example, when the vehicle information of the vehicle to be inspected is transmitted from the store device 2, the processing unit 11 of the information processing device 1 starts the following processing. The processing unit 11 receives the vehicle information of the vehicle to be inspected from the store device 2 (S1). The processing unit 11 stores the vehicle information received from the store device 2 in the vehicle information table 121 (S2). The processing unit 11 inputs the vehicle information of the vehicle to be inspected into the learning model M (S3), and outputs necessity information of items requiring maintenance for each item (S4). The processing unit 11 calculates the contribution degree of each information included in the vehicle information input in S4 to the output of the necessity information (S5), and identifies the main factors of the vehicle information based on the calculated contribution degree (S6). The processing unit 11 transmits the necessity information output in S4 and the main factors identified in S6 to the store device 2 (S7), and ends the processing.

[0052] FIG. 15 is a flowchart for explaining an example of the processing of the store device. The processing unit 21 of the store device 2 reads out the latest record from the store vehicle information table 221 of past records related to the vehicle to be inspected based on the registration number (S11). The processing unit 21 updates the vehicle information stored in the read record (S12) and inserts a new record into the store vehicle information table 221 (S13). The processing unit 21 transmits the vehicle information to be stored in the new record to the information processing device 1 (S14). The processing unit 21 receives necessity information and main factors of each item from the information processing device 1 (S15). The processing unit 21 displays the necessity information and main factors received from the information processing device 1 on the output result display screen 3 for each item (S16). Note that in S16, items with a low degree of recommendation may not be displayed on the output result display screen 3. The processing unit 21 reads out vehicle information and maintenance information from the store vehicle information table 221 (S17). The processing unit 21 also reads out the vehicle inspection history from the vehicle inspection history table (S18). The processing unit 21 displays the vehicle information, the vehicle inspection history, and the maintenance information on the output result display screen 3 (S19). The processing unit 21 also displays the first proposal creation command 351 and the second proposal creation command 352 in the first proposal 41 (S20). The processing unit 21 accepts the selection of the first proposal creation command 351 (S21) and creates the first proposal 41 (S22). Specifically, the processing unit 21 creates the first proposal 41 by filling in the vehicle information column 411, the previous vehicle inspection maintenance column 412, the latest maintenance column 413, and the recommended vehicle inspection maintenance column 414. The processing unit 21 causes the printer 243 to print (S23). The processing unit 21 accepts the selection of the second proposal creation command 352 (S24) and determines the items in which a message is to be written in the second proposal 42 based on the maintenance recommendation level of each item (S25). The processing unit 21 reads the message table 222 (S26) and determines a message to be written in the second proposal 42 based on the items determined in S25 and the main factors for the items (S27). The processing unit 21 creates the second proposal 42 (S28). Specifically, the processing unit 21 creates the second proposal 42 by filling in the vehicle information column 421, the recommended vehicle inspection maintenance column 424, and the message column 425. The processing unit 21 causes the printing device 243 to print the second proposal 42 (S29), and the processing ends.

[0053] According to the above process, a proposal for items to be maintained can be created without inspecting the vehicle to be inspected by an inspector, so that estimates can be made at the time of vehicle inspection in a short time. In addition, the contents of the maintenance items proposed by the vehicle inspection company can be standardized, and the omission and variation of the proposal can be reduced. In this embodiment, the information processing device 1 outputs the necessity information using the learning model M, and the store device 2 creates the proposal, but this is not limited to this. The information processing device 1 and the store device 2 may be configured as an integrated device, and the functions of the information processing device 1 and the store device 2 may be realized by one device.

[0054] (Embodiment 2) Hereinafter, the present invention according to the second embodiment will be described with reference to the drawings. Among the configurations according to the second embodiment, the same configurations as those according to the first embodiment are given the same reference numerals, and detailed description thereof will be omitted. The store device 2 according to the second embodiment describes the required items in the second proposal 42 regardless of the necessity information. Furthermore, the store device 2 according to the second embodiment describes in the second proposal 42 the items corresponding to the free items for which no fee is required for part replacement or inspection (maintenance) among the items for which the necessity information is output. FIG. 16 is a block diagram showing an example of the configuration of the store device according to the second embodiment. The storage unit 22 of the store device 2 according to the second embodiment stores a required item table 223 and a maintenance package table 224.

[0055] 17 is an explanatory diagram showing an example of a required item table. Required item table 223 includes, for example, a vehicle type field and a required item field. The vehicle type field stores the vehicle type name. The required item field stores items that must be written in second proposal 42 for the vehicle type of the vehicle to be inspected. Note that multiple items may be stored in the required item field, and required item table 223 may include multiple required item fields.

[0056] FIG. 18 is an explanatory diagram showing an example of a store vehicle information table according to the second embodiment. The store vehicle information table 221 according to the second embodiment includes a maintenance pack field. A maintenance pack is a package product that a customer subscribes to when purchasing a vehicle, etc., and allows subsequent replacement or inspection of parts to be free of charge. There are multiple types of maintenance packs, and the maintenance pack field of the store vehicle information table 221 stores the types of maintenance packs.

[0057] 19 is an explanatory diagram showing an example of a maintenance pack table. The maintenance pack table 224 includes a maintenance pack field and a plurality of free item fields. The maintenance pack field stores the type of maintenance pack. The free item field stores items that are free items for which no fee is required for part replacement or inspection (maintenance) for the type of maintenance pack.

[0058] FIG. 20 is an explanatory diagram showing an example of the second proposal according to the second embodiment. The second proposal 42 according to the second embodiment includes a required item field 426 and a maintenance pack field 427. The required item field 426 describes required items for the vehicle type of the vehicle to be inspected, which are identified by referring to the required item table. The required item field 426 also describes that the required items are recommended for the vehicle type of the vehicle to be inspected. The maintenance pack field 427 describes the type of maintenance pack to which the customer who is the owner of the vehicle to be inspected subscribes, which are identified by referring to the store vehicle information table 221, and free items. The processing unit 21 may highlight items that correspond to free items among the items described in the recommended vehicle inspection maintenance field 424. The processing unit 21 may also describe in the second proposal 42 an estimated amount obtained by accumulating the amount required for maintenance of only items that do not correspond to free items among items whose maintenance recommendation degree is equal to or higher than a certain degree. Furthermore, the processing unit 21 may describe in the second proposal 42 an estimated amount obtained by adding the amount required for maintenance of the required items.

[0059] FIG. 21 is a flowchart showing an example of the process of the store device according to the second embodiment. When the second proposal creation command 352 is selected on the output result display screen 3, the processing unit 21 of the store device 2 according to the second embodiment starts the following process. The processing unit 21 reads out the store vehicle information table 221 (S31) and acquires the vehicle type of the vehicle to be inspected (S32). The processing unit 21 reads out the required item table 223 (S33) and identifies the required items for the vehicle type acquired in S32 (S34). The processing unit 21 reads out the store vehicle information table 221 (S35) and acquires the type of maintenance package for the vehicle to be inspected (S36). The processing unit 21 reads out the maintenance package table 224 (S37) and identifies the free items for the type of maintenance package read out in S32 (S38). The processing unit 21 writes the required items identified in S34 and the free items identified in S37 in the second proposal 42 (S39). The processing unit 21 causes the printing device 243 to print the second proposal 42 (S40), and ends the process. Note that the required items or free items may also be described in the first proposal 41.

[0060] According to the above process, recommended items and free items for which no maintenance fee is required are written in the second proposal 42, so that it is possible to improve the customer's motivation to carry out a vehicle inspection. Note that the processing unit 21 of the store device 2 may write in the second proposal 42 only one of the information on the required items and the information on the maintenance pack.

[0061] (Embodiment 3) Hereinafter, the present invention according to the third embodiment will be described with reference to the drawings. Among the configurations according to the third embodiment, the same configurations as those in the first embodiment are given the same reference numerals, and detailed description thereof will be omitted. In the third embodiment, the generation of the learning model M will be described.

[0062] 22 is a block diagram showing an example of the configuration of an information processing device according to embodiment 3. The storage unit 12 of the information processing device 1 according to embodiment 3 stores an item master table 123. The item master table 123 will be described in detail later.

[0063] 23 is a block diagram showing an example of the configuration of a store device according to embodiment 3. The storage unit 22 of the store device 2 according to embodiment 3 includes a vehicle inspection history table 225. The vehicle inspection history table 225 will be described later in detail.

[0064] FIG. 24 is an explanatory diagram showing an example of the vehicle inspection history table according to the third embodiment. The vehicle inspection history table 225 according to the third embodiment includes, for example, a customer name field, a customer number field, a registration number field, a vehicle inspection date field, and a plurality of performed maintenance code / performed item fields. The customer name field stores the customer name of the customer who is the owner of the vehicle to be inspected. The customer number field stores the number assigned to the customer. The registration number field stores the registration number written on the license plate of the vehicle registered by the local government. The vehicle inspection date field stores the date on which the vehicle inspection was performed. The performed maintenance code / performed item field stores a unique code assigned to the maintenance performed at the time of the vehicle inspection at each store in the upper row, and stores the name of the maintenance performed at the time of the vehicle inspection in the lower row. The vehicle inspection history table 225 stores data on the items of maintenance performed by the vehicle inspection company.

[0065] FIG. 25 is an explanatory diagram showing an example of an item master table. The item master table 123 includes a store field, a store management number field, a store maintenance name field, a common management number field, and a common management item field. The store field stores the store name. The store management number field stores a code for maintenance uniquely assigned to each store. The store maintenance name field stores the name of the maintenance corresponding to the store management number. The common management number field stores the item number (common management number) of the item to which the maintenance at the store belongs. The common management item field stores the item name of the item to which the maintenance at the store belongs.

[0066] FIG. 26 is a flowchart showing an example of a generation process procedure of the learning model M. The processing unit 11 of the information processing device 1 reads out vehicle information from the store vehicle information table 221 stored in the storage unit 22 of the store device 2 (S51). The processing unit 11 reads out the vehicle inspection history related to the vehicle whose vehicle information has been read out from the vehicle inspection history table 225 stored in the storage unit 22 of the store device 2 (S52). The vehicle inspection history read out from the vehicle inspection history table 225 becomes the original data of the training data. The processing unit 11 of the information processing device 1 refers to the item master table and converts the store management number and the store maintenance name included in the read out vehicle inspection history into an item number and an item name (S53). The processing unit 11 associates the vehicle information on the vehicle inspection date included in the vehicle inspection history read out in S52 with the converted vehicle inspection history to create training data (S54). In addition, in S54, the processing unit 11 associates a large amount of vehicle information read out from the store device 2 of each store with the vehicle inspection history to create a large amount of training data. The processing unit 11 creates a learning model M based on the generated training data (S55).

[0067] Specifically, the processing unit 11 takes vehicle information as input data and optimizes each learning model M so that 1 is output for the likelihood of an item included in the vehicle inspection history and 0 is output for the likelihood of an item not included in the vehicle inspection history. When the learning model M is constructed by LGBoost, the processing unit 11 adjusts parameters using, for example, gradient descent so as to optimize (minimize) the loss function in the learning model M, and sequentially learns the outcome. Note that, in order to prevent so-called overfitting (overlearning), when optimizing the loss function, the control unit 10 does not use all explanatory variables, but selects and uses the number of explanatory variables at a randomly determined ratio. The processing unit 11 completes the learning when the loss function satisfies a predetermined criterion.

[0068] The processing unit 11 of the information processing device stores the learned learning model M in the storage unit 12 (S56), and ends the process. The processing unit 11 executes the above-mentioned process for each learning model M.

[0069] According to the above process, when vehicle information is input, it is possible to generate a learning model M that has been trained to be able to appropriately output necessity information (likelihood) of items that require part replacement or inspection (maintenance). Even if the items to be maintained are classified uniquely for each store, they can be included in the training data by converting them into common management numbers and common management items, that is, by performing so-called biasing. This makes it possible to generate the learning model M using vehicle inspection histories from multiple stores, and it is possible to increase the number of training data and generate a more accurate learning model M. Note that, after generating the learning model M, the processing unit 11 may re-train the learning model M by a method similar to the above-mentioned process. Note that, in this embodiment, the storage unit 12 of the information processing device 1 stores the item master table 123, and the conversion from the store management number to the common management number is performed by the processing unit 11 of the information processing device 1, but the storage unit 22 of the store device 2 may store the item master table, and the conversion from the store management number to the common management number may be performed by the processing unit 21 of the store device 2.

[0070] (Modification) In the above-described embodiments, the vehicle inspection is described as a statutory vehicle inspection (a vehicle inspection conducted in accordance with the Road Transport Vehicle Law), but is not limited thereto. The configurations and processes described in the embodiments may be similarly applied to any vehicle inspection or vehicle maintenance conducted by a private company.

[0071] The embodiments disclosed herein should be considered to be illustrative in all respects and not restrictive. The technical features described in each embodiment can be combined with each other, and the scope of the present invention is intended to include all modifications within the scope of the claims and equivalents to the scope of the claims. In addition, the independent claims and dependent claims described in the claims can be combined with each other in all combinations regardless of the reference format. Furthermore, the claims use a format in which a claim references two or more other claims (multiple claim format), but this is not limited to this. A format in which a multiple claim (multi-multi claim) that references at least one multiple claim may be used. [Explanation of symbols]

[0072] 1. Information processing device 10 Control section 11 Processing section 12 Storage section 121 Vehicle Information Table 122 Item Table 123 Item Master Table 12a Storage medium 13. Communications Department 2. Store equipment 21 Processing section 22 Memory section 221 Store vehicle information table 222 Message Table 223 Required Fields Table 224 Maintenance Pack Table 225 Vehicle Inspection History Table 23 Communications Department 24 Input / Output Interface 241 Display section 242 Input section 243 Printing device 3 Output result display screen 31 Vehicle information section 32 Last vehicle inspection and maintenance column 33 Latest maintenance column 34 Recommended vehicle inspection and maintenance column 35 Command column 351 First Proposal Creation Command 352 Second Proposal Creation Command 41 First proposal 411 Vehicle information section 412 Last vehicle inspection and maintenance column 413 Latest maintenance column 414 Recommended vehicle inspection and maintenance column 42 Second proposal 421 Vehicle information section 424 Recommended vehicle inspection and maintenance column 425 Message field 426 Required field 427 Maintenance Pack Column M Learning Model N Network P Program S Vehicle Inspection Proposal System

Claims

1. Obtain vehicle information on vehicles subject to inspection, The acquired vehicle information is input to a learning model that has been trained to output information on whether or not a part needs to be replaced or inspected when the vehicle information is input, and the information on whether or not the part needs to be replaced or inspected is output. Information processing methods.

2. The vehicle information includes maintenance information regarding maintenance of the vehicle to be inspected. The information processing method according to claim 1 .

3. The maintenance information includes information regarding the number of days since the last maintenance of the vehicle to be inspected, the mileage, the number of days since the first year of registration, the number of inspections, or the replacement of parts. The information processing method according to claim 2 .

4. The necessity information includes whether or not the part needs to be replaced or inspected, the likelihood of the necessity, or a classification of the likelihood of the necessity.

3. The information processing method according to claim 1.

5. The items, the necessity information, and the main factors among the vehicle information that had the greatest effect on the output of the necessity information are listed and output.

3. The information processing method according to claim 1.

6. A command for creating a first proposal for the vehicle inspection company in which the necessity information is described, and a command for creating a second proposal for the customer, which is different from the first proposal, are output.

3. The information processing method according to claim 1.

7. In the second proposal, the item, the necessity information, a major factor among the vehicle information that had the greatest effect on the output of the necessity information, and a message corresponding to the item and the major factor are output. The information processing method according to claim 6.

8. Outputting the items for which the likelihood of the need for part replacement or inspection is equal to or greater than a predetermined value 3. The information processing method according to claim 1.

9. A table is referred to, which stores required items that must be written in the proposal regardless of the necessity information, and the required items are written in the proposal regardless of the output of the necessity information.

3. The information processing method according to claim 1.

10. A table is referred to which stores free items for which no fee is required for replacement or inspection of the part, and among the items for which necessity information has been output, the items corresponding to the free items are written in the proposal.

3. The information processing method according to claim 1.

11. Acquire training data including vehicle information on a vehicle to be inspected and necessity information on items that require replacement or inspection of parts of the vehicle to be inspected; A learning model is generated that outputs the necessity information when the vehicle information is input based on the acquired training data. How to generate the model.

12. The original data of the training data is collected from store devices installed in stores of a plurality of vehicle inspection companies, the training data is generated by converting the store management number included in the original data into the common management number by referring to a table that stores a correspondence between a store management number assigned to each item, which differs for each of the store devices, and a common management number common to all stores; The vehicle information and the necessity information included in the training data are associated with each other by the common management number. A method for generating a model according to claim 11.

13. Obtain vehicle information on vehicles subject to inspection, The acquired vehicle information is input to a learning model that has been trained to output information on whether or not a part needs to be replaced or inspected when the vehicle information is input, and the information on whether or not the part needs to be replaced or inspected is output. A computer program that causes a computer to carry out processing.

14. Obtain vehicle information on vehicles subject to inspection, The acquired vehicle information is input to a learning model that has been trained to output information on whether or not a part needs to be replaced or inspected when the vehicle information is input, and the information on whether or not the part needs to be replaced or inspected is output. Processing section An information processing device comprising:

Images