JP2023178787A - Price information providing device - Google Patents

Abstract

To provide, to a user, appropriate price information of an industrial product on which a learned machine learning model is installed instead of a sensor.SOLUTION: A price information providing device includes: a price determination unit 17 which determines a price of an industrial product on which a learned machine learning model is installed instead of a prescribed sensor; and an information providing unit 18 which provides the user with the price information regarding the price of the industrial product in response to a request from the user. When a large number of training data sets have been used for the training of the machine learning model, the price determination unit sets the price of the industrial product higher than in cases with fewer training data sets.SELECTED DRAWING: Figure 5

Description

The present invention relates to a price information providing device.

Patent Document 1 describes that a machine learning model is trained using data acquired by a sensor in a vehicle, and the learned machine learning model is used for controlling another vehicle.

JP2020-067911A

Such a machine learning model can output a predetermined output parameter from a predetermined input parameter, and can be used as a substitute for a sensor that outputs the same output parameter. Basically, industrial products (such as vehicles) equipped with trained machine learning models can be manufactured at a lower cost than industrial products equipped with sensors.

However, at present, no method has been established for appropriately evaluating the value of trained machine learning models and providing appropriate price information to users.

Therefore, in view of the above-mentioned problems, an object of the present invention is to provide users with appropriate price information on industrial products that are equipped with trained machine learning models instead of sensors.

In order to solve the above problems, the present invention includes a price determining unit that determines the price of an industrial product equipped with a learned machine learning model instead of a predetermined sensor; an information providing unit that provides the user with price information regarding the price of the product, and when the number of training datasets used for learning the machine learning model is large, the price determining unit Provided is a price information providing device that makes the price of the industrial product higher than when the price is low.

According to the present invention, it is possible to provide a user with appropriate price information for an industrial product equipped with a trained machine learning model instead of a sensor.

FIG. 1 is a schematic configuration diagram of a client server system including a price information providing device according to a first embodiment of the present invention. FIG. 2 is a diagram schematically showing the configuration of the server. FIG. 3 is a functional block diagram of the processor of the server in the first embodiment. FIG. 4 is a flowchart showing a control routine for price determination processing in the first embodiment. FIG. 5 is a flowchart showing a control routine for information provision processing in the first embodiment. FIG. 6 is a diagram showing an example of a price display screen displayed on a user terminal in the first embodiment. FIG. 7 is a functional block diagram of a processor of a server in the second embodiment. FIG. 8 is a flowchart showing a control routine for information provision processing in the second embodiment. FIG. 9 is a diagram showing an example of a price display screen displayed on a user terminal in the second embodiment.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In addition, in the following description, the same reference number is attached to the same component.

<First embodiment>
First, a first embodiment of the present invention will be described with reference to FIGS. 1 to 6. FIG. 1 is a schematic configuration diagram of a client server system 100 including a price information providing device according to a first embodiment of the present invention. The client server system 100 includes a server 1, multiple vehicles 2, and multiple user terminals 3.

The vehicle 2 includes a communication device (for example, a data communication module (DCM)) that enables communication between the vehicle 2 and the outside of the vehicle 2. The communication device connects the vehicle 2 to the communication network 4 via the radio base station 5 by accessing the radio base station 5 . Vehicle 2 communicates with server 1 via wireless base station 5 and communication network 4 . The vehicle 2 also includes a plurality of sensors that detect state quantities of the vehicle 2, and data acquired by the plurality of sensors is transmitted from the vehicle 2 to the server 1.

The user terminal 3 is, for example, a smartphone, a tablet terminal, a personal computer, etc., and is owned by the user or installed in a store or the like. The user terminal 3 includes a processor that performs various processes on the user terminal 3, an input section (touch panel, operation buttons, microphone, keyboard, mouse, etc.), an output section (display, speaker, etc.), and a communication section. The communication unit connects the user terminal 3 to the communication network 4 via the radio base station 5 by accessing the radio base station 5 . User terminal 3 communicates with server 1 via wireless base station 5 and communication network 4 . Note that the user terminal 3 may be connected to the communication network 4 without going through the wireless base station 5.

FIG. 2 is a diagram schematically showing the configuration of the server 1. As shown in FIG. As shown in FIG. 2, the server 1 includes a communication interface 11, a storage device 12, a memory 13, and a processor 14. Communication interface 11, storage device 12, and memory 13 are connected to processor 14 via signal lines. Note that the server 1 may further include an input device such as a keyboard and a mouse, an output device such as a display, and the like. Further, the server 1 may be composed of a plurality of computers.

Communication interface 11 has an interface circuit for connecting server 1 to communication network 4 . The server 1 communicates with the outside of the server 1 (for example, a plurality of vehicles 2) via a communication interface 11 and a communication network 4. The communication interface 11 is an example of a communication section of the server 1.

The storage device 12 includes, for example, a hard disk drive (HDD), a solid state drive (SDD), or an optical recording medium and its access device. The storage device 12 stores various data, such as computer programs for the processor 14 to execute various processes. The storage device 12 is an example of a storage unit of the server 1.

The memory 13 includes a nonvolatile semiconductor memory (eg, RAM). The memory 13 temporarily stores various data used when the processor 14 executes various processes, for example. The memory 13 is another example of the storage section of the server 1.

The processor 14 includes one or more CPUs and their peripheral circuits, and executes various processes. Note that the processor 14 may further include other arithmetic circuits such as a logic arithmetic unit, a numerical arithmetic unit, or a graphic processing unit.

In this embodiment, the server 1 uses data transmitted from the vehicle 2 to train a machine learning model that is used as a sensor substitute in the vehicle. The server 1 also functions as a price information providing device, and provides the user with price information for the user to compare and consider multiple types of vehicles (for example, new cars) with different sensor mounting conditions.

FIG. 3 is a functional block diagram of the processor 14 of the server 1 in the first embodiment. In this embodiment, the processor 14 includes a data acquisition section 15, a learning section 16, a price determination section 17, and an information provision section 18. The data acquisition unit 15, the learning unit 16, the price determination unit 17, and the information providing unit 18 are functional modules realized by the processor 14 of the server 1 executing a computer program stored in the storage device 12 of the server 1. . Note that these functional modules may be realized by a dedicated arithmetic circuit provided in the processor 14.

A machine learning model outputs predetermined output parameters from a plurality of input parameters. In a vehicle equipped with a machine learning model instead of a sensor, the vehicle is controlled based on the output of the machine learning model instead of the output of the sensor. In learning such a machine learning model, a training data set is used, which is a combination of actually measured values of a plurality of input parameters and actually measured values (correct data) of predetermined output parameters corresponding to these actually measured values.

The data acquisition unit 15 receives data from a plurality of vehicles 2, and creates a training data set by combining corresponding actual measured values of input parameters and output parameters of the machine learning model. The training data set created by the data acquisition unit 15 is stored in the storage device 12 of the server 1. Note that the training data set may be created by a person such as an operator of the server 1.

The learning unit 16 performs learning of a machine learning model using the training data set created by the data acquisition unit 15. For example, when the machine learning model is a neural network model, the learning unit 16 uses a training data set to reduce the difference between the output value of the neural network model and the actual measured value of the output parameter. Iteratively updates the weights and biases in the neural network model using error backpropagation. As a result, the neural network model is trained, and a trained neural network model is generated. Information regarding the machine learning model generated by the learning unit 16 (for example, weights, biases, etc. of the neural network model) is stored in the storage device 12 of the server 1.

The price determination unit 17 determines the price of the vehicle when a predetermined sensor in the vehicle is replaced with a trained machine learning model, that is, the price of a vehicle in which a trained machine learning model is installed instead of the predetermined sensor. do. Basically, the greater the number of training datasets used to train a machine learning model, the more accurate the trained machine learning model will be, and the more reliable the vehicle in which the trained machine learning model will be installed will be. do. Therefore, when the number of training datasets used for learning the machine learning model is large, the price determining unit 17 determines that the amount of training data set used for learning the machine learning model is higher than that of the trained Increase the price of vehicles equipped with machine learning models. For example, the price determination unit 17 increases the price of a vehicle equipped with a trained machine learning model as the number of training datasets used for learning increases. This makes it possible to set appropriate prices for vehicles equipped with trained machine learning models.

The information providing unit 18 provides the user with price information regarding the price of the vehicle determined by the price determining unit 17 in response to a request from the user. This makes it possible to provide users with appropriate price information for vehicles equipped with trained machine learning models instead of sensors.

The control flow for the above processing will be described below with reference to the flowcharts of FIGS. 4 and 5. FIG. 4 is a flowchart showing a control routine for price determination processing in the first embodiment. This control routine is repeatedly executed by the processor 14 of the server 1, and when there are multiple machine learning models used in the vehicle, it is executed for each of the multiple machine learning models.

First, in step S101, the learning unit 16 determines whether the number of training datasets newly created by the data acquisition unit 15 after the previous learning of the machine learning model is greater than or equal to a predetermined number. do. If it is determined that the number of newly created training data sets is less than the predetermined number, this control routine ends. On the other hand, if it is determined that the number of newly created training data sets is greater than or equal to the predetermined number, the control routine proceeds to step S102. Note that if the machine learning model has never been trained, the determination in step S101 is affirmed when the number of training data sets created by the data acquisition unit 15 reaches a predetermined number, and the present control routine The process proceeds to step S102.

In step S102, the learning unit 16 performs learning of the machine learning model using a predetermined number of newly created training data sets. For example, if the machine learning model is a neural network model, the learning unit 16 updates predetermined parameters (weights and biases) of the machine learning model using a known error backpropagation method.

Next, in step S103, the price determination unit 17 determines the price of a vehicle (hereinafter referred to as a "sensorless vehicle") in which a trained machine learning model is installed instead of a predetermined sensor. Specifically, the price determination unit 17 lowers the price of the sensorless vehicle by making the price of the sensorless vehicle after the learning in step S102 higher than the price of the sensorless vehicle before the learning in step S102. Update. That is, when the number of training data sets used for learning the machine learning model is large, the price determination unit 17 determines that the price of the sensorless vehicle is higher than when the number of training data sets used for learning the machine learning model is small. raise the price.

In this embodiment, price information regarding the price of a sensorless vehicle is calculated as the difference between the price of the sensorless vehicle and the price of a vehicle equipped with a predetermined sensor, and is stored in the storage device 12 of the server 1. For example, when the price of a sensorless vehicle is 30,000 yen cheaper than the price of a vehicle equipped with a predetermined sensor, the price information regarding the price of the sensorless vehicle is -30,000 yen. After step S103, this control routine ends.

FIG. 5 is a flowchart showing a control routine for information provision processing in the first embodiment. This control routine is repeatedly executed by the processor 14 of the server 1 at predetermined execution intervals.

First, in step S201, the information providing unit 18 determines whether the user has requested information provision. For example, when an information request signal is transmitted from the user terminal 3 to the server 1, the information providing unit 18 determines that the user has requested information provision. For example, the user performs an operation to request price information of a specific vehicle on an application or website that presents vehicle information, and in response to this operation, an information request signal is transmitted from the user terminal 3 to the server 1. .

If it is determined in step S201 that the user has not requested information provision, this control routine ends. On the other hand, if it is determined in step S201 that the user has requested information provision, the control routine proceeds to step S202.

In step S202, the information providing unit 18 acquires price information regarding the price of a sensorless vehicle corresponding to a specific vehicle specified by the user from the storage device 12.

Next, in step S203, the information providing unit 18 provides the user with price information regarding the price of the sensorless vehicle. Specifically, the information providing unit 18 transmits price information to the user terminal 3. As a result, the user terminal 3 displays the price information on the output section of the user terminal 3. After step S203, this control routine ends.

FIG. 6 is a diagram showing an example of a price display screen displayed on the user terminal 3 in the first embodiment. In the example in Figure 6, price information is shown when the exhaust temperature sensor in the vehicle is replaced with a trained machine learning model (AI1), and when the air-fuel ratio sensor is replaced in the vehicle by the trained machine learning model (AI2). Price information is shown at the time. Note that the price information regarding the price of the sensorless vehicle may be the price of the sensorless vehicle, the price of a trained machine learning model that is replaced with a predetermined sensor, or the like.

<Second embodiment>
The configuration and control of the price information providing apparatus according to the second embodiment are basically the same as the configuration and control of the price information providing apparatus according to the first embodiment, except for the points described below. Therefore, the second embodiment of the present invention will be described below, focusing on the differences from the first embodiment.

FIG. 7 is a functional block diagram of the processor 14 of the server 1 in the second embodiment. In the second embodiment, the processor 14 includes a sensor selection section 19 in addition to a data acquisition section 15, a learning section 16, a price determination section 17, and an information provision section 18. These are functional modules realized by the processor 14 of the server 1 executing a computer program stored in the storage device 12 of the server 1. Note that these functional modules may be realized by a dedicated arithmetic circuit provided in the processor 14.

As described above, the information providing unit 18 provides the user with price information regarding the price of the sensorless vehicle. However, if there are multiple sensors that can be replaced with a trained machine learning model in a vehicle that the user wants to purchase, it is difficult for the user to appropriately judge which sensor should be replaced with the machine learning model. It is.

Therefore, in the second embodiment, the sensor selection unit 19 selects a sensor suitable for replacement with a learned machine learning model based on preference information input in advance by the user, and the information provision unit 18 A proposal is made to the user to replace the sensor selected by the sensor selection unit 19 with a trained machine learning model. This can assist the user in appropriately selecting a trained machine learning model to be replaced by the sensor.

FIG. 8 is a flowchart showing a control routine for information provision processing in the second embodiment. This control routine is repeatedly executed by the processor 14 of the server 1 at predetermined execution intervals.

First, similar to step S201 in FIG. 5, in step S301, the information providing unit 18 determines whether the user has requested information provision. If it is determined that the user does not request information provision, this control routine ends. On the other hand, if it is determined in step S301 that the user has requested information provision, the control routine proceeds to step S302.

In step S302, similar to step S202 in FIG. 5, the information providing unit 18 acquires price information regarding the price of the sensorless vehicle corresponding to the specific vehicle designated by the user from the storage device 12.

Next, in step S303, the sensor selection unit 19 selects a sensor suitable for replacement with the learned machine learning model based on preference information input in advance by the user. For example, the user inputs preference information into the user terminal 3 in advance, and the preference information is transmitted from the user terminal 3 to the server 1 along with an information request signal. The preference information is, for example, performance (eg, acceleration performance, fuel efficiency, ride comfort, etc.) that the user preferentially desires from the vehicle when purchasing the vehicle. For example, when acceleration performance is selected as preference information, a sensor other than a sensor that outputs a parameter that has a strong correlation with acceleration performance is selected as a sensor suitable for replacement with a learned machine learning model.

Next, in step S304, the information providing unit 18 provides the user with price information regarding the price of the sensorless vehicle. At this time, the information providing unit 18 suggests to the user that the sensor selected by the sensor selection unit 19 be replaced with a trained machine learning model. For example, the information providing unit 18 instructs the user terminal 3 to highlight and display a trained machine learning model that is replaced by the sensor selected by the sensor selection unit 19. After step S304, this control routine ends.

FIG. 9 is a diagram showing an example of a price display screen displayed on the user terminal 3 in the second embodiment. In the example of FIG. 9, the display of the learned machine learning model (AI1) that is replaced by the exhaust temperature sensor is highlighted with an asterisk.

<Other embodiments>
Although the preferred embodiments of the present invention have been described above, the present invention is not limited to these embodiments, and various modifications and changes can be made within the scope of the claims.

For example, the price determining unit 17 determines the price of the sensorless vehicle when the trained machine learning model is updated, and the price of the sensorless vehicle when the trained machine learning model is not updated, and the information providing unit 18 may provide the user with price information regarding the prices of both. This allows the user to select whether or not to update the trained machine learning model, taking into account the price difference between the two. When a trained machine learning model is updated, information on the machine learning model that has been further trained in the server 1 (for example, updated weights and bias values) is sent from the server 1 to the vehicle, and The trained machine learning model used in is updated.

Furthermore, a trained machine learning model may be installed in place of a predetermined sensor in an industrial product other than a vehicle, and price information regarding the price of such an industrial product may be provided to the user. Further, the trained machine learning model that is replaced with a predetermined sensor in an industrial product may be a machine learning model other than a neural network such as a random forest, a k-nearest neighbor method, or a support vector machine.

1 Server 2 Vehicle 14 Processor 17 Price determining section 18 Information providing section

Claims (1)

a price determination unit that determines the price of an industrial product equipped with a learned machine learning model instead of a predetermined sensor;
an information providing unit that provides price information regarding the price of the industrial product to the user in response to a request from the user;
The price determination unit provides price information that increases the price of the industrial product when the number of training data sets used for learning the machine learning model is large compared to when the number of training data sets is small. Device.

Images