JP7553275B2 - PROGRAM, INFORMATION PROCESSING APPARATUS, INFORMATION PROCESSING METHOD, AND METHOD FOR GENERATING LEARNING MODEL - Google Patents

Description

The present disclosure relates to a program, an information processing device, an information processing method, and a method for generating a learning model.

Patent Document 1 proposes a system that estimates the amount of contamination in each indoor area based on relationship information that indicates the relationship between the characteristics of human use of the area and the amount of contamination by pollutants. The system disclosed in Patent Document 1 acquires usage characteristics such as purpose, floor area, and number of occupants for each indoor area, and estimates the contamination situation based on the usage characteristics.

International Publication No. 2018/147075

However, pathogens such as viruses and bacteria that cause infectious diseases can be either infectious or non-infectious, so even if the amount of contamination is high, it cannot necessarily be said that the risk of infection is high. Therefore, the system disclosed in Patent Document 1 has the problem that it is difficult to estimate the risk of infection caused by pathogens.

This disclosure has been made in light of these circumstances, and its purpose is to provide a program or the like capable of estimating the risk of infection by a pathogen.

A program according to one aspect of the present disclosure causes a computer to execute a process of acquiring environmental information, including temperature, humidity, or sunlight, at each location within a facility, identifying the pathogen infection risk at each location based on the acquired environmental information, and outputting the identified infection risk for each location.

This disclosure makes it possible to estimate the risk of infection by pathogens.

1 is a block diagram showing an example of the configuration of an information processing device; FIG. 2 is a schematic diagram showing an example of the configuration of an environmental information DB; FIG. 2 is a schematic diagram showing an example of the configuration of a usage status DB; FIG. 2 is a schematic diagram showing an example of the configuration of an infection risk determination model. 13 is a flowchart illustrating an example of a registration process procedure. FIG. 13 is a schematic diagram showing an example of a screen. 13 is a flowchart illustrating an example of a procedure for a process of determining an infection risk. FIG. 13 is a schematic diagram showing an example of a screen. FIG. 13 is a schematic diagram showing an example of a screen. FIG. 13 is a schematic diagram showing another example of the determination screen. FIG. 2 is a block diagram showing a configuration example of a learning device. 13 is a flowchart showing an example of a generation process procedure of an infection risk determination model. FIG. 13 is a schematic diagram showing an example of the configuration of an infection risk determination model of embodiment 3. 13 is a flowchart illustrating an example of a procedure for a process of determining an infection risk according to the fourth embodiment. FIG. 13 is a schematic diagram showing an example of a screen.

Below, the program, information processing device, information processing method, and learning model generation method disclosed herein will be described in detail with reference to drawings showing embodiments thereof.

(Embodiment 1)
An information processing device that determines the infection risk of pathogens for each location in a facility based on the environment of each location will be described. In the present disclosure, the facilities for which the infection risk of pathogens is determined can be schools, medical institutions, nursing homes, public transportation stations and airports, stores, etc. The locations for which the infection risk is determined can be various locations in the facility, such as doorknobs (door handles), operation panels of elevators, handrails of stairs and escalators, ticket slots and card reading units of automatic ticket gates installed at ticket gates of stations and boarding gates of airports, operation buttons and coin slots of vending machines and automatic ticket vending machines, top surfaces of desks and tables, handles of drawers and storage shelves, touch panels of tablet terminals, etc. In particular, it is preferable to determine the infection risk from locations where pathogens are likely to exist, locations that are frequently touched by people, etc. In addition, in the locations for which the infection risk is determined, a test sample is collected from each location by wiping the surface of each location with a collection tool such as a cotton swab, and an examination is performed to measure the amount of each pathogen present in the test sample.

FIG. 1 is a block diagram showing an example of the configuration of an information processing device. The information processing device 10 of this embodiment is configured using, for example, a server computer, a personal computer, a tablet terminal, or the like. The information processing device 10 may be configured to perform distributed processing by providing a plurality of devices, may be realized by a plurality of virtual machines provided in one device, or may be realized by using a cloud server. The information processing device 10 includes a control unit 11, a storage unit 12, a communication unit 13, an input unit 14, a display unit 15, a reading unit 16, and the like, and these units are connected to each other via a bus. The control unit 11 includes one or more processors such as a CPU (Central Processing Unit), an MPU (Micro-Processing Unit), or a GPU (Graphics Processing Unit). The control unit 11 appropriately executes a control program 12P stored in the storage unit 12 to perform various information processing, control processing, and the like that should be performed by the information processing device 10.

The storage unit 12 includes a RAM (Random Access Memory), a flash memory, a hard disk, an SSD (Solid State Drive), etc. The storage unit 12 prestores the control program 12P executed by the control unit 11 and various data necessary for the execution of the control program 12P. The storage unit 12 also temporarily stores data generated when the control unit 11 executes the control program 12P. The storage unit 12 also stores an infection risk determination model 12M, which is a trained model constructed by machine learning, for example. The infection risk determination model 12M is expected to be used as a program module that functions as part of artificial intelligence software. The infection risk determination model 12M is a trained model that has been trained to output information related to the infection risk of a pathogen in a location to be determined when various information on the location to be determined for infection risk is input. The infection risk determination model 12M is provided for each pathogen to be determined for infection risk. In this embodiment, the pathogens to be judged are, for example, coronaviruses (SARS-Cov, MERS-Cov, SARS-Cov-2, etc.), influenza viruses, hepatitis B viruses, hepatitis C viruses, noroviruses, pneumococci, adenoviruses, etc., and infection risk judgment models 12M corresponding to these pathogens are provided. The trained model performs a predetermined calculation on the input values and outputs the calculation results, and the storage unit 12 stores data such as coefficients and thresholds of functions that define this calculation as the infection risk judgment model 12M. The storage unit 12 also stores an environmental information DB (database) 12a and a usage status DB 12b, which will be described later. The environmental information DB 12a and the usage status DB 12b may be stored in another storage device connected to the information processing device 10, or may be stored in another storage device with which the information processing device 10 can communicate.

The communication unit 13 is an interface for connecting to a network such as the Internet by wired or wireless communication, and transmits and receives information to and from other devices via the network. The input unit 14 accepts operation input by the user and sends a control signal corresponding to the operation content to the control unit 11. The display unit 15 is a liquid crystal display or an organic EL display, etc., and displays various information according to instructions from the control unit 11. The input unit 14 and the display unit 15 may be a touch panel configured as an integrated unit.

The reading unit 16 reads information stored in portable storage medium 1a, including CD (Compact Disc)-ROM, DVD (Digital Versatile Disc)-ROM, USB (Universal Serial Bus) memory, SD (Secure Digital) card, etc. The control program 12P and various data stored in the storage unit 12 may be read by the control unit 11 from the portable storage medium 1a via the reading unit 16 and stored in the storage unit 12. The control program 12P and various data stored in the storage unit 12 may also be downloaded by the control unit 11 from another device via the communication unit 13 and stored in the storage unit 12.

Figure 2 is a schematic diagram showing an example of the configuration of the environmental information DB 12a. The environmental information DB 12a stores information about the environment of a place to be judged for infection risk. The environmental information DB 12a shown in Figure 2 includes a facility ID column, a facility name column, a location ID column, a location information column, a material column, an environmental information column, etc., and stores various information on each location in the facility in association with the facility ID and the location ID. The facility ID column stores identification information (facility ID) assigned to each facility to be judged for infection risk, and the facility name column stores the name of each facility in association with the facility ID. The location ID column stores identification information (location ID) assigned to each location (place) to be judged for infection risk in the facility in association with the facility ID, and the location information column stores information for identifying each location in association with the location ID. The material column stores the material of each location to be judged for infection risk in association with the location ID. Materials that can be used include, for example, steel, stainless steel, aluminum, wood, paper, plastic, glass, vinyl, silicone rubber, clothing, ceramics, Teflon (registered trademark), etc. The environmental information column stores information about the environment of each location in association with the location ID. Environmental information includes, for example, temperature, humidity, presence or absence of direct sunlight, hours of sunshine, amount of solar radiation, wind speed, weather, season, etc., and the measurement results of each item are stored in association with the measurement date and time (date and time). Note that the environmental information is not limited to information at the time when the user measures using various measuring instruments, but may be the maximum and minimum temperature, maximum and minimum humidity, etc. on the day of measurement, or average values for each season or weather may be used.

The facility ID stored in the environmental information DB 12a is issued and stored by the control unit 11 when a new facility is registered. The location ID stored in the environmental information DB 12a is issued and stored by the control unit 11 when a new location within the facility is registered. Other information stored in the environmental information DB 12a is stored by the control unit 11 when the control unit 11 acquires the information via the input unit 14 or the communication unit 13. The contents stored in the environmental information DB 12a are not limited to the example shown in FIG. 2, and various information related to the facility and each location within the facility may be stored. For example, the results (measurement results) of a test to measure the amount of each pathogen present on a test sample collected at each location may be stored in the environmental information DB 12a.

Figure 3 is a schematic diagram showing an example of the configuration of the usage status DB 12b. The usage status DB 12b stores information about the usage status (usage state) of users for each location in each facility registered in the environmental information DB 12a. The usage status DB 12b shown in Figure 3 includes a facility ID column, a facility name column, a location ID column, a location information column, a usage information column, etc., and stores each piece of information about the usage status in association with the facility ID and the location ID. The facility ID column, the facility name column, the location ID column, and the location information column store the facility ID, facility name, location ID, and location information registered in the environmental information DB 12a, respectively. The usage information column stores each piece of information about the usage status of users at each location in association with the location ID. The usage status information includes, for example, the number of users, the number of contacts, the average contact time, and the contact location, and each piece of information is stored in association with a date. The number of users is the number of users at each location, for example, in one day. The number of users may be the number of users counted using a human sensor, or may be the number of users for each predetermined time period. The number of contacts is the number of times that a user contacts each location when using the location, and may be the total number of contacts. For example, when the determination location is an operation button, the number of operations may be used as the number of contacts. The average contact time is the average time that a contact state continues with one contact action by a user at each location, and the contact location is a part of the body when the user contacts each location, such as a fingertip or a palm. The number of contacts, the average contact time, and the contact location can be detected based on image data of a user photographed using a camera, for example. The facility ID, facility name, location ID, and location information stored in the usage status DB 12b are each information registered in the environmental information DB 12a and stored in advance by the control unit 11. Other information stored in the usage status DB 12b is stored by the control unit 11, for example, when the control unit 11 detects each information or when the control unit 11 acquires each information via the input unit 14. The contents stored in the usage status DB 12b are not limited to the example shown in FIG. 3A, and various information related to the usage status at each location within the facility may be stored. In addition, the usage status DB 12b may be configured not to include a facility name column and a location information column.

FIG. 4 is a schematic diagram showing an example of the configuration of the infection risk determination model 12M. The infection risk determination model 12M is configured with a learning model such as a decision tree, a random forest, or an SVM (Support Vector Machine). The infection risk determination model 12M may be configured using various algorithms such as a CNN (Convolution Neural Network), an RNN (Recurrent Neural Network), or an LSTM (Long Short-Term Memory), or may be configured by combining a plurality of algorithms. The infection risk determination model 12M of this embodiment is provided for each pathogen to be determined, and each infection risk determination model 12M can be realized with a similar configuration. The infection risk determination model 12M receives information about a location to be determined for infection risk via each input node of the input layer. Specifically, environmental information about the location to be determined is input via input node 0, information indicating the usage status (usage information) of the location to be determined is input via input node 1, and information indicating the material of the location to be determined (material information) is input via input node 2. Each piece of information input through the input layer is input to an intermediate layer (hidden layer). The intermediate layer calculates an output value from each piece of input information using various functions and thresholds, and outputs the calculated output value from each output node of the output layer. As a result, the infection risk determination model 12M of this embodiment outputs information related to the infection risk of a pathogen in a place to be determined when environmental information, usage information, and material information related to the place to be determined are input. The infection risk determination model 12M of this embodiment is configured to determine whether the infection risk is high, medium, or low. Therefore, in the infection risk determination model 12M, the output layer has three output nodes 0 to 2, and output node 0 outputs the probability that the risk should be determined to be high, output node 1 outputs the probability that the risk should be determined to be medium, and output node 2 outputs the probability that the risk should be determined to be low. Note that the output value of each output node is, for example, a value of 0 to 1, and the sum of the probabilities output from the three output nodes is 1.0 (100%). In addition to being configured to output discrimination probabilities corresponding to high, medium, and low infection risks, the infection risk determination model 12M may also be configured to output discrimination probabilities for each value indicating the degree of infection risk on a scale of four or more levels.

The infection risk determination model 12M learns using training data including information on a location to be determined for infection risk and information (correct answer label) indicating the level of infection risk (high, medium, low). When location-related information included in the training data is input, the infection risk determination model 12M learns so that the output value from the output node corresponding to the correct answer label included in the training data approaches 1 and the output values from the other output nodes approach 0. In the learning process, the infection risk determination model 12M optimizes data such as coefficients and thresholds of various functions that define a predetermined calculation performed on the input value. This allows for the acquisition of a trained infection risk determination model 12M that has been trained to output information indicating the infection risk of a pathogen when location-related information is input. The learning of the infection risk determination model 12M is performed, for example, on another learning device. The trained infection risk determination model 12M generated by learning on the other learning device is downloaded from the learning device to the information processing device 10 via a network or a portable storage medium 1a, for example, and stored in the storage unit 12.

The infection risk determination model 12M is not limited to the configuration shown in FIG. 4. For example, the infection risk determination model 12M may be configured to input the measurement results of the amount of pathogens detected from a test sample collected at the location to be determined, cleaning information indicating the cleaning situation (cleaning state) at the location to be determined, and the like. The cleaning information may include, for example, whether or not disinfection work can be performed at each location, the type of disinfectant that can be used, and the frequency at which disinfection work should be performed. By using such information, the infection risk determination model 12M may be configured to output information indicating the infection risk of pathogens when each of these pieces of information is input.

The following describes the process of registering information about locations that should be subject to infection risk assessment in the information processing device 10 of this embodiment. Figure 5 is a flowchart showing an example of the registration process procedure, and Figure 6 is a schematic diagram showing an example screen. The following process is realized by the control unit 11 in accordance with the control program 12P stored in the memory unit 12 of the information processing device 10. Part of the following process may be realized by a dedicated hardware circuit.

In the information processing device 10 of this embodiment, when a user wants to determine the risk of pathogen infection for any location in a facility, the user registers information about the location to be determined in the information processing device 10. For example, when a user collects a sample for testing at each location in the facility, the user may measure the environmental information of the location and register it in the information processing device 10. The control unit 11 of the information processing device 10 displays a registration screen for registering environmental information on the display unit 15 in accordance with the user's operation via the input unit 14 (S11). FIG. 6 shows an example of a registration screen, and the registration screen shown in FIG. 6 has an input field for the facility name, an input field for the weather, a selection field for the measurement location, an input field for the environmental information of each measurement location, and the like. The input field for the facility name has a pull-down menu for selecting one of the facility names already registered in the information processing device 10. Note that a new registration button for instructing the registration of a new facility name is provided near the input field for the facility name. The input field for the weather has a pull-down menu for selecting one of the weather conditions registered in advance. Weather conditions include, for example, sunny, cloudy, rain, and snow. The measurement location selection field displays, for example, a floor map of the facility, and has check boxes for selecting each location on the floor map that may be selected. The measurement location selection field may be configured to allow any location on the floor map to be specified. The environmental information input field has input fields for measurement date and time, temperature, humidity, direct sunlight, ventilation, wind speed, sunshine hours, etc. for each measurement location selected via the measurement location selection field. The measurement date and time input field has an input field in which any date can be input, and a pull-down menu for selecting any time (hour and minute). The temperature, humidity, wind speed, and sunshine hours input fields are configured to allow any numerical value to be input. The direct sunlight input field has a pull-down menu for selecting the direct sunlight irradiation state, such as hit or not hit, and the ventilation input field has a pull-down menu for selecting the ventilation state, such as good or bad. The registration screen has a registration button (not shown) for instructing the registration of the input contents. The environmental information registration screen is not limited to the configuration shown in FIG. 6. For example, the registration screen may display an input field for the test items to be performed on the test samples collected at each measurement location.

In the registration screen shown in FIG. 6, the user inputs the contents to be registered in each input field via the input unit 14, and operates the registration button when instructing to register the input contents. The control unit 11 accepts the registration contents for each input field in the registration screen (S12), and displays the accepted information in each input field. The control unit 11 then determines whether the registration button on the registration screen has been operated (S13), and if it is determined that it has not been operated (S13: NO), repeats the processing of step S12. If it is determined that the registration button has been operated (S13: YES), the control unit 11 (acquisition unit) acquires each piece of information via the registration screen, and registers each piece of acquired information in the environmental information DB 12a (S14). Specifically, the control unit 11 stores the measurement date and time and the measured environmental information in the environmental information DB 12a in association with the location ID of each measurement location in the selected facility (facility name). If a measurement location is selected whose location ID is not stored in the environmental information DB 12a, the control unit 11 issues a new location ID and registers the measurement date and time and environmental information in the environmental information DB 12a in association with the location ID.

By the above-mentioned process, the date and time when the environmental information was measured and the environmental information such as the temperature and humidity at the time of measurement can be registered in the environmental information DB 12a for each location in the facility. Note that the environmental information is not limited to actual values measured by a user using a measuring device, but may be information predicted by a weather forecast, for example. Furthermore, the environmental information is not limited to a configuration in which it is registered in the environmental information DB 12a by a user's operation input via the input unit 14. For example, IOT (Internet of Things) devices such as a temperature sensor, humidity sensor, sunshine sensor, wind speed sensor, etc. that can be connected to a network may be installed in each location, and the information processing device 10 may acquire the measurement results from these sensors and register them in the environmental information DB 12a. Note that when the test samples collected at each location are tested for each test item and the test results (measurement results) are obtained, for example, the user may register the measurement results in the environmental information DB 12a via the input unit 14.

Next, a process for determining the risk of pathogen infection at each location using the environmental information for each location registered in the environmental information DB 12a by the above-mentioned process will be described. FIG. 7 is a flowchart showing an example of the infection risk determination process procedure, and FIG. 8 and FIG. 9 are schematic diagrams showing example screens. The following process is realized by the control unit 11 in accordance with the control program 12P stored in the storage unit 12 of the information processing device 10. A part of the following process may be realized by a dedicated hardware circuit. Here, it is assumed that the environmental information for each location has already been registered in the environmental information DB 12a, and the usage status of each location has already been registered in the usage status DB 12b. Note that the usage status (usage information) registered in the usage status DB 12b may be input by the user via the input unit 14, or the control unit 11 may use the usage status detected based on image data of the user captured using a camera, for example.

In the information processing device 10 of this embodiment, when a user wants to judge the infection risk of a pathogen for a facility registered in the environmental information DB 12a, the user performs a predetermined operation via the input unit 14. The control unit 11 of the information processing device 10 displays on the display unit 15 a selection screen (input screen for judgment target) for selecting a judgment target facility for judging the infection risk of a pathogen according to the user's operation via the input unit 14 (S21). FIG. 8 shows an example of the input screen, and the input screen shown in FIG. 8 has an input field for a facility and an input field for an examination item. The input field for the facility has a pull-down menu for selecting one of the facilities registered in the environmental information DB 12a, and the input field for the examination item has a pull-down menu for selecting one of the examination items prepared in advance. The input screen has an OK button for instructing the execution of the infection risk judgment process for the facility and the examination item inputted in the input field for the facility and the input field for the examination item, respectively. The input screen for the judgment target is not limited to the configuration shown in FIG. 8.

In the input screen shown in FIG. 8, the user inputs the facility and test items to be judged in each input field via the input unit 14 and operates the OK button. The control unit 11 accepts the facility and test items to be judged by accepting input for each input field in the input screen (S22), and displays the accepted contents in each input field. The control unit 11 then determines whether the OK button on the input screen has been operated (S23), and if it is determined that it has not been operated (S23: NO), it repeats the processing of step S22. If it is determined that the OK button has been operated (S23: YES), the control unit 11 reads out the infection risk judgment model 12M corresponding to the test item selected via the input screen from the memory unit 12 (S24). The control unit 11 then reads out the registration information for one measurement location for the facility to be judged accepted via the input screen (S25). Specifically, for one measurement location, the control unit 11 reads out environmental information for the date (measurement date and time) for which the infection risk is to be determined from the environmental information DB 12a, and reads out usage information for the same date at the same measurement location from the usage status DB 12b. The control unit 11 also reads out the material at this measurement location from the environmental information DB 12a.

Then, the control unit 11 (identification unit) identifies the infection risk of the pathogen at this measurement location based on the read environmental information, usage information, and material information (S26). Specifically, the control unit 11 inputs the environmental information and material information read from the environmental information DB 12a to the infection risk determination model 12M from input nodes 0 and 2, and inputs the usage information read from the usage status DB 12b from input node 1. For example, input node 0 has multiple input nodes to which each piece of information included in the environmental information, such as temperature and humidity, is input, and pre-associated information is input to each input node. Similarly, input node 1 has multiple input nodes to which each piece of information included in the usage information, such as the number of users and the number of contacts, is input, and pre-associated information is input to each input node. Information (identification information) previously assigned to each material is input to input node 2.

The control unit 11 identifies the infection risk for this measurement location based on the output information from the infection risk determination model 12M. For example, the control unit 11 identifies the output node in the infection risk determination model 12M that outputs the maximum output value (discrimination probability), and identifies the infection risk associated with the identified output node as the infection risk for this measurement location. In this embodiment, the control unit 11 identifies the infection risk as either high risk, medium risk, or low risk. The control unit 11 stores the identified infection risk in the memory unit 12 in association with the location ID of the processing target.

The control unit 11 judges whether the processing for all the measurement locations for the facility to be judged has been completed (S27), and if it is judged that the processing has not been completed (S27: NO), the process returns to step S25. Then, the control unit 11 performs steps S25 to S26 for the unprocessed measurement locations to identify the infection risk for each location. If it is judged that the processing for all the measurement locations has been completed (S27: YES), the control unit 11 identifies the recommended behavior according to the infection risk identified for each location (S28). Information on the recommended behavior is stored in advance in the storage unit 12 according to the infection risk situation of each location, for example, when there is a high-risk location, when there is a medium-risk location, when there are a predetermined number or more high-risk locations, etc. Therefore, the control unit 11 identifies the recommended behavior to be performed to reduce the infection risk based on the infection risk identified for each location in the facility. In addition, the information on the recommended behavior may be set according to the infection risk and environmental information of each location. In this case, the control unit 11 can identify the recommended behavior according to the infection risk and environmental information of each location. The control unit 11 generates an infection risk determination screen that displays the calculated infection risk for each location (S29). The control unit 11 (output unit) then displays the generated infection risk screen on the display unit 15 (S30). This makes it possible to notify the user of the infection risk status within the facility.

9 shows an example of a judgment screen, and the screen shown in FIG. 9 displays the name of the facility to be judged for the infection risk and the test item (type of pathogen). In addition, a pull-down menu for selecting one of the selectable test items is provided in the display area of the test items, and the test item can be changed using the pull-down menu. In addition, the screen shown in FIG. 9 displays the infection risk (high, medium, or low) specified for each location, and a mark in a manner according to the infection risk is displayed at a position corresponding to each location on the floor map of the facility. In FIG. 9, each mark is hatched with a different color, but for example, each mark may be displayed in a different color, and any display manner may be used as long as each infection risk is displayed in a manner according to the infection risk. In addition, each number of the measurement locations 1, 2, 3, etc. corresponds to the number assigned to each mark on the floor map. Furthermore, the screen shown in FIG. 9 displays information on recommended actions to be taken according to the displayed infection risk situation. The judgment screen has a save button for instructing to store the judgment result of the infection risk being displayed in the memory unit 12, and a close button for instructing to end the process without saving the judgment result. When the save button on the screen shown in FIG. 9 is operated, the control unit 11 stores the infection risk assessment result being displayed, specifically, the floor map with marks according to the infection risk, in the memory unit 12, and when the close button is operated, the control unit 11 ends the display of the assessment screen.

By the above-mentioned process, the information processing device 10 of this embodiment can determine the infection risk at each location by considering the environmental information, usage status, and material information of each location. In this embodiment, the infection risk is determined using an infection risk determination model 12M that differs for each type of pathogen, so that the infection risk according to the characteristics of the pathogen can be obtained. That is, the infection risk according to the characteristics of each pathogen, such as strength against temperature, strength against humidity, strength against sunlight, strength against ultraviolet rays, strength against each material, and strength against each disinfectant, can be determined. In addition, when the measurement results (measurement amount) from the test using the test sample collected from each location in the facility are used, the infection risk according to the environmental information, usage status, and material information of each location in addition to the measurement results can be determined. Therefore, for example, even if the measurement results (abundance) of pathogens such as viruses and bacteria that cause infectious diseases are high, it cannot necessarily be said that the infection risk is high, but by using the above-mentioned information, the infection risk can be determined more accurately.

The infection risk judgment screen is not limited to the configuration shown in FIG. 9. FIG. 10 is a schematic diagram showing another example of the judgment screen. In addition to the configuration shown in FIG. 9, the judgment screen shown in FIG. 10 displays information on recommended actions to be taken for each measurement location according to the infection risk of each location. The information on the recommended actions here is stored in advance in the memory unit 12 according to, for example, the material and the infection risk, and the control unit 11 reads out the material information for each location from the environmental information DB 12a and identifies the recommended action based on the material information and the infection risk. Such a judgment screen can notify the action to be taken for each location, so that the cleaning staff can clean and disinfect each location in a manner according to the notified recommended action, and each location can be kept clean.

In this embodiment, the information processing device 10 may be configured as a server computer, and the process of determining the infection risk at each facility using the infection risk determination model 12M may be configured to be performed by the server computer. In this case, environmental information and usage information of each location, etc., can be input via an input device having an input unit and a display unit and transmitted to the server computer, and an infection risk determination screen generated by the server computer can be acquired and displayed. Even with such a configuration, the same processing as in this embodiment is possible, and the same effects can be obtained.

In addition, in this embodiment, the determination of the infection risk at each location in the facility is not limited to a configuration using the infection risk determination model 12M. For example, the infection risk may be determined for each type of pathogen (test item) using a table in which the infection risk corresponding to each combination of environmental information, usage information, and material information is registered in advance. In this case, the control unit 11 identifies a combination that matches the contents of the environmental information, usage information, and material information for each location from the combinations registered in the table, and identifies the infection risk corresponding to the identified combination as the infection risk for this location.

(Embodiment 2)
A process for generating the infection risk determination model 12M used in the information processing device 10 of the first embodiment will be described. The learning of the infection risk determination model 12M is performed, for example, by a predetermined learning device, and the learned infection risk determination model 12M generated by learning in the learning device is downloaded from the learning device to the information processing device 10 via a network or a portable storage medium 1a and stored in the storage unit 12. FIG. 11 is a block diagram showing an example of the configuration of the learning device. The learning device 30 is composed of a personal computer, a server computer, or the like, and includes a control unit 31, a storage unit 32, a communication unit 33, an input unit 34, a display unit 35, a reading unit 36, and the like, and these units are connected to each other via a bus. The units 31 to 36 of the learning device 30 are similar to the units 11 to 16 of the information processing device 10 of the first embodiment, so detailed description will be omitted. In addition to the control program 32P and the infection risk determination model 32M, the storage unit 32 of the learning device 30 stores a training data DB 32a in which training data used for learning the infection risk determination model 32M is stored. The training data DB 32a stores training data for each type of pathogen, in which input data including environmental information and usage information of each location at the time when the test sample was collected at each location, and information on the material of each location are associated with an infection risk (correct label) identified based on the pathogen test results for the test sample collected at each location. The infection risk of the correct label can be, for example, high risk, medium risk, or low risk. In addition, it is preferable that the infection risk of the correct label is identified taking into consideration not only the amount of pathogen present in the test sample, but also the presence or absence of infectivity of the pathogen.

Figure 12 is a flowchart showing an example of the generation process procedure of the infection risk determination model 32M. The control unit 31 of the learning device 30 acquires one piece of training data from the training data DB 32a (S41). The control unit 31 performs a learning process of the infection risk determination model 12M using the read training data (S42). Here, the control unit 31 inputs each piece of information of the input data included in the training data to the infection risk determination model 32M, and at this time, trains the infection risk determination model 32M so that the output value from the output node corresponding to the correct label included in the training data approaches 1 and the output values from the other output nodes approach 0. Note that in the learning process, the infection risk determination model 32M learns to optimize the coefficients of the functions used in the calculation process performed on the input values in the intermediate layer, as well as various weighting coefficients or thresholds.

The control unit 31 judges whether there is unprocessed training data in the training data DB 32a (S43), and if it is judged that there is unprocessed training data (S43: YES), the process returns to step S41. Then, the control unit 31 reads out the unprocessed training data from the training data DB 32a and repeats the learning process using the read out training data. If it is judged that there is no unprocessed training data (S43: NO), the control unit 31 ends the learning process. As a result, when various information (environmental information, usage information, and material information) for each pathogen for each location in the facility is input, an infection risk determination model 32M is generated that is trained to output a discrimination probability for each of high risk, medium risk, and low risk as the infection risk at each location. The control unit 31 trains the infection risk determination model 32M for each type of pathogen by performing the above-mentioned learning process for each type of pathogen. In addition, the infection risk determination model 32M that has already been trained can be retrained by performing the above-mentioned process, in which case it is possible to generate an infection risk determination model 32M with higher discrimination accuracy. In addition, the infection risk determination model 32M can be further optimized by repeatedly performing the learning process shown in FIG. 12 using the training data stored in the training data DB 32a.

By using the infection risk determination model 32M as described above, it is possible to accurately determine the infection risk at each location within the facility based on various information at each location. Furthermore, if the infection risk determination model 32M is trained using a huge amount of training data, it is possible to realize an infection risk determination model 12M that can determine the infection risk more accurately. By using the infection risk determination model 12M generated in this way, it is possible to determine the infection risk based on environmental information, usage information, and material information, for example, even in locations where testing samples are not being tested.

In this embodiment, the information input to the infection risk determination models 12M, 32M is not limited to environmental information, usage information, and material information for each location. For example, the infection risk determination models 12M, 32M may be configured to use only environmental information as input data. In this case, when environmental information for each location is input, the infection risk for each location can be determined using the infection risk determination models 12M, 32M that output the infection risk at this location. In addition, the infection risk determination models 12M, 32M may be configured to use environmental information and material information as input data, and may be configured to include test results (amount of pathogens present) for test samples collected at each location, information indicating the cleaning status of each location, etc. as input data.

(Embodiment 3)
An information processing device that determines the risk of pathogen infection based on past environmental information (time-series data) at each location in a facility will be described. The information processing device of this embodiment has a similar configuration to the information processing device 10 of the first embodiment, so a detailed description of the configuration will be omitted. Note that the infection risk determination model of this embodiment is slightly different from the configuration of the first embodiment shown in FIG.

FIG. 13 is a schematic diagram showing a configuration example of the infection risk determination model 12Ma of the third embodiment. The infection risk determination model 12Ma of the present embodiment is configured with a learning model such as LSTM or RNN. In the infection risk determination model 12Ma of the present embodiment, time series data of environmental information at the location to be determined is input from input node 0, time series data of usage information at the location to be determined is input from input node 1, and material information of the location to be determined is input from input node 2. The time series data of environmental information is past environmental information accumulated in the environmental information DB 12a, and may be, for example, the average temperature, average humidity, presence or absence of direct sunlight, hours of sunshine, average wind speed, weather, and season for each day. In addition, the environmental information may be the average temperature, average humidity, presence or absence of direct sunlight, hours of sunshine, average wind speed, weather, and season for each three hours. Furthermore, the maximum temperature or minimum temperature may be used instead of the average temperature, the maximum humidity or minimum humidity may be used instead of the average humidity, and the maximum wind speed or minimum wind speed may be used instead of the average wind speed. The time series data of the usage information is past usage information accumulated in the usage status DB 12b, and may be, for example, the number of users, the number of contacts, the average contact time, and the contact location per day. The usage information may also be the number of users, the number of contacts, the average contact time, and the contact location per three hours. In the infection risk determination model 12Ma of this embodiment, the intermediate layer also calculates output values using various functions and thresholds, etc., based on each piece of input information, and outputs the calculated output values from each output node of the output layer. As a result, the infection risk determination model 12Ma of this embodiment can determine the infection risk of a pathogen at this location based on the time series data of environmental information and the time series data of usage information related to the location to be determined. The infection risk determination model 12Ma of this embodiment has three output nodes 0 to 2, and outputs a discrimination probability for each of the three types of infection risk, high, medium, and low, from each output node.

The infection risk determination model 12Ma of this embodiment learns using training data including input data including time series data of environmental information of each location, time series data of usage information, and material information, and information (correct label) indicating the degree of infection risk (high, medium, low) identified based on the pathogen test results for the test sample collected at each location. Here, too, it is preferable that the infection risk of the correct label is identified taking into consideration not only the amount of pathogen present in the test sample, but also the presence or absence of infectiousness in the pathogen. In addition, the infection risk determination model 12Ma of this embodiment can also learn by the learning device 30 of the above-mentioned embodiment 2 executing a process similar to the process shown in FIG. 12. In the learning process, the infection risk determination model 12Ma learns that when each piece of information of the input data included in the training data is input, the output value from the output node corresponding to the correct label included in the training data approaches 1, and the output values from the other output nodes approach 0. In addition, the infection risk determination model 12Ma learns to optimize the coefficients of functions used in the calculation processing performed on the input values, as well as various weighting coefficients or thresholds. This results in a trained infection risk determination model 12Ma that has been trained to output information indicating the infection risk of pathogens when time-series data of environmental information for each location, time-series data of usage information, and material information are input.

The infection risk determination model 12Ma of this embodiment is not limited to the configuration shown in FIG. 13. For example, in addition to the environmental information, usage information, and material information, the model may be configured to input measurement results of the amount of pathogens detected from a test sample collected at the location to be determined, cleaning information indicating the cleaning status at the location to be determined, and the like. By using an infection risk determination model 12Ma configured in this way, it becomes possible to determine the infection risk of pathogens when each of these pieces of information is input.

The control unit 11 of the information processing device 10 of this embodiment executes the same process as that shown in FIG. 5. As a result, in this embodiment as well, environmental information such as temperature and humidity for each location in the facility can be registered in the environmental information DB 12a stored in the information processing device 10. In addition, the number of users counted using a human presence sensor, the number of contacts detected based on image data of users photographed using a camera, the average contact time, and the contact location are registered in the usage status DB 12b.

The control unit 11 of the information processing device 10 of this embodiment executes the same process as that shown in FIG. 7. As a result, in this embodiment as well, the infection risk at each location in the facility can be determined based on the environmental information, usage status, and material information of each location. In this embodiment, in step S25 in FIG. 7, the control unit 11 reads environmental information for a specific period in the past, including the date on which the infection risk is to be determined, from the environmental information DB 12a for one measurement location, and reads usage information for the same period at the same measurement location from the usage status DB 12b. Then, in step S26 in FIG. 7, the control unit 11 inputs the read environmental information for the specific period (time series data of environmental information for multiple days), usage information for the specific period (time series data of usage information), and material information to the infection risk determination model 12Ma read in step S24, and identifies the infection risk of the pathogen at the location to be determined based on the output information from the infection risk determination model 12Ma.

By performing the above-described processing, in this embodiment, the infection risk at each location within the facility can be determined using time series data of environmental information, time series data of usage information, and material information. In this embodiment, the infection risk is determined based on environmental information and usage information over a specified period of time in the past, so that the infection risk that fluctuates with changes in the environment or usage conditions can be determined with greater accuracy.

In this embodiment, the same effects as those of the above-mentioned embodiments can be obtained. Furthermore, in this embodiment, the infection risk is determined for each location in the facility using environmental information and usage information for multiple days. Therefore, by performing disinfection and cleaning work for each location in the facility according to the more accurately determined infection risk, it is possible to avoid an increase in the infection risk at each location. In this embodiment as well, the modified examples described as appropriate in each of the above-mentioned embodiments can be applied.

(Embodiment 4)
An information processing device that predicts the infection risk of pathogens based on prediction information of future environmental information at each location in a facility will be described. The information processing device of this embodiment has the same configuration as the information processing device 10 of the first embodiment, so detailed description of the configuration will be omitted. In this embodiment, the future infection risk of pathogens is predicted by inputting prediction information of future environmental information, prediction information of future usage information, and material information into the infection risk determination model 12M of the first embodiment shown in FIG. 4. For the prediction information of future environmental information, for example, each piece of meteorological information obtained from a website that publishes meteorological information via a network can be used, or it may be obtained in advance from the website and stored in the environmental information DB 12a. Specifically, average temperature, average humidity, presence or absence of direct sunlight, sunshine hours, average wind speed, weather, season, etc. can be extracted and used from the weather forecast (weather information) for a future day for which the infection risk is to be determined. Note that, when the infection risk is determined indoors, there may be no direct sunlight, and the sunshine hours may be a constant value set indoors. For the prediction information of future usage information, for example, each piece of information estimated from past usage conditions can be used, or it may be estimated in advance and stored in the usage condition DB 12b. Specifically, the usage status of a future day for which the infection risk is to be determined can be estimated and used based on the usage status (number of users, number of contacts, average contact time, contact location) of a past day that matches the season, weather, day of the week, time of day, etc. of the future day for which the infection risk is to be determined.

The infection risk determination model 12M of this embodiment is not limited to the configuration shown in FIG. 4. For example, in addition to predicted information on environmental information, predicted information on usage information, and material information, predicted information on cleaning information indicating the cleaning status at the location to be determined may be input. Note that the cleaning information may be, for example, predicted information on future cleaning status estimated from past cleaning status. By using an infection risk determination model 12M configured in this way, it is possible to predict the future risk of pathogen infection when each of these pieces of information is input.

The control unit 11 of the information processing device 10 of this embodiment executes a process similar to that shown in FIG. 5. As a result, in this embodiment as well, environmental information such as temperature and humidity for each location in the facility can be registered in the environmental information DB 12a stored in the information processing device 10. In addition, the number of users counted using a human presence sensor, the number of contacts detected based on image data of users captured using a camera, the average contact time, and the contact location are registered in the usage status DB 12b. In addition, in this embodiment, prediction information of future environmental information is registered in the environmental information DB 12a, and prediction information of future usage information is registered in the usage status DB 12b.

Figure 14 is a flowchart showing an example of the infection risk determination process procedure of the fourth embodiment, and Figure 15 is a schematic diagram showing an example screen. The process shown in Figure 14 is the process shown in Figure 7 with steps S61 to S62 added instead of step S26. The same steps as in Figure 7 will not be described. In the information processing device 10 of this embodiment, the control unit 11 performs the same process as steps S21 to S25 shown in Figure 7. In this embodiment, in step S25, the control unit 11 reads out, for example, prediction information on environmental information for multiple future days for which the infection risk is to be determined for one measurement location from the environmental information DB 12a, and reads out prediction information on usage information for the same period at the same measurement location from the usage status DB 12b. In this embodiment, the control unit 11 sequentially inputs the prediction information on environmental information and usage information for each future day read out as described above, and material information, to the infection risk determination model 12M. The control unit 11 then identifies (predicts) the infection risk for multiple future days at the location being assessed based on the output information sequentially output from the infection risk assessment model 12M (S61).

The control unit 11 identifies a change in the infection risk based on the infection risk for multiple days calculated by the above-mentioned process (S62). For example, the control unit 11 determines whether the infection risk level is increasing, maintaining, or decreasing for the infection risk for multiple days. The control unit 11 stores the identified change in the infection risk (increasing, maintaining, or decreasing) in the memory unit 12. Thereafter, the control unit 11 performs processing from step S27 onwards. In this embodiment, in step S29, the control unit 11 generates an infection risk judgment screen that displays the infection risk and the change in the infection risk calculated for each measurement location. FIG. 15 shows an example of the judgment screen of this embodiment, and the judgment screen shown in FIG. 15 displays a risk forecast that shows the change in the infection risk identified for each measurement location in addition to the configuration of embodiment 1 shown in FIG. 9. The change in the infection risk is indicated by an arrow as increasing, maintaining, or decreasing, and indicates a prediction of the change in the infection risk in the future (risk forecast). The judgment screen shown in FIG. 15 may display the predicted change in infection risk along with the predicted infection risk values for multiple days used to predict the change in infection risk. Through the above-described process, the information processing device 10 of this embodiment can estimate and notify the predicted change in future infection risk at each location in the facility.

In this embodiment, the same effects as those of the above-mentioned embodiments can be obtained. Furthermore, in this embodiment, the future infection risk at each location in the facility can be predicted using future predicted information on environmental information, future predicted information on usage information, and material information. In this embodiment, the user can grasp the future infection risk, and therefore can infer the risk of the spread of an infectious disease from the risk of infection with an infectious disease such as the influenza virus, making it possible to detect in advance the timing of the spread of the infectious disease, etc.

In this embodiment, the infection risk determination at each location in the facility is not limited to the configuration using the infection risk determination model 12M. For example, the infection risk may be determined using a table in which the infection risk corresponding to each combination of the predicted information of the environmental information, the predicted information of the usage information, and the material information is registered in advance for each type of pathogen (test item). In this case, the control unit 11 may identify a combination that matches the contents of the predicted information of the environmental information, the predicted information of the usage information, and the material information of each location from the combinations registered in the table, and identify the infection risk corresponding to the identified combination as the infection risk of this location. In addition, the infection risk determination model 12M of this embodiment may have a configuration similar to that of the infection risk determination model 12Ma of the third embodiment. In this case, the infection risk can be predicted based on the predicted information of the environmental information (time series data) and the predicted information of the usage information (time series data) over a predetermined future period. Therefore, the infection risk that changes with changes in the environment or the usage situation can be predicted more accurately, and the increase in the infection risk at each location can be suppressed by performing disinfection and cleaning work on each location in the facility according to the accurately predicted infection risk. In this embodiment as well, the modified examples described in each of the above embodiments can be applied as appropriate.

The embodiments disclosed herein are illustrative in all respects and should not be considered limiting. The scope of the present disclosure is indicated by the claims, not by the above meaning, and is intended to include all modifications within the scope and meaning equivalent to the claims.

REFERENCE SIGNS LIST 10 Information processing device 11 Control unit 12 Memory unit 15 Display unit 30 Learning device 31 Control unit 32 Memory unit 12M Infection risk determination model 12Ma Infection risk determination model 32M Infection risk determination model

Claims (14)

Acquire environmental information, including temperature, humidity, and sunlight, at each location in the facility that people touch with their hands ,
obtaining a measure of the pathogen at each of said locations;
Identifying an infection risk of the pathogen at each of the locations based on the acquired environmental information and the measured amount of the pathogen ;
A program that causes a computer to execute a process to output the identified infection risk for each location. Obtaining material information indicating a material of each of the locations;
Identifying an infection risk of the pathogen in each of the locations based on the acquired environmental information, the measured amount of the pathogen, and the material information.
The program according to claim 1 , which causes the computer to execute a process. Obtain cleaning information including whether each location is capable of disinfection, the type of disinfectant that can be used, or the frequency at which disinfection should be performed;
Identifying a risk of infection of the pathogen in each of the locations based on the acquired environmental information, the measured amount of the pathogen, and the cleaning information.
3. The program according to claim 1, which causes the computer to execute a process. Acquire usage information regarding the usage status of users at each of the locations;
The program according to claim 1 , further comprising: causing the computer to execute a process of identifying an infection risk of the pathogen in each of the locations based on the acquired environmental information, the measured amount of the pathogen, and the usage information . A program as described in any one of claims 1 to 4, which causes the computer to execute a process of inputting the acquired environmental information and the measured amount of the pathogen into a trained model trained to output information related to the infection risk of the pathogen at the location when the environmental information and the measured amount of the pathogen at the location are input, and determining the infection risk of the pathogen at the location. The trained model is trained to output information related to a pathogen infection risk at the location when time series data of environmental information for multiple days in the past and a measured amount of a pathogen are input,
The program according to claim 5, which causes the computer to execute a process of inputting the time series data of the environmental information and the measured amount of the pathogen into the trained model to determine a pathogen infection risk in the location. The program according to any one of claims 1 to 6, which causes the computer to execute a process of: identifying recommended actions to be taken in each of the locations based on the acquired environmental information and the identified infection risk; and outputting the identified recommended actions for each location. Obtaining prediction information of the environmental information at each of the locations;
determining a predicted value of a pathogen infection risk at each of the locations based on the obtained prediction information;
The program according to claim 1 , further comprising: predicting a change in an infection risk based on the identified infection risk and a predicted value of the infection risk. The program according to claim 1 , further comprising: displaying an infection risk at each of the locations in a manner corresponding to the infection risk. The program according to claim 1 , which causes the computer to execute a process of displaying the infection risk at each of the locations on a floor map of the facility. An acquisition unit that acquires environmental information including temperature, humidity, or sunlight at each location in the facility that is touched by people with their hands ;
a measurement acquisition unit for acquiring a measurement amount of a pathogen at each of the locations;
an identification unit that identifies an infection risk of the pathogen in each of the locations based on the acquired environmental information and the measured amount of the pathogen ;
and an output unit that outputs the identified infection risk for each location. Acquire environmental information, including temperature, humidity, and sunlight, at each location in the facility that people touch with their hands ,
obtaining a measure of the pathogen at each of said locations;
Identifying an infection risk of the pathogen at each of the locations based on the acquired environmental information and the measured amount of the pathogen ;
An information processing method in which a computer executes a process of outputting the identified infection risk for each location. Acquire training data including environmental information including temperature, humidity, or sunlight at each location in the facility that is touched by people with their hands , a measured amount of a pathogen at each of the locations, and an infection risk of the pathogen at each of the locations;
A method for generating a learning model in which a computer executes a process of generating a learning model using the acquired training data, which outputs information related to the risk of infection of a pathogen in a location within a facility that people touch with their hands when environmental information and measured amounts of the pathogen are input. Obtain training data including an infection risk identified based on the measured amount of the pathogen contained in the test sample collected at each of the locations, and environmental information and the measured amount of the pathogen at the time of collection of the test sample;
The method for generating a learning model according to claim 13 , wherein the computer executes a process of generating the learning model using the training data.

Images