JP7186207B2 - Information processing device, information processing program and information processing system - Google Patents

Description

The present invention relates to an information processing device, an information processing program, and an information processing system.

Conventionally, techniques for providing comfortable spaces for users are known. For example, the degree of concentration of the user in the work space is evaluated based on an image of the user captured by a camera. A technique is known that changes the air environment in the work space at the timing when the degree of concentration of the user is lowered.

Japanese Patent Application Laid-Open No. 2020-20571

However, in the above-described conventional technology, the air environment in the work space is only changed at the timing when the concentration level of the user decreases, so it is not always possible to improve the productivity of the group in the space.

The information processing apparatus according to the embodiment is based on the work productivity of each person estimated based on the amount of work of each person in a space where a plurality of people are present and air conditioning is performed by an air conditioner. an air conditioning control model for outputting air conditioning control information for controlling air conditioning by said air conditioning equipment; a learning unit that performs reinforcement learning based on the air-conditioning control information when the air-conditioning is performed by and a value indicating the reward for the group.

FIG. 1 is a diagram illustrating a configuration example of an information processing system according to an embodiment. FIG. 2 is a diagram illustrating a configuration example of an information processing apparatus according to the embodiment; FIG. 3 is a diagram for explaining an example of an individual person's productivity estimation process according to the embodiment. FIG. 4 is a diagram for explaining information about workload according to the embodiment. FIG. 5 is a diagram for explaining an example of a remuneration table for determining individual remunerations according to the embodiment. FIG. 6 is a diagram showing an information processing procedure according to the embodiment. FIG. 7 is a diagram illustrating an information processing procedure according to the embodiment; FIG. 8 is a diagram illustrating an information processing procedure according to the embodiment; FIG. 9 is a diagram illustrating an information processing procedure according to the embodiment; FIG. 10 is a diagram illustrating an information processing procedure according to the embodiment; FIG. 11 is a diagram illustrating an information processing procedure according to the embodiment; FIG. 12 is a diagram illustrating an information processing procedure according to the embodiment; FIG. 13 is a diagram for explaining an example of a remuneration table for determining individual remunerations according to the modification. FIG. 14 is a hardware configuration diagram showing an example of a computer that implements the functions of the information processing apparatus.

Hereinafter, modes for implementing an information processing apparatus, an information processing program, and an information processing system according to the present application (hereinafter referred to as "embodiments") will be described in detail with reference to the drawings. Note that the information processing apparatus, information processing program, and information processing system according to the present application are not limited to this embodiment. Also, in each of the following embodiments, the same parts are denoted by the same reference numerals, and overlapping descriptions are omitted.

(embodiment)
[1. Introduction]
2. Description of the Related Art Conventionally, there has been known a technique for controlling air conditioning in order to improve work productivity of people in a space. However, by only controlling the air conditioning in consideration of the productivity of each individual person present in the space, it is possible to improve the productivity of all the plural persons present in the space (hereinafter also referred to as collective productivity). Not necessarily.

On the other hand, the information processing apparatus 100 according to the embodiment estimates the work volume of each person in a space where a plurality of people are present and is air-conditioned by an air conditioner. Based on gender, a value is determined that indicates the collective reward for air conditioning for multiple persons located in the space. Further, the information processing apparatus 100 creates an air-conditioning control model that outputs air-conditioning control information for controlling air-conditioning by the air-conditioning equipment based on the air-conditioning control information when the air-conditioning is performed by the air-conditioning equipment and a value indicating the reward of the group. Reinforcement learning. Accordingly, the information processing apparatus 100 can control air conditioning using an air conditioning control model that outputs an air conditioning operation that increases the productivity of a group located in the space. Therefore, the information processing device 100 can improve the productivity of a group located in the space.

[2. Configuration of information processing system]
First, the configuration of the information processing system according to the embodiment will be described using FIG. FIG. 1 is a diagram illustrating a configuration example of an information processing system according to an embodiment. The information processing system 1 may include an air conditioner 10 , a sensor device 20 , and an information processing device 100 . The air conditioner 10, the sensor device 20, and the information processing device 100 may be communicatively connected via a predetermined network N by wire or wirelessly. The information processing system 1 may include an arbitrary number of air conditioners 10 , an arbitrary number of sensor devices 20 , and an arbitrary number of information processing devices 100 .

The air conditioner 10 is a device that adjusts the temperature and humidity of the air in the space. For example, the air conditioner 10 may be an air conditioner (for example, an air conditioner) installed in each space subject to air conditioning control by the information processing device 100 . Moreover, the air conditioner 10 may be able to communicate with other devices by IoT (Internet of Things) technology. The air conditioner 10 can air-condition a space to be controlled under the control of the information processing device 100 . Specifically, the air conditioner 10 can receive air conditioning control information for controlling air conditioning by air conditioners installed in the space from the information processing device 100 . For example, the air conditioner 10 may receive, as the air conditioning control information, cooling information indicating that air conditioning is to be performed according to the cooling setting or heating information indicating that air conditioning is to be performed according to the heating setting. Further, the air conditioner 10 may receive temperature information indicating the set temperature of the space and humidity information indicating the set humidity of the space as the air conditioning control information. Subsequently, the air conditioner 10 air-conditions the space according to the air-conditioning control information received from the information processing device 100 . For example, upon receiving the cooling information (heating information), the air conditioner 10 may adjust the temperature and humidity of the space according to the cooling setting (heating setting). Further, when the temperature information is received, the air conditioner 10 may adjust the temperature of the space so that it is kept at the set temperature or the set humidity. Further, upon receiving the humidity information, the air conditioner 10 may adjust the humidity of the space so that the set humidity is maintained. In addition, below, the air conditioner 10 may be described as an "air conditioner."

Incidentally, as shown in FIG. 1, depending on the space in which the air conditioner 10 is installed, the air conditioner 10 may be distinguished and described as air conditioners 10-1 to 10-2. For example, the air conditioner 10-1 may be the air conditioner 10 installed in the conference room R1. Further, for example, the air conditioner 10-2 may be the air conditioner 10 installed in the conference room R2. Further, hereinafter, the air conditioners 10-1 and 10-2 will be referred to as the air conditioner 10 when they are not distinguished from each other.

The sensor device 20 is a device that detects the physical state of a space and the physical state of a user present in the space. The sensor device 20 may acquire, as sensor information, information indicating the physical state of the space and the physical state of users present in the space. For example, sensor device 20 may be an image sensor. For example, sensor device 20 may be an RGB camera. That is, the sensor device 20 may acquire, as sensor information, an RGB image including, as a subject, a user existing in space, for example. Also, for example, the sensor device 20 may be an infrared camera. That is, the sensor device 20 may acquire, as sensor information, a thermal image including, as a subject, a user existing in space, for example. Also, the sensor device 20 may be an environmental sensor. For example, sensor device 20 may be a USB-type environmental sensor. That is, the sensor device 20 may acquire, for example, information about the environment such as the temperature and humidity of the space (hereinafter also referred to as environment information) as sensor information. Further, the sensor device 20 may transmit the acquired sensor information to the information processing device 100 . Moreover, the environmental information acquired by the sensor device 20 is not limited to the above example, as long as it is information that can be used in determining comfort, which will be described later. For example, the sensor device 20 may acquire arbitrary information such as atmospheric pressure and illuminance as environmental information.

As shown in FIG. 1, depending on the space in which the sensor device 20 is installed, the sensor device 20 may be distinguished and described as sensor devices 20-1 to 20-2. For example, the sensor device 20-1 is the sensor device 20 installed in the conference room R1. Further, for example, the sensor device 20-2 is the sensor device 20 installed in the conference room R2. Further, hereinafter, the sensor devices 20-1 and 20-2 will be referred to as the sensor device 20 when they are described without particular distinction.

The terminal device 30 is an information processing device used by a user present in the space. The terminal device 30 is, for example, a smart phone, a tablet terminal, a notebook PC (Personal Computer), a desktop PC, a mobile phone, a PDA (Personal Digital Assistant), a head-mounted display, or the like. In the embodiment, an example in which the terminal device 30 is a notebook PC or a desktop PC (hereinafter sometimes simply referred to as PC) will be described.

In addition, the terminal device 30 is equipped with or connected to various sensors for detecting the physical state of the user. For example, the terminal device 30 is connected to sensors such as a camera and a microphone. The terminal device 30 may detect human sensor information indicating the physical state of the user by various sensors. For example, the terminal device 30 may detect an image of the user using a camera as an example of person sensor information. In addition, the terminal device 30 may detect the user's voice with a microphone as an example of person sensor information. In addition, the terminal device 30 may detect, as an example of person sensor information, operation log information regarding the operation history of the user on the PC. Upon detecting the person sensor information, the terminal device 30 may transmit the detected person sensor information to the information processing device 100 .

Note that, as shown in FIG. 1, the terminal devices 30 may be described separately as terminal devices 30-1 to 30-2 depending on the user who uses the terminal device 30. FIG. For example, the terminal device 30-1 is the terminal device 30 used by the user U1 identified by the user ID "U1". Also, for example, the terminal device 30-2 is the terminal device 30 used by the user U2 identified by the user ID "U2". In addition, hereinafter, the terminal devices 30-1 and 30-2 will be referred to as the terminal device 30 when they are not distinguished from each other. In addition, below, a user may be described as a "person."

The information processing apparatus 100 is based on the work productivity of each person estimated based on the amount of work of each person in a space where a plurality of people are present and is air-conditioned by an air conditioner. determine a value that indicates the collective reward for air conditioning for multiple persons who Further, the information processing apparatus 100 creates an air-conditioning control model that outputs air-conditioning control information for controlling air-conditioning by the air-conditioning equipment based on the air-conditioning control information when the air-conditioning is performed by the air-conditioning equipment and a value indicating the reward of the group. Reinforcement learning.

[3. Configuration of Information Processing Device]
Next, the configuration of the information processing apparatus 100 according to the embodiment will be described using FIG. FIG. 2 is a diagram illustrating a configuration example of the information processing apparatus 100 according to the embodiment. The information processing device 100 may have a communication section 110 , a storage section 120 and a control section 130 . The information processing apparatus 100 includes an input unit (for example, a keyboard, a mouse, etc.) for receiving various operations from an administrator of the information processing apparatus 100, and a display unit (for example, a liquid crystal display) for displaying various information. may have.

(Communication unit 110)
The communication unit 110 is realized by, for example, a NIC (Network Interface Card) or the like. The communication unit 110 is connected to a network by wire or wirelessly, and transmits and receives information to and from the air conditioner 10 and the sensor device 20, for example.

(storage unit 120)
The storage unit 120 is realized by, for example, a semiconductor memory device such as a RAM (Random Access Memory) or flash memory, or a storage device such as a hard disk or an optical disk.

The storage unit 120 may store sensor information detected by the sensor device 20 . For example, the storage unit 120 may store, as an example of sensor information, an RGB image and a thermal image of the space in which the user is located in association with identification information for identifying the space. In addition, the storage unit 120 may store, as an example of sensor information, environment information regarding the temperature and humidity of the space where the user is located in association with identification information for identifying the space.

The storage unit 120 may also store information on various learning models. For example, the storage unit 120 may store information on the productivity discrimination model. In addition, the storage unit 120 may store information regarding air conditioning control models. For example, the storage unit 120 may store information about connection coefficients of various learning models.

The storage unit 120 may also store person sensor information detected by the terminal device 30 . For example, the storage unit 120 may store, as an example of person sensor information, an image of the user detected by the camera mounted on the terminal device 30 and identification information for identifying the user in association with each other. Further, the storage unit 120 may store, as an example of human sensor information, the user's voice information detected by a microphone mounted on the terminal device 30 and identification information for identifying the user in association with each other. In addition, the storage unit 120 may store, as an example of the person sensor information, the operation log information regarding the operation history of the PC by the user detected by the terminal device 30 and the identification information for identifying the user in association with each other.

(control unit 130)
The control unit 130 is a controller, and for example, various programs (information processing programs) stored in a storage device inside the information processing apparatus 100 are controlled by a CPU (Central Processing Unit), an MPU (Micro Processing Unit), or the like. (equivalent to one example) is implemented by executing the RAM as a work area. Also, the control unit 130 is a controller, and is implemented by an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array).

The control unit 130 has, as functional units, an acquisition unit 131, a generation unit 132, an estimation unit 133, an air conditioning control unit 134, a determination unit 135, and a learning unit 136, and performs information processing operations described below. may be realized or performed. Note that the internal configuration of the control unit 130 is not limited to the configuration shown in FIG. 2, and may be another configuration as long as it performs the information processing described later. Moreover, each functional unit indicates the function of the control unit 130 and may not necessarily be physically distinguished.

(Acquisition unit 131)
The acquisition unit 131 may acquire information about the workload of each person located in a space where a plurality of people are located. Specifically, the acquisition unit 131 may acquire person sensor information of each person from the terminal device 30 . Subsequently, when acquiring the person sensor information of each person, the acquiring unit 131 may calculate information about the amount of work of each person based on the acquired person sensor information of each person.

Further, when acquiring the person sensor information of each person, the acquiring unit 131 may store the acquired person sensor information and the identification information that can identify each person in the storage unit 120 in association with each other. For example, the acquisition unit 131 obtains the person sensor information of each person in the space before air conditioning is performed by the air conditioning equipment (hereinafter sometimes referred to as “before air conditioning”) and the identification that can identify each person. The information may be stored in the storage unit 120 in association with identification information that identifies the space before air conditioning. In addition, the acquisition unit 131 can identify the person sensor information of each person in the space after the air conditioning is performed by the air conditioning equipment (hereinafter sometimes referred to as “after air conditioning”) and the individual person. The identification information and the identification information for identifying the air-conditioned space may be stored in the storage unit 120 in association with each other. That is, the acquisition unit 131 may store the person sensor information of each person acquired before and after the air conditioning in the storage unit 120 by distinguishing between them.

Here, the information about the amount of work done by each person means the information that quantifies the amount of work done by each person. Information about the amount of work will be described in detail with reference to FIG. FIG. 4 is a diagram for explaining information about workload. The information on the amount of work shown in FIG. 4 can be used as an index for evaluating the work productivity of each person. In FIG. 4, PC (information equipment) operation information, voice information, and human activity information will be described as information related to workload.

In the example shown on the left side of FIG. 4, the acquisition unit 131 obtains, as an example of information related to the amount of work of each person, PC operation information related to the amount of operation of the information device by each person (hereinafter referred to as PC operation information of each person). ) may be obtained. For example, the acquisition unit 131 acquires, as an example of the person sensor information of each person, operation log information (hereinafter sometimes referred to as operation log information of each person) regarding the operation history of PCs by each person. It may be obtained from device 30 . Subsequently, after acquiring the operation log information of each person, the acquiring unit 131 may calculate the PC operation information of each person based on the acquired operation log information of each person. For example, the acquisition unit 131 obtains, as an example of the PC operation information of each person, the operation amount of keyboard input, the operation amount of mouse operation, the length of the PC startup time, the amount of power consumption, the number of printer outputs, and communication Information about quantity may be calculated.

Further, after calculating the PC operation information of each person, the acquisition unit 131 may associate the calculated PC operation information of each person with identification information that can identify each person, and store the information in the storage unit 120. . For example, the acquisition unit 131 associates the PC operation information of each person in the space before air conditioning, the identification information that can identify each person, and the identification information that identifies the space before air conditioning, and stores the information in the storage unit 120. can be stored. In addition, the acquisition unit 131 stores the PC operation information of each person in the air-conditioned space, the identification information that can identify each person, and the identification information that identifies the space after the air-conditioning in association with each other, and stores the information in the storage unit 120. can be stored. That is, the acquisition unit 131 may store in the storage unit 120 the PC operation information of each person acquired before and after the air conditioning, separately.

In the example shown in the center of FIG. 4 , the acquisition unit 131 obtains, as an example of information related to the amount of work of each person, voice information related to the amount of speech of each person (hereinafter, may be referred to as voice information of each person). available). For example, the acquiring unit 131 may acquire voice information detected by a microphone mounted on the terminal device 30 from the terminal device 30 as an example of person sensor information of each person. Subsequently, after acquiring the voice information, the acquiring unit 131 may acquire the voice information of each person based on the acquired voice information. For example, if the acquired voice information includes voices of multiple speakers, the acquisition unit 131 may identify the voices of each speaker using a known multi-speaker identification technology. Subsequently, the acquisition unit 131 may identify each person's voice and then remove noise included in each person's voice information using a known sound source separation (noise removal) technique. Subsequently, the acquisition unit 131 may acquire the voice information of each person from which noise has been removed as the voice information of each person.

Further, when acquiring the voice information of each person, the acquiring unit 131 may associate the acquired voice information of each person with identification information that can identify each person, and store them in the storage unit 120 . For example, the acquisition unit 131 associates the voice information of each person present in the space before air conditioning, the identification information that can identify each person, and the identification information that identifies the space before air conditioning, and stores them in the storage unit 120. You can In addition, the acquisition unit 131 associates the voice information of each person present in the air-conditioned space with the identification information that can identify each person and the identification information that identifies the space after the air-conditioning, and stores them in the storage unit 120 . You can In other words, the acquisition unit 131 may store the voice information of each individual acquired before and after the air conditioning in the storage unit 120 by distinguishing between them.

In the example shown on the right side of FIG. 4, the acquisition unit 131 obtains human activity information related to the amount of movement of each person (hereinafter referred to as human activity information of each person) as an example of information related to the amount of work of each person. may be obtained). For example, the acquiring unit 131 may acquire from the terminal device 30 an image detected by a camera mounted on the terminal device 30 as an example of person sensor information of each person. Subsequently, after acquiring the image, the acquiring unit 131 may acquire the person activity information of each person based on the acquired image. For example, when acquiring an image, the acquiring unit 131 identifies an area in which each person included in the image is captured (hereinafter also referred to as an individual person area). For example, the acquisition unit 131 may identify each person region included in the image based on posture estimation processing for estimating the posture of each person. Subsequently, after identifying each person area included in the image, the acquiring unit 131 acquires information about each person area included in the image as person activity information of each person.

Further, when acquiring the person activity information of each person, the acquiring unit 131 may associate the acquired person activity information of each person with identification information that can identify each person, and store the information in the storage unit 120. . For example, the acquisition unit 131 associates the person activity information of each person located in the space before air conditioning, the identification information that can identify each person, and the identification information that identifies the space before air conditioning, and stores the information in the storage unit 120. can be stored. In addition, the acquisition unit 131 associates the person activity information of each person present in the air-conditioned space, the identification information that can identify each person, and the identification information that identifies the space after the air-conditioning, and stores the information in the storage unit 120. can be stored. In other words, the acquiring unit 131 may store in the storage unit 120 the individual person's activity information acquired before and after the air conditioning, separately.

Further, the acquisition unit 131 may acquire from the sensor device 20 sensor information indicating the physical state of the space where the person is located and the physical state of the person located in the space. Specifically, the acquiring unit 131 may acquire, as an example of sensor information, an RGB image of a space in which a plurality of people are located from an RGB camera installed in the space in which the plurality of people are located. Subsequently, when acquiring the RGB image, the acquiring unit 131 may associate the acquired RGB image with the identification information for identifying the space, and store them in the storage unit 120 . For example, the acquisition unit 131 may associate an RGB image of the space before air conditioning with identification information for identifying the space before air conditioning, and store the information in the storage unit 120 . Further, the obtaining unit 131 may store the RGB image of the air-conditioned space in the storage unit 120 in association with the identification information for identifying the air-conditioned space. That is, the acquisition unit 131 may distinguish between RGB images acquired before and after air conditioning and store them in the storage unit 120 .

As an example of sensor information, the acquisition unit 131 may acquire a thermal image of a space in which a plurality of people are located from an infrared camera installed in the space in which a plurality of people are located. Subsequently, when acquiring the thermal image, the acquiring unit 131 may store the acquired thermal image and the identification information for identifying the space in the storage unit 120 in association with each other. For example, the acquisition unit 131 may associate a thermal image obtained by capturing the space before air conditioning with identification information for identifying the space before air conditioning, and store the information in the storage unit 120 . Further, the acquisition unit 131 may associate a thermal image obtained by capturing the air-conditioned space with identification information for identifying the air-conditioned space and store the information in the storage unit 120 . That is, the acquisition unit 131 may store the thermal images acquired before and after the air conditioning in the storage unit 120 by distinguishing between them.

As an example of sensor information, the acquisition unit 131 may acquire environmental information about the temperature and humidity of the space where the plurality of people are located from an environment sensor installed in the space where the plurality of people are located. Subsequently, the acquiring unit 131 may associate the acquired environment information regarding the temperature and humidity with the identification information for identifying the space, and store the information in the storage unit 120 . For example, the acquisition unit 131 may associate environmental information regarding the temperature and humidity of the space before air conditioning with identification information for identifying the space before air conditioning, and store the information in the storage unit 120 . In addition, the acquisition unit 131 may associate environmental information regarding the temperature and humidity of the air-conditioned space with identification information for identifying the air-conditioned space and store the associated information in the storage unit 120 . That is, the acquisition unit 131 may store the environment information acquired before and after the air conditioning in the storage unit 120 by distinguishing between them.

In addition, the acquisition unit 131 may acquire information about various machine learning models (hereinafter sometimes simply referred to as “learning models”) from an external information processing device. Specifically, as an example of a learning model, the acquiring unit 131 outputs a feature vector V1 indicating a feature of the PC operation information of each person when PC operation information of each person is input as input information. You may acquire the learning model M1 output as . For example, the acquisition unit 131 may acquire a learning model M1 that has been trained to estimate a feature vector indicating the characteristics of each person's PC operation information from each person's PC operation information. For example, the acquisition unit 131 may acquire the learning model M1, which is a neural network. For example, when PC operation information of an individual person is input as input information, the acquisition unit 131 is a learning model M1 that outputs a 2048-dimensional feature vector V1 indicating the characteristics of the PC operation information of an individual person as output information. can be obtained. Subsequently, when acquiring the learning model M1, the acquiring unit 131 may associate the information about the learning model M1 with the identification information “M1” for identifying the learning model M1 and store them in the storage unit 120 .

In addition, as an example of a learning model, the acquisition unit 131 outputs, as output information, a feature vector V2 indicating a feature of the voice information of each person when voice information of each person is input as input information. You may get M2. For example, the acquisition unit 131 may acquire a learning model M2 that has been trained to estimate a feature vector representing a feature of the individual person's audio information from the individual person's audio information. For example, the acquisition unit 131 may acquire the learning model M2, which is a neural network. For example, the acquisition unit 131 acquires a learning model M2 that outputs, as output information, a 2048-dimensional feature vector V2 indicating the characteristics of the voice information of each person when the voice information of each person is input as input information. You can Subsequently, when acquiring the learning model M2, the acquiring unit 131 may associate the information about the learning model M2 with the identification information “M2” for identifying the learning model M2 and store them in the storage unit 120 .

In addition, as an example of a learning model, the acquisition unit 131 outputs a feature vector V3 indicating the characteristics of the individual person's activity information as output information when the individual person's activity information is input as input information. A learning model M3 may be obtained. For example, the acquisition unit 131 uses a learning model M3 trained to estimate object names (classes) in images from a large number of images (for example, over 10 million images) registered in an image database such as ImageNet. can be obtained. For example, the acquisition unit 131 may acquire the learning model M3, which is a convolutional neural network. For example, the acquisition unit 131 may acquire the learning model M3, which is Resnet50. For example, when personal activity information of each person is input as input information, the acquisition unit 131 is a learning model M3 that outputs, as output information, a 2048-dimensional feature vector V3 indicating the characteristics of the individual person's activity information. can be obtained. Subsequently, when acquiring the learning model M3, the acquiring unit 131 may associate the information about the learning model M3 with the identification information “M3” for identifying the learning model M3 and store them in the storage unit 120 .

Further, the acquiring unit 131 may acquire, as an example of a learning model, a learning model M5 that outputs a feature vector V5 representing a feature of an RGB image as output information when an RGB image is input as input information. For example, the acquisition unit 131 uses a learning model M5 trained to estimate object names (classes) in images from a large number of images (for example, over 10 million images) registered in an image database such as ImageNet. can be obtained. For example, the acquisition unit 131 may acquire the learning model M5, which is a convolutional neural network. For example, the acquisition unit 131 may acquire the learning model M5, which is Resnet50. For example, when an RGB image is input as input information, the acquisition unit 131 may acquire a learning model M5 that outputs a 2048-dimensional feature vector V5 representing the characteristics of the RGB image as output information. Subsequently, when acquiring the learning model M5, the acquiring unit 131 may associate the information about the learning model M5 with the identification information “M5” for identifying the learning model M5 and store them in the storage unit 120 .

Further, the acquisition unit 131 may acquire, as an example of a learning model, a learning model M6 that outputs a feature vector V6 representing a feature of a thermal image as output information when a thermal image is input as input information. For example, the acquisition unit 131 is trained to estimate object names (classes) appearing in thermal images from a large number (for example, over 10 million) of thermal images registered in an image database such as ImageNet. A model M6 may be obtained. For example, the acquisition unit 131 may acquire the learning model M6, which is a convolutional neural network. For example, the acquisition unit 131 may acquire the learning model M6, which is Resnet50. For example, when a thermal image is input as input information, the acquisition unit 131 may acquire a learning model M6 that outputs a 2048-dimensional feature vector V6 representing the characteristics of the thermal image as output information. Subsequently, when acquiring the learning model M6, the acquiring unit 131 may associate the information about the learning model M6 with the identification information “M6” for identifying the learning model M6 and store them in the storage unit 120 .

In addition, as an example of a learning model, when environmental information about the temperature and humidity of a space is input as input information, the acquisition unit 131 outputs a feature vector V8 indicating the characteristics of the environmental information about the temperature and humidity of the space as output information. A learning model M8 to be output may be obtained. For example, the acquisition unit 131 may acquire a learning model M8 that has been trained to estimate a feature vector that indicates the characteristics of information about the temperature and humidity of the space from environmental information about the temperature and humidity of the space. For example, the acquisition unit 131 may acquire the learning model M8, which is a neural network. For example, the acquisition unit 131 acquires a learning model M8 that outputs, as output information, a 2048-dimensional feature vector V8 indicating the characteristics of the environmental information regarding temperature and humidity when environmental information regarding temperature and humidity is input as input information. You can Subsequently, when acquiring the learning model M8, the acquisition unit 131 may associate the information about the learning model M8 with the identification information “M8” for identifying the learning model M8 and store them in the storage unit 120 .

(Generating unit 132)
The generation unit 132 calculates the work productivity of each person (hereinafter, "work productivity" may be referred to as "productivity" based on information about the amount of work of each person in the space. ) may be generated. Specifically, the generation unit 132 receives, as input information, a feature vector V4 indicating the characteristics of the information on the workload of each person in the space, as an example of the information on the workload of each person in the space. A productivity determination model M4 may be generated that outputs the work productivity of each individual person as output information when the information is given. For example, when the feature vector V4 is input as input information, the generation unit 132 is a productivity discrimination model trained to output the probability that the feature vector V4 is classified into each category C1 indicating productivity as output information. M4 may be generated. For example, when the feature vector V4 is input as input information, the generating unit 132 determines the probability of being classified into the “high” category indicating high productivity and the “normal” category indicating normal productivity. and the probability of being classified into the "low" category indicating low productivity as output information.

More specifically, the generating unit 132 generates a feature vector V4 indicating the characteristics of the information about the workload of each person in the space as an example of the information about the workload of each person in the space. good. For example, the generation unit 132 may refer to the storage unit 120 to acquire the PC operation information of each person and the information on the learning model M1 acquired by the acquisition unit 131 . Next, when the generation unit 132 acquires the PC operation information of each person and the information about the learning model M1, the generation unit 132 uses the learning model M1 to generate a feature vector V1 indicating the characteristics of the PC operation information of each person. good. In addition, the generation unit 132 may refer to the storage unit 120 to acquire the voice information of each person and the information related to the learning model M2 acquired by the acquisition unit 131 . Subsequently, when acquiring the voice information of each person and the information about the learning model M2, the generation unit 132 may use the learning model M2 to generate a feature vector V2 indicating the features of the voice information of each person. In addition, the generation unit 132 may refer to the storage unit 120 to acquire the person activity information of each person acquired by the acquisition unit 131 and the information related to the learning model M3. Next, when the generation unit 132 acquires the person activity information of each person and the information about the learning model M3, the generation unit 132 uses the learning model M3 to generate a feature vector V3 indicating the characteristics of the person activity information of each person. good.

After generating the feature vectors V1 to V3, the generator 132 may then generate a feature map MP1 indicating the features of the feature vectors V1 to V3 based on the generated feature vectors V1 to V3. Subsequently, after generating the feature map MP1, the generator 132 may generate a feature vector V4 representing the features of the feature map MP1 from the feature map MP1 by performing maximum value pooling on the feature map MP1. For example, the generation unit 132 may generate a 2048-dimensional feature vector V4 indicating the features of the feature map MP1 by maximum value pooling. In this way, the generation unit 132 may generate the feature vector V4 that indicates the feature of the information regarding the amount of work of each person located in the space.

Subsequently, when the feature vector V4 is generated, the generating unit 132 generates the productivity discrimination model M4 so that the work productivity of each person is output as output information when the generated feature vector V4 is input as input information. to learn For example, the generation unit 132 learns the productivity discrimination model M4 so that when the feature vector V4 is input as input information, the probability that the feature vector V4 is classified into each category C1 indicating productivity is output as output information. Let For example, when the feature vector V4 is input as input information, the generating unit 132 determines the probability of being classified into the “high” category indicating high productivity and the “normal” category indicating normal productivity. The productivity discrimination model M4 is trained to output as output information the probability of being classified into 1 and the probability of being classified into the "low" category indicating that the productivity is low.

Note that the category C1 indicating productivity may be any classification, and is not limited to the above example. For example, the productivity classification may include other classifications such as "somewhat high" and "somewhat low." Further, the generating unit 132 generates a productivity discrimination model M4 that has been trained to output, as output information, the label of the productivity category C1 into which the feature vector V4 is classified when the feature vector V4 is input as input information. may be generated. For example, when the feature vector V4 is input as input information, the generation unit 132 learns to output the label of the category with the highest probability that the feature vector V4 is classified into each category C1 indicating productivity as the output information. A productivity discrimination model M4 may be generated. Subsequently, when generating the productivity discrimination model M4, the generation unit 132 associates the information about the productivity discrimination model M4 with the identification information “M4” for identifying the productivity discrimination model M4, and stores the information in the storage unit 120. good.

(Estimation unit 133)
The estimating unit 133 estimates the work productivity of each person located in the space where the plurality of people is located based on the information about the amount of work done by each person located in the space where the plurality of people are located. Specifically, the estimation unit 133 may estimate the work productivity of each person before air conditioning based on the amount of work of each person in the space before air conditioning where a plurality of people are located. In addition, the estimation unit 133 may estimate the work productivity of each person after air conditioning based on the amount of work of each person in the space after air conditioning.

More specifically, the estimating unit 133 uses productivity discrimination learned to estimate the work productivity of each person located in a predetermined space from information on the amount of work of each person located in the predetermined space. Using the model M4, the work productivity of each person in the space is estimated from the information on the amount of work done by each person in the space. In addition, the estimating unit 133 uses, as information about the amount of work of each person, PC operation information about the amount of operation of the information device by each person, voice information about the amount of speech of each person, or person information about the amount of operation of each person. The work productivity of each person is estimated based on at least one of the activity information. Here, a process of estimating the productivity of each person according to the embodiment will be described in detail with reference to FIG. 3 . FIG. 3 is a diagram for explaining an example of an individual person's productivity estimation process according to the embodiment.

In FIG. 3, the estimating unit 133 may generate a feature vector V1 indicating the features of the PC operation information D1 regarding the operation amount of the information device by each person, as an example of the information regarding the amount of work of each person. Specifically, the estimating unit 133 may refer to the storage unit 120 to acquire information about the learning model M1 acquired by the acquiring unit 131 . Subsequently, when acquiring the information about the learning model M1, the estimation unit 133 may input the PC operation information D1 of each person to the learning model M1 as input information for the learning model M1. Subsequently, the estimating unit 133 may output from the learning model M1, as the output information of the learning model M1, a feature vector V1 indicating the features of the PC operation information of each person.

Further, the estimation unit 133 may generate a feature vector V2 indicating the features of the speech information D2 regarding the amount of speech of each person as an example of information regarding the amount of work of each person. Specifically, the estimating unit 133 may refer to the storage unit 120 to acquire information about the learning model M2 acquired by the acquiring unit 131 . Subsequently, when acquiring the information about the learning model M2, the estimating unit 133 may input the voice information D2 of each person to the learning model M2 as the input information for the learning model M2. Subsequently, the estimating unit 133 may output the feature vector V2 representing the feature of the speech information D2 of each person from the learning model M2 as the output information of the learning model M2.

Moreover, the estimation unit 133 may generate a feature vector V3 indicating the features of the person activity information D3 regarding the amount of movement of each person, as an example of the information regarding the amount of work of each person. Specifically, the estimating unit 133 may refer to the storage unit 120 to acquire information about the learning model M3 acquired by the acquiring unit 131 . Subsequently, when acquiring the information about the learning model M3, the estimating unit 133 may input the person activity information D3 of each person to the learning model M3 as input information for the learning model M3. Subsequently, the estimating unit 133 may output the feature vector V3 indicating the feature of the person activity information D3 of each person from the learning model M3 as the output information of the learning model M3.

After generating the feature vectors V1 to V3, the estimation unit 133 may then generate a feature map MP1 indicating the features of the feature vectors V1 to V3 based on the generated feature vectors V1 to V3. Subsequently, after generating the feature map MP1, the estimating unit 133 may generate a feature vector V4 representing the feature of the feature map MP1 from the feature map MP1 by performing maximum value pooling on the feature map MP1. For example, the estimating unit 133 may generate a 2048-dimensional feature vector V4 indicating the features of the feature map MP1 by maximum value pooling. In this way, the estimating unit 133 may generate the feature vector V4 that indicates the feature of the information regarding the amount of work of each person located in the space.

Subsequently, the estimation unit 133 may generate the feature vector V4 indicating the characteristics of the information on the workload of each person, and estimate the productivity of each person based on the generated feature vector V4. For example, the estimating unit 133 may refer to the storage unit 120 to obtain information regarding the productivity discrimination model M4 generated by the generating unit 132 . Subsequently, when acquiring the information on the productivity discrimination model M4, the estimation unit 133 may input the generated feature vector V4 to the productivity discrimination model M4 as input information for the productivity discrimination model M4. Subsequently, the estimating unit 133 may output a category C1 (either “high”, “normal”, or “low”) indicating the productivity of each person as the output information of the productivity discrimination model M4.

(Air conditioning control unit 134)
The air conditioning control unit 134 controls air conditioning using an air conditioning control model. The air-conditioning control unit 134 controls air-conditioning using the air-conditioning control model M7 that has undergone reinforcement learning by the learning unit 136 . Specifically, the air-conditioning control unit 134 calculates the work productivity of each person estimated based on the amount of work of each person in a space where a plurality of people are present and is air-conditioned by an air-conditioning device. Air conditioning is controlled by reinforcement learning based on the value indicating the group's reward for air conditioning for multiple people located in the space determined by the method and the air conditioning control information when the air conditioning is performed by the air conditioning equipment. Air conditioning is controlled using the air conditioning control model M7.

More specifically, the air conditioning control unit 134 outputs from the air conditioning control model as output information of the air conditioning control model when information about an image of a space is input to the air conditioning control model as input information of the air conditioning control model. Air conditioning is controlled based on the air conditioning control information. For example, when information about an RGB image of a space is input to the air conditioning control model M7 as an example of information about an image of the space, the air conditioning control unit 134 outputs the air conditioning control model M7 as the output information of the air conditioning control model M7. Air conditioning is controlled based on the air conditioning control information output from M7. For example, the air-conditioning control unit 134 may generate a feature vector V5 representing features of an RGB image obtained by imaging a space in which a plurality of people are present, as an example of information related to the RGB image. For example, the air-conditioning control unit 134 may refer to the storage unit 120 to obtain an RGB image of a space in which a plurality of people are located and information related to the learning model M5. Subsequently, after acquiring the RGB image, the air conditioning control unit 134 may use the learning model M5 to generate a feature vector V5 representing the features of the RGB image from the RGB image. Based on the generated feature vector V5, the air conditioning control unit 134 may generate a feature vector V7 that indicates the feature of the space in which the plurality of persons are located. Subsequently, the air-conditioning control unit 134 may use the air-conditioning control model M7 to estimate air-conditioning control information for controlling air-conditioning by air-conditioning equipment installed in the space where a plurality of people are located from the feature vector V7.

Further, when the information on the thermal image of the space is input to the air conditioning control model M7 as an example of the information on the image of the space, the air conditioning control unit 134 outputs the air conditioning control model M7 as the output information of the air conditioning control model M7. Air conditioning is controlled based on the air conditioning control information output from M7. For example, the air-conditioning control unit 134 may generate, as an example of information about the thermal image, a feature vector V6 indicating the features of a thermal image obtained by imaging a space in which a plurality of people are present. For example, the air-conditioning control unit 134 may refer to the storage unit 120 to acquire a thermal image of a space in which a plurality of people are located and information on the learning model M6. Subsequently, when acquiring the thermal image, the air conditioning control unit 134 may use the learning model M6 to generate a feature vector V6 representing the features of the thermal image from the thermal image. Then, based on the generated feature vector V6, the air conditioning control unit 134 may generate a feature vector V7 indicating the feature of the space in which the plurality of persons are located. Subsequently, the air-conditioning control unit 134 may use the air-conditioning control model M7 to estimate air-conditioning control information for controlling air-conditioning by air-conditioning equipment installed in the space where a plurality of people are located from the feature vector V7.

Further, the air conditioning control unit 134 outputs from the air conditioning control model M7 as the output information of the air conditioning control model M7 when the environmental information regarding the temperature and humidity of the space where a plurality of persons are located is input to the air conditioning control model M7. Air conditioning is controlled based on the air conditioning control information. For example, the air-conditioning control unit 134 may generate a feature vector V8 indicating the characteristics of the environmental information regarding the temperature and humidity of the space in which a plurality of people are present, as an example of the environmental information regarding the temperature and humidity of the space. For example, the air-conditioning control unit 134 may refer to the storage unit 120 to obtain environment information regarding the temperature and humidity of the space where a plurality of people are located and information regarding the learning model M8. Subsequently, when the environmental information regarding the temperature and humidity is acquired, the air conditioning control unit 134 uses the learning model M8 to generate a feature vector V8 representing the characteristics of the environmental information regarding the temperature and humidity from the information regarding the temperature and humidity. good. Then, based on the generated feature vector V8, the air conditioning control unit 134 may generate a feature vector V7 that indicates the feature of the space where a plurality of persons are located. Subsequently, the air-conditioning control unit 134 may use the air-conditioning control model M7 to estimate air-conditioning control information for controlling air-conditioning by air-conditioning equipment installed in the space where a plurality of people are located from the feature vector V7.

More specifically, after generating the feature vectors V5 to V6 and V8, the air conditioning control unit 134 generates a feature vector V7 representing the feature of the feature vectors V5 to V6 and V8 based on the generated feature vectors V5 to V6 and V8. can be generated. For example, based on the generated feature vectors V5 to V6 and V8, the air conditioning control unit 134 may generate a feature map MP2 indicating the features of the feature vectors V5 to V6 and V8. Subsequently, after generating the feature map MP2, the air conditioning control unit 134 may generate a feature vector V7 representing the feature of the feature map MP2 from the feature map MP2 by performing maximum value pooling on the feature map MP2. For example, the air-conditioning control unit 134 may generate a 2048-dimensional feature vector V7 indicating the feature of the feature map MP2 by maximum value pooling. In this manner, the air-conditioning control unit 134 may generate the feature vector V7 that indicates the features of the space in which a plurality of persons are located. Subsequently, after generating the feature vector V7, the air conditioning control unit 134 uses the air conditioning control model M7 to obtain air conditioning control information for controlling air conditioning by air conditioning equipment installed in the space where a plurality of people are located from the feature vector V7. You can guess. More specifically, the air conditioning control unit 134 may input the feature vector V7 as input information for the air conditioning control model M7. Subsequently, the air-conditioning control unit 134 may output air-conditioning control information for controlling air-conditioning by air-conditioning equipment installed in a space where a plurality of people are present as output information of the air-conditioning control model M7.

Subsequently, the air conditioning control unit 134 may control the air conditioning of the space based on the air conditioning control information output from the air conditioning control model M7. Specifically, the air conditioning control unit 134 may transmit the air conditioning control information output from the air conditioning control model M7 to the air conditioner 10 installed in the space subject to air conditioning control. In this way, the air conditioning control unit 134 may control the air conditioning of the space using the air conditioning control model M7.

(Determination unit 135)
The determination unit 135 determines the number of people in the space based on the work productivity of each person estimated based on the amount of work done by each person in the space air-conditioned by the air conditioner. A value is determined that indicates the collective reward for air conditioning for multiple persons. Specifically, the determination unit 135 compares the work productivity of each person before air conditioning by the air conditioning equipment and the work productivity of each person after air conditioning by the air conditioning equipment, based on the group remuneration determine the value that indicates More specifically, the determination unit 135 determines the individual performance of air conditioning for each person based on a comparison between the work productivity of each person before air conditioning and the work productivity of each person after air conditioning. Determines the value that indicates the reward for . Subsequently, the determining unit 135 determines a value indicating the group reward based on the value indicating the individual reward. This point will be described in detail with reference to FIG. FIG. 5 is a diagram for explaining an example of a remuneration table for determining individual remunerations according to the embodiment.

In FIG. 5 , the state before air-conditioning control indicates a state before the air-conditioning equipment air-conditions the space where a plurality of people are located according to the air-conditioning control information of the air-conditioning control unit 134 . On the other hand, the state after air-conditioning control indicates the state after the air-conditioning equipment has air-conditioned the space where a plurality of persons are located according to the air-conditioning control information of the air-conditioning control unit 134 .

In the example shown in FIG. 5, the determination unit 135 determines that when the productivity of each person before air-conditioning is "normal" and the productivity of each person after air-conditioning is "high" , the value that indicates the individual reward for air conditioning may be determined to be "+1", which is a positive reward. In addition, if the productivity of the individual person before air conditioning is "low" and the productivity of the individual person after air conditioning is "high", the determination unit 135 determines the individual reward for air conditioning. may be determined to be a positive reward of "+3". In addition, if the productivity of each person before air conditioning is in a “low” state and the productivity of each person after air conditioning is in a “normal” state, the determination unit 135 determines that the individual person’s productivity for air conditioning is “low”. A value indicative of the reward may be determined to be "+1", which is a positive reward. In this way, when the individual person's work productivity after air-conditioning is higher than the individual person's work productivity before air-conditioning, the determination unit 135 determines the individual reward as a positive reward. You may determine the value shown. In addition, in the example of FIG. 5, the value of the reward is indicated by the magnitude relationship of the numbers. For example, FIG. 5 shows that "+3" is higher than "+1" as the reward value. That is, in FIG. 5, rather than the air conditioning that changes the productivity of each person from the state of "normal" before air conditioning to the state of "high" productivity of each person after air conditioning. Air-conditioning that changes the productivity of an individual person from a state of “low” to a state of “high” productivity after air-conditioning is considered to be an air-conditioning operation that improves the comfort of each person. This indicates that the value of the air conditioning is high. In this way, the determination unit 135 may determine to give a reward of higher value to the executed air conditioning operation, as the executed air conditioning operation is control that further improves comfort for the individual person.

In addition, if the productivity of the individual person before air conditioning is "high" and the productivity of the individual person after air conditioning is "high", the determination unit 135 determines the individual reward for air conditioning. may be determined to be "0". In addition, if the productivity of each person before air conditioning is in a "normal" state and the productivity of each person after air conditioning is in a state of "normal", the determination unit 135 may be determined to be "0". In addition, if the productivity of the individual person before air conditioning is "low" and the productivity of the individual person after air conditioning is "low", the determination unit 135 determines the individual reward for air conditioning. may be determined to be "0". In this way, the determination unit 135 determines the value indicating the individual remuneration to be zero when the individual person's work productivity before air conditioning is the same as the individual person's work productivity after air conditioning.

In addition, when the productivity of each person before air conditioning is in a state of “high” and the productivity of each person after air conditioning is in a state of “normal,” the determining unit 135 A value indicative of the reward may be determined to be "-1", which is a negative reward. In addition, if the productivity of the individual person before air conditioning is "high" and the productivity of the individual person after air conditioning is "low", the determination unit 135 determines the individual reward for air conditioning. may be determined to be "-3", which is a negative reward. In addition, when the productivity of each person before air conditioning is in a “normal” state and the productivity of each person after air conditioning is in a “low” state, the determination unit 135 determines whether the individual person’s productivity for air conditioning is “normal”. A value indicative of the reward may be determined to be "-1", which is a negative reward. In this way, when the individual person's work productivity after air-conditioning is lower than the individual person's work productivity before air-conditioning, the determination unit 135 gives the individual reward as a negative reward. determine the value shown. Also, FIG. 5 shows that "-3" is lower in value than "-1" as the reward value. That is, in FIG. 5, the individual person's productivity before air conditioning is changed from the state where the productivity of each person is "high" before air conditioning to the state where the productivity of each person is "normal" after air conditioning. Air conditioning that changes the productivity of an individual person from a state of “high” to a state of “low” productivity for each person after air conditioning is considered to be an air conditioning operation that reduces the comfort of each person. It indicates that the value of the air conditioning is low. In this way, the determination unit 135 may determine to give a lower reward to the performed air conditioning operation as the performed air conditioning operation is a control that lowers the comfort level of the individual person.

Subsequently, when the value indicating the individual reward is determined, the determining unit 135 determines the value indicating the collective reward for air conditioning for a plurality of persons located in the space based on the determined individual reward value. you can For example, the determining unit 135 determines the value indicating the group reward by adding the values indicating the individual rewards for all of the plurality of persons based on a predetermined condition.

(learning unit 136)
The learning unit 136 performs reinforcement learning of an air-conditioning control model that outputs air-conditioning control information for controlling air-conditioning by the air-conditioning equipment based on the air-conditioning control information when the air-conditioning is performed by the air-conditioning equipment and a value indicating the reward of the group. . Specifically, the learning unit 136 may perform reinforcement learning on the air conditioning control model M7 so that the value indicating the reward for the group is maximized. After performing reinforcement learning on the air conditioning control model M7, the learning unit 136 may associate information about the air conditioning control model M7 that has undergone reinforcement learning with identification information “M7” that identifies the air conditioning control model M7, and store the information in the storage unit 120. .

[4. Information processing flow]
Next, the procedure of information processing according to the embodiment will be described with reference to FIGS. 6 to 12. FIG. 6 to 12 are flowcharts showing information processing procedures according to the embodiment.

First, the procedure of information processing according to the embodiment will be described with reference to FIG. In FIG. 6, the acquisition unit 131 may acquire information about the amount of work of each person located in a space where a plurality of people are located. Specifically, the acquisition unit 131 may acquire PC operation information regarding the amount of operation of the information device by each person as the information regarding the amount of work of each person. In addition, the acquisition unit 131 may acquire voice information regarding the amount of speech of each person as the information regarding the amount of work of each person. In addition, the acquisition unit 131 may acquire person activity information related to the amount of movement of each person as the information related to the amount of work of each person.

Subsequently, when the acquisition unit 131 acquires the information about the amount of work of each person, the estimation unit 133 uses the productivity discrimination model M4 to estimate the productivity of each person in the space (step S11). Specifically, the estimating unit 133 generates a feature vector V1 indicating the characteristics of the PC operation information regarding the operation amount of the information device by each person, a feature vector V2 indicating the characteristics of the voice information regarding the amount of speech of each person, and an individual The feature vector V4 generated based on the feature vector V3 indicating the feature of the person activity information regarding the amount of motion of the person may be input to the productivity determination model M4 as input information. Subsequently, the estimation unit 133 may output the productivity of each person located in the space as output information of the productivity discrimination model M4.

Subsequently, when the estimation unit 133 outputs the productivity of each person located in the space as the output information of the productivity discrimination model M4, the estimation unit 133 saves the output data (step S12). Specifically, the estimating unit 133 may store the data output as the output information of the productivity discrimination model M4 in the storage unit 120 as the productivity of each person located in the previous space.

Next, the procedure of information processing according to the embodiment will be described with reference to FIG. In FIG. 7, subsequent to FIG. 6, the air conditioning control unit 134 uses the air conditioning control model M7 to estimate air conditioning control information for controlling air conditioning by air conditioning equipment installed in the space (step S21). Specifically, the air conditioning control unit 134 may input sensor information acquired by the acquiring unit 131 as input information for the air conditioning control model M7. More specifically, the air-conditioning control unit 134 may input a feature vector V7 indicating the characteristics of the sensor information acquired by the acquisition unit 131 as input information for the air-conditioning control model M7. For example, the air-conditioning control unit 134 is generated based on a feature vector V5 indicating the features of the RGB image, a feature vector V6 indicating the features of the thermal image, and a feature vector V8 indicating the features of the environmental information regarding the temperature and humidity of the space. The characteristic vector V7 may be input to the air conditioning control model M7 as input information. Subsequently, the air-conditioning control unit 134 may output air-conditioning control information for controlling air-conditioning by air-conditioning equipment installed in the space as output information of the air-conditioning control model M7.

Next, the procedure of information processing according to the embodiment will be described with reference to FIG. In FIG. 8, subsequent to FIG. 7, the air conditioner 10 air-conditions the space based on the air-conditioning control information output from the information processing device 100 (step S31). Specifically, the air conditioning device 10 may acquire air conditioning control information from the information processing device 100 . Subsequently, when acquiring the air-conditioning control information, the air conditioner 10 may air-condition the space by performing an air-conditioning operation based on the acquired air-conditioning control information. For example, the air conditioner 10 may air-condition the space according to the cooling/heating settings included in the air-conditioning control information. Also, the air conditioner 10 may adjust the temperature of the space to the set temperature included in the air conditioning control information. Further, the air conditioner 10 may adjust the humidity of the space so as to reach the set humidity included in the air conditioning control information.

Moreover, as a result of adjusting the air conditioning of the space by the air conditioner 10, the environment of the space changes (step S32). Specifically, the temperature of the space may change from the temperature before air conditioning by the air conditioner 10 to the set temperature. Further, the humidity of the space may change from the humidity before air conditioning by the air conditioner 10 to the set humidity.

Further, the sensor device 20 monitors the environment of the space where a plurality of people are present after air conditioning (step S33). Specifically, the sensor device 20 may acquire sensor information indicating the physical state of an air-conditioned space in which a plurality of people are located and the physical state of a person in the air-conditioned space. Subsequently, the sensor device 20 may transmit the acquired sensor information to the information processing device 100 .

Next, the procedure of information processing according to the embodiment will be described with reference to FIG. 9 . In FIG. 9 , following FIG. 8 , the acquisition unit 131 may acquire information about the amount of work of each person in the air-conditioned space in which a plurality of people are present. Specifically, the acquiring unit 131 may acquire PC operation information related to the amount of operation of the information device by each person in the air-conditioned space as the information related to the amount of work done by each person after air conditioning. In addition, the acquisition unit 131 may acquire voice information about the amount of speech of each person in the air-conditioned space as the information about the amount of work of each person after the air conditioning. In addition, the acquisition unit 131 may acquire human activity information related to the amount of movement of each person in the air-conditioned space as the information related to the amount of work done by each person after air conditioning.

Subsequently, when the acquisition unit 131 acquires the information about the amount of work done by each person after air conditioning, the estimation unit 133 uses the productivity discrimination model M4 to calculate the productivity of each person in the space after air conditioning. The gender is estimated (step S41). Specifically, the estimating unit 133 may input a feature vector V4' indicating the characteristics of the amount of work of each person in the air-conditioned space as input information for the productivity discrimination model M4. More specifically, the estimating unit 133 generates a feature vector V1′ that indicates the characteristics of the PC operation information related to the operation amount of the information device by each person in the air-conditioned space, and the individual person in the air-conditioned space. A feature vector generated based on a feature vector V2′ representing the feature of voice information related to the amount of utterances and a feature vector V3′ representing the feature of human activity information related to the amount of motion of each person in the air-conditioned space. V4' may be input to the productivity discrimination model M4 as input information. Subsequently, the estimation unit 133 may output the productivity of each person located in the air-conditioned space as output information of the productivity discrimination model M4.

Next, the procedure of information processing according to the embodiment will be described with reference to FIG. 10 . In FIG. 10, following FIG. 9, the determination unit 135 outputs the productivity of each person in the previous space stored in the storage unit 120 and the output information of the productivity discrimination model M4. Values indicative of individual rewards for air conditioning for individual persons may be determined based on a comparison with the productivity of individual persons in the air-conditioned space. Subsequently, when the value indicating the individual reward is determined, the determining unit 135 adds the values indicating the individual reward for all of the plurality of persons present in the space, thereby determining the air conditioning for the plurality of persons present in the space. A value indicating a collective reward (also referred to as an overall reward) for is determined (step S51).

Next, the procedure of information processing according to the embodiment will be described with reference to FIG. 11 . In FIG. 11, following FIG. 10, the estimating unit 133 outputs the productivity of each person in the air-conditioned space as the output information of the productivity discrimination model M4, and saves the output data (step S61 ). Specifically, the estimation unit 133 may store the data output as the output information of the productivity discrimination model M4 in the storage unit 120 as the productivity of the previous individual person.

Next, the procedure of information processing according to the embodiment will be described with reference to FIG. 12 . In FIG. 12 , following FIG. 11 , the learning unit 136 generates an air conditioning control model based on a value indicating a group reward for air conditioning (also referred to as an overall reward) and air conditioning control information indicating air conditioning operation related to air conditioning control. Reinforcement learning of M7 is performed (step S71).

[5. Modification]
The information processing system 1 according to the above-described embodiments may be implemented in various different forms other than the above-described embodiments. Therefore, other embodiments of the information processing system 1 will be described below. In addition, the same code|symbol is attached|subjected to the same part as embodiment, and description is abbreviate|omitted.

In the above-described embodiment, an example in which the estimation unit 133 estimates the productivity of each person using the productivity discrimination model M4 has been described. Productivity may be estimated. Specifically, as an example of the productivity of each person, the estimation unit 133 calculates a productivity score (hereinafter referred to as an individual productivity score) may be calculated.

For example, the estimating unit 133, as an example of information about the amount of work of each person, based on the PC operation information about the amount of operation of the information device by each person, as an example of the productivity of the work of each person, A productivity score related to PC operation, which is a value obtained by dividing the operation amount of a person by the duration of operation of the information device by each person, may be calculated. Here, the calculation formula of the productivity score p (PC operation information) regarding PC operation is shown by the following formula (1).

For example, when the acquisition unit 131 acquires the operation log information of each person, the estimation unit 133 may identify each operation (keyboard input operation, mouse operation, printer output, communication, etc.) by each person. . Subsequently, when each operation of each person is specified, the estimation unit 133 may calculate the duration of each operation by each person for each operation. Further, the estimating unit 133 calculates the amount of keyboard input, the amount of mouse operation, the length of PC startup time, the amount of power consumption, the number of printer outputs, and the amount of communication as examples of the operation amount of each person. may be calculated. Subsequently, when the operation amount of each person is calculated, the estimation unit 133 obtains a value obtained by dividing each operation amount of each person by the duration of each operation of each person as an example of a productivity score related to PC operation. It may be calculated for each operation. In addition, the estimating unit 133 may calculate a value obtained by dividing each operation amount of each person by the duration of each operation of each person and adding the value as the productivity score related to the PC operation.

In addition, the estimation unit 133 calculates, as an example of the work productivity of each person, the speech amount A productivity score for an utterance may be calculated, which is the quotient divided by the duration of an utterance by an individual person. Here, the formula for calculating the productivity score p (speech information) related to speech is given by the following formula (2).

For example, when the speech information acquired by the acquisition unit 131 includes voices of a plurality of speakers, the estimation unit 133 identifies the voices of each speaker using a known multi-speaker identification technology. you can Subsequently, the estimation unit 133 identifies the voice of each speaker (individual person), and uses a known sound source separation (noise removal) technique to remove noise contained in the voice of each person. you can Subsequently, the estimating unit 133 may calculate the utterance duration of each person after removing the noise. Note that the estimating unit 133 may calculate the length of the sentence related to the utterance of each person instead of the duration of the utterance of each person.

Further, after identifying the voice of each person, the estimation unit 133 may calculate the number of times of speech by each person as an example of information related to the amount of speech of each person. In addition, the estimation unit 133 may calculate silence time (length of time between utterances) as an example of information related to the amount of utterances of individual persons. In addition, the estimation unit 133 may calculate the speed of speech as an example of information regarding the speech volume of each person. In addition, the estimation unit 133 may calculate the frequency of occurrence of interjections (such as “eh”) as an example of information related to the amount of speech of each person. In addition, the estimation unit 133 may calculate the frequency of occurrence of laughter as an example of information regarding the amount of speech of each person. In addition, the estimation unit 133 may calculate the speaker switching frequency as an example of information regarding the amount of speech of each person. In addition, the estimating unit 133 also collects the number of times each person speaks, silence time (length of time between utterances), utterance speed, interjections (e.g. ), the frequency of laughter, and the frequency of speaker switching may be calculated. After calculating the information about the amount of speech of each person, the estimation unit 133 may calculate a value obtained by dividing the information about the amount of speech of each person by the duration of the speech of each person as the productivity score of speech. .

In addition, the estimation unit 133 calculates, as an example of the work productivity of each person, based on the person activity information related to the amount of movement of each person, as an example of information related to the amount of work of each person. A productivity score may be calculated for person activity, which is the amount divided by the duration of action by an individual person. Here, the formula for calculating the productivity score p (person activity information) related to person activity is shown in Equation (3) below.

For example, when an image is acquired by the acquiring unit 131, the estimating unit 133 may specify an area in which each person included in the image is captured (hereinafter also referred to as an individual person area). For example, the estimation unit 133 may identify individual human regions included in the image based on posture estimation processing for estimating the posture of each individual person. Subsequently, after specifying each person region, the estimating unit 133 may specify each action (standing, sitting, moving, etc.) of each person based on the change in posture of each person over time. Subsequently, when each action of each person is specified, the estimation unit 133 may calculate the duration of each action by each person for each action. Further, when each action of each person is specified, the estimation unit 133 may calculate the number of times each action of each person is performed for each action as an example of the amount of movement of each person. Subsequently, when calculating the number of times of each action of the individual person, the estimation unit 133 divides the number of times of each action of the individual person by the duration of each action of the individual person as an example of a productivity score related to the person's activity. The value may be calculated for each action.

In addition, after identifying each human region, the estimating unit 133 determines each movement of the head of each person (nodding, shaking the head, shaking the head, etc.) based on the temporal change in the posture of the head of each person. tilting, etc.) may be specified. After identifying each movement of the head of each person, the estimation unit 133 may then calculate the duration of each movement of the head of each person for each head movement. In addition, after specifying each movement of the head of each person, the estimation unit 133 calculates the number of times each movement of the head of each person is performed for each movement of the head as an example of the amount of movement of each person. you can Subsequently, when the estimation unit 133 calculates the number of times of each movement of the head of each person, the estimation unit 133 calculates the number of times of each movement of the head of each person as an example of a productivity score related to the person's activity. A value divided by the duration of each head movement may be calculated for each head movement. In addition, the estimating unit 133 divides the number of actions of each individual person by the duration of each action of each individual person and the number of head movements of each individual person as productivity scores related to human activities. divided by the duration of each movement of the individual person's head.

In addition, the estimation unit 133 calculates a value obtained by adding a productivity score related to PC operation, a productivity score related to speech, and a productivity score related to human activity as an individual productivity score, as an example of the work productivity of each person. You can Here, the calculation formula of each productivity score P is shown by the following formula (4).

Also, the determination unit 135 may determine the reward for the group based on the individual productivity scores calculated by the estimation unit 133 . This point will be described in detail with reference to FIG. FIG. 13 is a diagram for explaining an example of a remuneration table for determining individual remunerations according to the modification. In FIG. 13, individual productivity scores before air conditioning (the state before air conditioning) are indicated by P(t-1), and individual productivity scores after air conditioning are indicated by P(t). Based on the individual productivity score P(t-1) before air conditioning and the individual productivity score P(t) after air conditioning, the determining unit 135 determines "(P(t)/ P(t−1))−1” may be calculated.

Specifically, when the individual productivity score P(t) after air-conditioning is greater than the individual productivity score P(t-1) before air-conditioning, the determining unit 135 determines the individual reward. The indicated value may be determined to be a positive reward expressed as "(P(t)/P(t-1))-1". Further, when the individual productivity score P(t-1) before air conditioning and the individual productivity score P(t) after air conditioning are the same, the determination unit 135 sets the value indicating the individual reward to "(P (t)/P(t-1))-1". Further, when the individual productivity score P(t) after air-conditioning is smaller than the individual productivity score P(t-1) before air-conditioning, the determining unit 135 sets a value indicating the individual reward. A negative reward expressed as "(P(t)/P(t-1))-1" may be determined.

[6. effect〕
As described above, the information processing apparatus 100 according to the embodiment includes the determination unit 135 and the learning unit 136. FIG. The determination unit 135 determines the number of people in the space based on the work productivity of each person estimated based on the amount of work done by each person in the space air-conditioned by the air conditioner. A value is determined that indicates the collective reward for air conditioning for multiple persons. The learning unit 136 performs reinforcement learning of an air-conditioning control model that outputs air-conditioning control information for controlling air-conditioning by the air-conditioning equipment based on the air-conditioning control information when the air-conditioning is performed by the air-conditioning equipment and a value indicating the reward of the group. .

As a result, the information processing apparatus 100 can perform air conditioning control that achieves optimal productivity for a group based on the work productivity of each person. Therefore, the information processing apparatus 100 can control air conditioning in consideration of group productivity.

In addition, the determination unit 135 determines a value indicating the remuneration of the group based on a comparison between the work productivity of each person before air conditioning by the air conditioning equipment and the work productivity of each person after air conditioning by the air conditioning equipment. to decide. Specifically, the determining unit 135 determines the individual person's work productivity for air conditioning based on a comparison between the individual person's work productivity before air conditioning and the individual person's work productivity after air conditioning. Determine the value that indicates the reward. Subsequently, the determining unit 135 determines a value indicating the group reward based on the value indicating the individual reward. For example, the determining unit 135 determines the value indicating the group reward by adding the values indicating the individual rewards for all of the plurality of persons based on a predetermined condition.

As a result, the information processing apparatus 100 can determine the reward for the group for air conditioning based on the change in the productivity of each person before and after air conditioning, thereby enabling air conditioning control that considers the productivity of the group. be able to.

Further, when the work productivity of the individual person after air-conditioning is higher than the work productivity of the individual person before air-conditioning, the determining unit 135 determines the individual reward as a positive reward. to decide.

As a result, the information processing apparatus 100 outputs the air conditioning operation that changes the productivity of the group to a higher state in order to give a positive reward to the air conditioning operation that changes the productivity of each person to a higher state. Reinforcement learning can be applied to the air-conditioning control model. Therefore, the information processing apparatus 100 can enable air conditioning control that changes the productivity of the group to a higher state.

Further, if the individual person's work productivity after air-conditioning is lower than the individual person's work productivity before air-conditioning, the determination unit 135 determines the individual reward as a negative reward. to decide.

As a result, the information processing apparatus 100 outputs an air-conditioning operation that lowers the productivity of the group in order to give a negative reward to the air-conditioning operation that lowers the productivity of each person. Reinforcement learning can be applied to the air conditioning control model so that it does not occur. Therefore, the information processing apparatus 100 can enable air conditioning control that does not change the productivity of the group to a lower state.

Moreover, the determination unit 135 determines the value indicating the individual remuneration to be zero when the individual person's work productivity before air conditioning is the same as the individual person's work productivity after air conditioning.

As a result, the information processing apparatus 100 does not give rewards for air conditioning operations that do not change the productivity of individual persons, so the information processing apparatus 100 performs reinforcement learning of the air conditioning control model so as to output air conditioning operations that change the productivity of the group. be able to.

The information processing device 100 further includes an estimation unit 133 . The estimation unit 133 estimates the work productivity of each person present in the space based on the information about the amount of work of each person present in the space. Specifically, the estimating unit 133 calculates the amount of operation as the productivity of the work of each person based on the PC operation information about the amount of operation of the information device by each person as the information about the amount of work of each person. Calculate a productivity score for PC operation, which is a value divided by the duration of operation of the information appliance by the individual person. Further, the estimation unit 133 calculates the amount of speech as the productivity of the work of each person based on the voice information about the amount of speech of each person as information about the amount of work of each person. Compute a productivity score for an utterance divided by time. In addition, the estimation unit 133 calculates the amount of movement as the productivity of the work of each person based on the person activity information related to the amount of movement of each person as information related to the amount of work of each person. Compute a productivity score for the person activity divided by the duration.

As a result, the information processing apparatus 100 can more accurately estimate the productivity of each person by quantifying the amount of work of each person.

In addition, the estimating unit 133 uses a productivity discrimination model that has been learned to estimate the work productivity of each person located in a predetermined space from information on the amount of work of each person located in the predetermined space. , the work productivity of each person in the space is estimated from the information on the amount of work of each person in the space. Specifically, the estimating unit 133 uses, as information about the amount of work of each person, PC operation information about the amount of operation of the information device by each person, voice information about the amount of speech of each person, or motion information of each person. Estimate job productivity of an individual person based on at least one of the quantitative person activity information.

As a result, the information processing apparatus 100 can more accurately estimate individual productivity by using a machine learning model learned to estimate individual productivity.

The information processing apparatus 100 further includes an air conditioning control section 134 . The air conditioning control unit 134 controls air conditioning using an air conditioning control model. Specifically, the air conditioning control unit 134 outputs from the air conditioning control model as output information of the air conditioning control model when information about an image of the space captured is input to the air conditioning control model as input information of the air conditioning control model. Air conditioning is controlled based on the air conditioning control information. For example, the air-conditioning control unit 134 inputs information about at least one of a thermal image of the space captured by an infrared camera and an RGB image of the space captured by an RGB camera into the air-conditioning control model as information about the image of the space. air-conditioning is controlled based on the air-conditioning control information output from the air-conditioning control model as the output information of the air-conditioning control model. Further, the air-conditioning control unit 134 controls the air-conditioning control output from the air-conditioning control model as the output information of the air-conditioning control model when the environmental information about the temperature and humidity of the space is input to the air-conditioning control model as the input information of the air-conditioning control model. Control air conditioning based on information.

As a result, the information processing apparatus 100 can control the air conditioning of the space using the air conditioning control model that has learned the air conditioning operation considering the productivity of the group. can be

In addition, the air conditioning control unit 134 is determined based on the work productivity of each person estimated based on the amount of work of each person in a space where a plurality of people are present and air conditioning is performed by the air conditioning equipment. An air-conditioning control model that controls air-conditioning by air-conditioning equipment that has undergone reinforcement learning based on a value that indicates the group's reward for air-conditioning for multiple people in the same space and air-conditioning control information when air-conditioning is performed by air-conditioning equipment. to control the air conditioning.

As a result, the information processing apparatus 100 can control the air conditioning of the space using the air conditioning control model that has learned the air conditioning operation in consideration of the productivity of the group based on the work productivity of each person. It is possible to control the air conditioning in consideration of productivity.

[7. Hardware configuration]
Also, the information processing apparatus 100 according to the above-described embodiments is implemented by a computer 1000 configured as shown in FIG. 14, for example. FIG. 14 is a hardware configuration diagram showing an example of a computer that implements the functions of the information processing apparatus 100. As shown in FIG. Computer 1000 includes CPU 1100 , RAM 1200 , ROM 1300 , HDD 1400 , communication interface (I/F) 1500 , input/output interface (I/F) 1600 and media interface (I/F) 1700 .

The CPU 1100 operates based on programs stored in the ROM 1300 or HDD 1400 and controls each section. The ROM 1300 stores a boot program executed by the CPU 1100 when the computer 1000 is started up, a program depending on the hardware of the computer 1000, and the like.

The HDD 1400 stores programs executed by the CPU 1100, data used by the programs, and the like. Communication interface 1500 receives data from another device via a predetermined communication network, sends the data to CPU 1100, and transmits data generated by CPU 1100 to another device via a predetermined communication network.

The CPU 1100 controls output devices such as displays and printers, and input devices such as keyboards and mice, through an input/output interface 1600 . CPU 1100 acquires data from an input device via input/output interface 1600 . CPU 1100 also outputs the generated data to an output device via input/output interface 1600 .

Media interface 1700 reads programs or data stored in recording medium 1800 and provides them to CPU 1100 via RAM 1200 . CPU 1100 loads such a program from recording medium 1800 onto RAM 1200 via media interface 1700, and executes the loaded program. The recording medium 1800 is, for example, an optical recording medium such as a DVD (Digital Versatile Disc) or a PD (Phase change rewritable disc), a magneto-optical recording medium such as an MO (Magneto-Optical disk), a tape medium, a magnetic recording medium, or a semiconductor memory. etc.

For example, when the computer 1000 functions as the information processing apparatus 100 according to the embodiment, the CPU 1100 of the computer 1000 implements the functions of the control unit 130 by executing programs loaded on the RAM 1200 . CPU 1100 of computer 1000 reads these programs from recording medium 1800 and executes them, but as another example, these programs may be obtained from another device via a predetermined communication network.

As described above, some of the embodiments of the present application have been described in detail based on the drawings. It is possible to carry out the invention in other forms with modifications.

[8. others〕
Further, among the processes described in the above embodiments and modifications, all or part of the processes described as being performed automatically can be performed manually, or described as being performed manually. All or part of the processing can also be performed automatically by known methods. In addition, information including processing procedures, specific names, various data and parameters shown in the above documents and drawings can be arbitrarily changed unless otherwise specified. For example, the various information shown in each drawing is not limited to the illustrated information.

Also, each component of each device illustrated is functionally conceptual, and does not necessarily need to be physically configured as illustrated. In other words, the specific form of distribution and integration of each device is not limited to the one shown in the figure, and all or part of them can be functionally or physically distributed and integrated in arbitrary units according to various loads and usage conditions. Can be integrated and configured. For example, in the above-described embodiment, the information processing apparatus 100 includes the air conditioning control unit 134, the determining unit 135, and the learning unit 136. may be provided. For example, the device including the air conditioning control unit 134 and the device including the determining unit 135 and the learning unit 136 may be separate devices.

Also, the above-described embodiments and modifications can be appropriately combined within a range that does not contradict the processing content.

Also, the above-mentioned "section, module, unit" can be read as "means" or "circuit". For example, the learning unit can be read as learning means or a learning circuit.

1 information processing system 10 air conditioner 20 sensor device 100 information processing device 110 communication unit 120 storage unit 130 control unit 131 acquisition unit 132 generation unit 133 estimation unit 134 air conditioning control unit 135 determination unit 136 learning unit

Claims (18)

Using a productivity discrimination model learned to estimate information about the work productivity of each person from information about the amount of work of each person in a space where a plurality of people are present and air-conditioned by an air conditioning device and estimating information on the work productivity of the individual person before and after the air conditioning from the information on the work amount of the individual person before and after the air conditioning by the air conditioning equipment, and An individual compensation for the air conditioning for the individual person based on a comparison of information about the individual person's work productivity before air conditioning and information about the individual person's work productivity after the air conditioning. and calculating the value indicating the individual reward based on a predetermined condition, thereby determining the value indicating the collective reward for the air conditioning for the plurality of persons located in the space When,
Reinforcement learning is performed on an air conditioning control model that outputs air conditioning control information for controlling air conditioning by the air conditioning equipment based on the air conditioning control information when the air conditioning is performed by the air conditioning equipment and a value indicating a reward for the group. the learning department;
Information processing device. An estimating unit that estimates information on the work productivity of the individual person from the information on the amount of work of the individual person using the productivity discrimination model ,
The estimation unit
As the information on the work productivity of the individual person, calculate the probability that the individual person falls into each category indicating the level of work productivity, and among the probabilities that fall into each category, the highest probability outputs the label of the corresponding category,
The decision unit
Information indicating the level of work productivity of the individual person corresponding to the label before the air conditioning and information indicating the level of work productivity of the individual person corresponding to the label after the air conditioning determining a value indicative of an individual reward for said air conditioning for said individual person based on the comparison with
The information processing device according to claim 1 . The decision unit
The work productivity of the individual person corresponding to the label after the air conditioning is indicated rather than the information indicating the work productivity of the individual person corresponding to the label before the air conditioning. determining the value indicative of the individual reward to be a positive value if the information indicates that the individual is more productive ;
The information processing apparatus according to claim 2 . The decision unit
The work productivity of the individual person corresponding to the label after the air conditioning is indicated rather than the information indicating the work productivity of the individual person corresponding to the label before the air conditioning. determining the value indicative of the individual reward to be negative if the information indicates a lower productivity ;
The information processing apparatus according to claim 2 or 3 . The decision unit
Information indicating the level of work productivity of the individual person corresponding to the label before air conditioning and information indicating the level of work productivity of the individual person corresponding to the label after air conditioning if the same, determine the value indicative of the individual reward to be zero;
The information processing apparatus according to any one of claims 2 to 4 . An estimating unit that calculates a productivity score that indicates the level of work productivity of the individual person as information on the work productivity of the individual person, based on the information on the amount of work of the individual person. prepared,
The estimation unit
Based on the PC operation information regarding the operation amount of the information device by the individual person as the information regarding the amount of work of the individual person, the amount of operation is used as the information regarding the productivity of the work of the individual person. calculating the productivity score for PC operation, which is the value divided by the duration of operation of the information device by
The decision unit
Based on a comparison of the productivity score calculated for the individual person before the air conditioning and the productivity score calculated for the individual person after the air conditioning, the air conditioning for the individual person determine a value that indicates an individual reward;
The information processing apparatus according to any one of claims 1 to 5 . An estimating unit that calculates a productivity score that indicates the level of work productivity of the individual person as information on the work productivity of the individual person, based on the information on the amount of work of the individual person. prepared,
The estimation unit
Based on voice information about the amount of speech of the individual person as information about the amount of work of the individual person, the amount of speech is used as information about the productivity of the work of the individual person, and the duration of speech by the individual person. calculating the productivity score for an utterance divided by time;
The decision unit
Based on a comparison of the productivity score calculated for the individual person before the air conditioning and the productivity score calculated for the individual person after the air conditioning, the air conditioning for the individual person determine a value that indicates an individual reward;
The information processing apparatus according to any one of claims 1 to 6 . An estimating unit that calculates a productivity score that indicates the level of work productivity of the individual person as information on the work productivity of the individual person, based on the information on the amount of work of the individual person. prepared,
The estimation unit
As information on the amount of work of the individual person, based on person activity information on the amount of operation of the individual person, the amount of operation is used as information on the productivity of the work of the individual person. calculating the productivity score for person activity that is divided by duration;
The decision unit
Based on a comparison of the productivity score calculated for the individual person before the air conditioning and the productivity score calculated for the individual person after the air conditioning, the air conditioning for the individual person determine a value that indicates an individual reward;
The information processing apparatus according to any one of claims 1 to 7 . An estimating unit that calculates a productivity score that indicates the level of work productivity of the individual person as information on the work productivity of the individual person, based on the information on the amount of work of the individual person. prepared,
The estimation unit
As the information on the amount of work of each person, PC operation information on the amount of operation of the information device by the individual, voice information on the amount of speech by the individual, or person activity information on the amount of operation of the individual. Calculate the productivity score based on at least one of
The decision unit
Based on a comparison of the productivity score calculated for the individual person before the air conditioning and the productivity score calculated for the individual person after the air conditioning, the air conditioning for the individual person determine a value that indicates an individual reward;
The information processing apparatus according to any one of claims 1 to 8 . an air conditioning control unit that controls air conditioning using the air conditioning control model;
The information processing apparatus according to any one of claims 1 to 9 , further comprising: The air conditioning control unit
When information about an image of a space is input to the air conditioning control model as input information for the air conditioning control model, air conditioning control information output from the air conditioning control model is used as output information for the air conditioning control model. to control the
The information processing apparatus according to claim 10 . The air conditioning control unit
When information about at least one of a thermal image of the space captured by an infrared camera and an RGB image of the space captured by an RGB camera is input to the air conditioning control model as the information about the image captured of the space. , controlling air conditioning based on the air conditioning control information output from the air conditioning control model as output information of the air conditioning control model;
The information processing apparatus according to claim 10 . The air conditioning control unit
When environmental information relating to the temperature and humidity of the space is input to the air conditioning control model as input information for the air conditioning control model, the air conditioning control information output from the air conditioning control model as output information for the air conditioning control model includes: to control air conditioning based on,
The information processing apparatus according to any one of claims 10 to 12 . Using a productivity discrimination model learned to estimate information about the work productivity of each person from information about the amount of work of each person in a space where a plurality of people are present and air-conditioned by an air conditioning device and estimating information on the work productivity of the individual person before and after the air conditioning from the information on the work amount of the individual person before and after the air conditioning by the air conditioning equipment, and An individual compensation for the air conditioning for the individual person based on a comparison of information about the individual person's work productivity before air conditioning and information about the individual person's work productivity after the air conditioning. and calculating the value indicating the individual reward based on a predetermined condition to determine the value indicating the collective reward for the air conditioning for the plurality of persons located in the space. When,
Reinforcement learning is performed on an air conditioning control model that outputs air conditioning control information for controlling air conditioning by the air conditioning equipment based on the air conditioning control information when the air conditioning is performed by the air conditioning equipment and a value indicating a reward for the group. a learning procedure;
An information processing program characterized by causing a computer to execute An information processing system including an information processing device and an air conditioner,
The information processing device is
Using a productivity discrimination model learned to estimate information about the work productivity of each person from information about the amount of work of each person in a space where a plurality of people are present and air-conditioned by an air conditioning device and estimating information on the work productivity of the individual person before and after the air conditioning from the information on the work amount of the individual person before and after the air conditioning by the air conditioning equipment, and An individual compensation for the air conditioning for the individual person based on a comparison of information about the individual person's work productivity before air conditioning and information about the individual person's work productivity after the air conditioning. and calculating the value indicating the individual reward based on a predetermined condition, thereby determining the value indicating the collective reward for the air conditioning for the plurality of persons located in the space When,
Reinforcement learning is performed on an air conditioning control model that outputs air conditioning control information for controlling air conditioning by the air conditioning equipment based on the air conditioning control information when the air conditioning is performed by the air conditioning equipment and a value indicating a reward for the group. the learning department;
with
The air conditioner is
performing air conditioning according to the air conditioning control information output from the air conditioning control model that has undergone reinforcement learning by the learning unit;
Information processing system. Using a productivity discrimination model learned to estimate information about the work productivity of each person from information about the amount of work of each person in a space where a plurality of people are present and air-conditioned by an air conditioning device and estimating information on the work productivity of the individual person before and after the air conditioning from the information on the work amount of the individual person before and after the air conditioning by the air conditioning equipment, and An individual reward for the air conditioning for the individual person based on a comparison of information about the work productivity of the individual person before air conditioning and information about the work productivity of the individual person after the air conditioning. A value indicating a group reward for the air conditioning for the plurality of persons in the space is determined by determining a value indicating the individual reward and calculating the value indicating the individual reward based on a predetermined condition. Air conditioning by the air conditioning equipment that undergoes reinforcement learning based on a value indicating a group reward for the air conditioning for the plurality of persons located in the space and air conditioning control information when the air conditioning is performed by the air conditioning equipment. An air conditioning control unit that controls air conditioning using an air conditioning control model that controls
Information processing device. Using a productivity discrimination model learned to estimate information about the work productivity of each person from information about the amount of work of each person in a space where a plurality of people are present and air-conditioned by an air conditioning device and estimating information on the work productivity of the individual person before and after the air conditioning from the information on the work amount of the individual person before and after the air conditioning by the air conditioning equipment, and An individual reward for the air conditioning for the individual person based on a comparison of information about the work productivity of the individual person before air conditioning and information about the work productivity of the individual person after the air conditioning. A value indicating a group reward for the air conditioning for the plurality of persons in the space is determined by determining a value indicating the individual reward and calculating the value indicating the individual reward based on a predetermined condition. Air conditioning by the air conditioning equipment that undergoes reinforcement learning based on a value indicating a group reward for the air conditioning for the plurality of persons located in the space and air conditioning control information when the air conditioning is performed by the air conditioning equipment. An air conditioning control procedure for controlling air conditioning using an air conditioning control model that controls
An information processing program characterized by causing a computer to execute An information processing system including an information processing device and an air conditioner,
The information processing device is
Using a productivity discrimination model learned to estimate information about the work productivity of each person from information about the amount of work of each person in a space where a plurality of people are present and air-conditioned by an air conditioning device and estimating information on the work productivity of the individual person before and after the air conditioning from the information on the work amount of the individual person before and after the air conditioning by the air conditioning equipment, and An individual reward for the air conditioning for the individual person based on a comparison of information about the work productivity of the individual person before air conditioning and information about the work productivity of the individual person after the air conditioning. A value indicating a group reward for the air conditioning for the plurality of persons in the space is determined by determining a value indicating the individual reward and calculating the value indicating the individual reward based on a predetermined condition. Air conditioning by the air conditioning equipment that undergoes reinforcement learning based on a value indicating a group reward for the air conditioning for the plurality of persons located in the space and air conditioning control information when the air conditioning is performed by the air conditioning equipment. An air conditioning control unit that controls air conditioning using an air conditioning control model that controls
with
The air conditioner is
performing air conditioning according to the air conditioning control information output from the air conditioning control unit;
Information processing system.

Images