Worker physical condition prediction system and program

Abstract

To provide a worker physical condition prediction system and a program capable of flexibly predicting a physical condition according to a situation of a worker.SOLUTION: A worker physical condition prediction system according to an aspect of the present disclosure includes an acquisition unit configured to acquire sensor information detected by a sensor attached to a worker at a construction site, and an output unit configured to output physical condition prediction information based on the sensor information acquired by the acquisition unit by referring to a prediction model learned using input data including the sensor information and output data including the physical condition prediction information indicating a physical condition of the worker as learning data.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a worker physical condition prediction system and program. [Background technology]

[0002] In recent years, there has been a demand for managing the physical condition of workers at construction sites based on information acquired from various sensors. As a technology for managing the physical condition of workers, for example, Patent Document 1 discloses a biological information measuring device.

[0003] Patent Document 1 discloses a bioinformation measuring device that is worn on the upper arm of a human body, the bioinformation measuring device comprising: a belt that is wrapped around the upper arm; electrocardiogram measuring means having a plurality of electrodes for detecting the electrocardiogram signal of the human body; pulse wave measuring means having a pulse wave sensor for detecting the pulse wave of the human body; heartbeat vibration measuring means having a vibration sensor for detecting vibrations caused by the beating of the human heart; and an analysis processing unit that calculates the pre-ejection period and pulse wave propagation time of the heart based on time series data of the electrocardiogram signal, time series data of the pulse wave, and time series data of vibrations caused by the beating of the heart. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Publication No. 2023-23136 Summary of the Invention [Problem to be solved by the invention]

[0005] However, the biological information measuring device disclosed in Patent Document 1 is premised on predicting the worker's physical condition uniquely from information acquired from a sensor, and is not intended to flexibly predict the worker's physical condition according to the worker's situation. Therefore, the biological information measuring device disclosed in Patent Document 1 has a problem in that it cannot flexibly predict the worker's physical condition according to the worker's situation.

[0006] The present invention was devised in consideration of the above-mentioned problems, and its purpose is to provide a worker physical condition prediction system and program that can flexibly predict the physical condition of a worker depending on the situation of the worker. [Means for solving the problem]

[0007] The worker physical condition prediction system according to the first invention is characterized by comprising an acquisition means for acquiring sensor information detected by a sensor attached to a worker at a construction site, and an output means for outputting physical condition prediction information based on the sensor information acquired by the acquisition means, by referring to a prediction model trained using input data including the sensor information and output data including physical condition prediction information indicating the worker's physical condition as learning data.

[0008] The worker physical condition prediction system according to the second invention is the first invention, wherein the acquisition means acquires at least one of worker information about the worker or environmental information about the environment of the construction site, and the output means references the prediction model trained using sensor information, information including at least one of worker information and environmental information, input data, and the output data as training data, and outputs the physical condition prediction information based on the sensor information acquired by the acquisition means and at least one of worker information and environmental information.

[0009] The worker physical condition prediction system according to the third invention is the system of the first invention, characterized in that the acquisition means acquires two or more types of sensor information detected by two or more types of sensors attached to the worker, and the output means outputs the physical condition prediction information based on the two or more types of sensor information acquired by the acquisition means.

[0010] The worker physical condition prediction system according to a fourth aspect of the present invention is the third aspect, wherein the acquisition means acquires, in accordance with the two or more types of sensor information acquired, status information indicating the worker's status that is preset for each of the two or more types of sensor information, and the output means references the prediction model trained using input data including the sensor information and status information and the output data as training data, and outputs the physical condition prediction information based on the two or more types of sensor information and status information acquired by the acquisition means.

[0011] The worker health condition prediction system according to the fifth invention is characterized in that, in the first invention, it further comprises a calculation means for calculating risk information indicating the risk of the worker's health condition based on the health condition prediction information output by the output means.

[0012] The worker physical condition prediction program according to the sixth aspect of the present invention is characterized in that it causes a computer to execute an acquisition step of acquiring sensor information detected by a sensor attached to a worker at a construction site, and an output step of outputting health condition prediction information based on the sensor information acquired by the acquisition step, by referring to a prediction model trained using input data including the sensor information and output data including health condition prediction information indicating the worker's physical condition as learning data. [Effects of the Invention]

[0013] According to the first to sixth aspects of the present invention, the worker's physical condition prediction system and program of the present invention refer to a prediction model and output physical condition prediction information based on sensor information. This makes it possible to flexibly predict the physical condition of a worker using a trained model according to the situation of the worker.

[0014] In particular, according to the second aspect of the present invention, the worker physical condition prediction system outputs physical condition prediction information based on sensor information and information including at least one of worker information and environmental information, which makes it possible to flexibly predict the physical condition of a worker by taking into account the worker's condition and the environment of the construction site.

[0015] In particular, according to the third aspect of the present invention, the worker's physical condition prediction system outputs physical condition prediction information based on two or more types of sensor information. This allows the system to predict the worker's physical condition by taking into account multiple pieces of sensor information acquired from multiple sensors. This makes it possible to predict the worker's physical condition with higher accuracy.

[0016] In particular, according to the fourth aspect of the present invention, the worker physical condition prediction system outputs physical condition prediction information based on two or more types of sensor information and situation information. This makes it possible to obtain situation information, such as whether a worker is working while wearing heavy equipment, depending on the sensor that obtained the sensor information. This makes it possible to predict the worker's physical condition more flexibly depending on the worker's situation.

[0017] In particular, according to the fifth aspect of the present invention, the worker physical condition prediction system calculates risk information based on predicted physical condition information, thereby making it possible to calculate the risk of the worker's physical condition according to the predicted physical condition of the worker. [Brief explanation of the drawings]

[0018] [Figure 1] FIG. 1 is a diagram showing the configuration of a worker physical condition prediction system to which this embodiment is applied. [Figure 2] FIG. 2 is a diagram showing the configuration of a worker's physical condition prediction device to which this embodiment is applied. [Figure 3] FIG. 3 is a diagram showing the functions of the worker physical condition prediction device to which this embodiment is applied. [Figure 4] FIG. 4 is a flowchart showing the operation of the worker physical condition prediction system to which this embodiment is applied. [Figure 5]FIG. 5 is a diagram illustrating inputs and outputs of a prediction model. [Figure 6] FIG. 6 is a diagram showing input and output of a prediction model when worker information and environmental information are used. [Figure 7] FIG. 7 is a diagram showing input and output of a prediction model when situation information is used. DETAILED DESCRIPTION OF THE INVENTION

[0019] An example of a worker physical condition prediction system according to an embodiment of the present invention will be described below with reference to the drawings.

[0020] 1 is a schematic diagram showing an example of the configuration of a worker physical condition prediction system 100 according to the first embodiment. As shown in FIG. 1, the worker physical condition prediction system 100 includes a worker physical condition prediction device 1, a server 3, a user terminal 2, and a sensor 5, which are connected via a public communication network 4.

[0021] The server 3 is a storage medium that stores various pieces of information transmitted from the worker physical condition prediction device 1, the user terminal 2, and the sensor 5. The server 3 also transmits the stored information to the worker physical condition prediction device 1 and the user terminal 2 as necessary. The server 3 may, for example, have at least some of the functions of the worker physical condition prediction device 1, and may, for example, perform at least some of the processing in place of the worker physical condition prediction device 1.

[0022] The public communication network 4 is, for example, an internet network to which the worker physical condition prediction device 1 is connected via a communication circuit. The public communication network 4 may be configured as a so-called optical fiber communication network. Furthermore, the public communication network 4 may be realized by known communication technologies such as a wired communication network or a wireless communication network.

[0023] The user terminal 2 is owned by a user, such as a worker 20, who receives a service using the worker physical condition prediction system 100, and is connected to the worker physical condition prediction device 1 via a public communication network 4. The user terminal 2 may represent, for example, an electronic device that generates a database. The user terminal 2 may be, for example, a personal computer, a tablet terminal, or other electronic device. The user terminal 2 may also acquire worker information about the worker 20 input by the user and transmit the information to the worker physical condition prediction device 1 via the public communication network 4. The user terminal 2 may have at least some of the functions of the worker physical condition prediction device 1, for example. The user terminal 2 may have a display or speaker (not shown) that can present information to the user.

[0024] The sensor 5 is a sensor installed at a construction site or the like. The sensor 5 may be a sensor that measures vital signs of the worker 20, such as blood pressure, blood oxygen saturation, pulse, heart rate, respiration, oxygen intake, skin and core body temperature, blood sugar level, and pulse variability. The sensor 5 may also be a sensor that measures the number of steps taken by the worker 20, calories burned, location, and sleep time of the worker 20. The sensor 5 may be, for example, a helmet-type sensor 21, a shirt-type sensor 22, a ring-type sensor 23, a wristwatch-type sensor 24, or the like, attached to the worker 20 at the construction site. The sensor 5 attached to the worker 20 may be two or more types, such as the helmet-type sensor 21, the shirt-type sensor 22, the ring-type sensor 23, and the wristwatch-type sensor 24.

[0025] The sensor 5 may also be a sensor that measures environmental information indicating the environment of the construction site, such as temperature, humidity, WBGT (Wet Bulb Globe Temperature), wind speed, illuminance, air pressure, amount of precipitation, amount of snowfall, etc.

[0026] The worker physical condition prediction device 1 outputs physical condition prediction information based on sensor information. The worker physical condition prediction device 1 may be an electronic device such as a personal computer (PC), or may be an electronic device such as a smartphone, a tablet terminal, a wearable terminal, or an IoT (Internet of Things) device, or a single-board computer such as Raspberry Pi (registered trademark).

[0027] Next, an example of the worker physical condition prediction device 1 in the first embodiment will be described with reference to Figures 2 and 3. Figure 2 is a schematic diagram showing an example of the configuration of the worker physical condition prediction device 1 in the first embodiment, and Figure 3 is a schematic diagram showing an example of the function of the worker physical condition prediction device 1 in the first embodiment.

[0028] 2, the worker physical condition prediction device 1 includes a housing 10, a CPU (Central Processing Unit) 101, a ROM (Read Only Memory) 102, a RAM (Random Access Memory) 103, a storage unit 104, and I / Fs 105 to 107. The CPU 101, ROM 102, RAM 103, storage unit 104, and I / Fs 105 to 107 are connected via an internal bus 110.

[0029] The CPU 101 controls the entire worker physical condition prediction device 1. The ROM 102 stores operation code for the CPU 101. The RAM 103 is a work area used when the CPU 101 is operating. The storage unit 104 stores various information such as sensor information and prediction models. The storage unit 104 may be, for example, a hard disk drive (HDD), a solid state drive (SSD), an SD card, a miniSD card, or other data storage device. Note that the worker physical condition prediction device 1 may also include a graphics processing unit (GPU), not shown, for example.

[0030] The I / F 105 is an interface for transmitting and receiving various types of information via the public communication network 4. The I / F 106 is an interface for transmitting and receiving information to and from the input unit 108. For example, a keyboard is used as the input unit 108, and a user or the like using the worker physical condition prediction device 1 inputs various types of information or control commands for the worker physical condition prediction device 1 via the input unit 108. The I / F 107 is an interface for transmitting and receiving various types of information to and from the display unit 109. The display unit 109 outputs various types of information such as deterioration information stored in the storage unit 104, or the processing status of the worker physical condition prediction device 1. A display is used as the display unit 109, and may be, for example, a touch panel type.

[0031] 3, the worker physical condition prediction device 1 includes an acquisition unit 11, an extraction unit 12 connected to the acquisition unit 11, an output unit 13 connected to the acquisition unit 11 and the extraction unit 12, a calculation unit 14 connected to the output unit 13, and a storage unit 15 connected to each function. Note that these various components are realized, for example, by a CPU 101 using a RAM 103 as a work area to execute a program stored in a storage unit 104 or the like, and may be controlled by, for example, artificial intelligence.

[0032] The acquisition unit 11 acquires various types of information. The acquisition unit 11 acquires, for example, sensor information measured by the sensor 5, environmental information, and worker information. The acquisition unit 11 may also acquire worker information from, for example, the user terminal 2 via the public communication network 4. The acquisition unit 11 may output the acquired various types of information to the extraction unit 12, the output unit 13, etc.

[0033] The extraction unit 12 extracts situation information indicating the situation of the worker 20 based on the sensor information output from the acquisition unit 11. The extraction unit 12 may refer to a database in which sensor information and situation information are stored in association with each other, for example, stored in the storage unit 15, and extract the situation information based on the sensor 5 that measured the sensor information. The extraction unit 12 outputs the extracted situation information to the output unit 13.

[0034] The output unit 13 outputs health condition prediction information indicating the health condition of the worker 20 based on the various output information. The output unit 13 outputs the health condition prediction information based on the sensor information, for example, by referring to a prediction model stored in the memory unit 15. The output unit 13 outputs the output health condition prediction information to the calculation unit 14.

[0035] The calculation unit 14 calculates risk information indicating the risk of the worker's 20 physical condition based on the physical condition prediction information output from the output unit 13.

[0036] Next, the operation of the worker's physical condition prediction system 100 to which this embodiment of the present invention is applied will be described. Fig. 4 is a flowchart showing the operation of the worker's physical condition prediction system 100 to which this embodiment is applied.

[0037] First, in step S11, the acquisition unit 11 acquires various types of information. The acquisition unit 11 acquires, for example, sensor information measured by a sensor 5 via the public communication network 4. The sensor information is information detected by a sensor 5 attached to a worker 20 at a construction site. The sensor information is information indicating vital signs of the worker 20 detected by the sensor 5, such as blood pressure, blood oxygen saturation, pulse, heart rate, respiration, oxygen intake, skin and core body temperature, blood glucose level, and pulse variability. The sensor information may also be information indicating changes in the vital signs of the worker 20 over time. The sensor information may also be information such as the number of steps taken by the worker 20, calories burned, location, and sleep time. The sensor information may also be identification information identifying the sensor 5 that measured the vital signs of the worker 20. The identification information may be, for example, information such as the name, ID, model number, and type of the sensor 5 that measured the vital signs. The type of the sensor 5 may also be, for example, a helmet-type sensor 21, a shirt-type sensor 22, a ring-type sensor 23, or a wristwatch-type sensor 24.

[0038] Furthermore, in step S11, the acquisition unit 11 may acquire two or more types of sensor information detected by two or more types of sensors 5. In such a case, for example, the acquisition unit 11 may acquire sensor information detected by a shirt-type sensor 22 worn by the worker 20 and sensor information detected by a helmet-type sensor 21.

[0039] Furthermore, in step S11, the acquisition unit 11 may acquire environmental information. The environmental information is information indicating the environment of the construction site, and may be information indicating, for example, the temperature, humidity, WBGT (Wet Bulb Globe Temperature), wind speed, illuminance, atmospheric pressure, amount of precipitation, amount of snowfall, etc. of the construction site. The environmental information may also be information indicating the weather. In step S11, the acquisition unit 11 may acquire environmental information detected by the sensor 5. In addition, in step S11, the acquisition unit 11 may acquire environmental information transmitted from any terminal via the public communication network 4.

[0040] In step S11, the acquiring unit 11 may also acquire worker information. The worker information is information about the worker 20. The worker information is qualitative data about the worker 20, and may be, for example, information about the work of the worker 20, such as a work review sheet filled out by the worker 20. The worker information may also be a record of the worker 20's chronic illnesses or past accidents. The worker information may also be, for example, a record of the worker 20's health check input in advance by the worker 20. The worker information may also be information indicating whether the worker 20 is sleep-deprived, whether they have a hangover, the amount of alcohol they drink, the amount of tobacco they smoke, etc. The worker information may also be information indicating the length of the work, the severity of the work, etc. The worker information may also be information indicating whether the worker 20 is afraid of heights, etc. The worker information may also be information indicating the worker 20's vital signs measured in advance. In step S11, the acquiring unit 11 may acquire worker information transmitted from an arbitrary terminal via the public communication network 4.

[0041] Next, in step S12, the extraction unit 12 extracts situation information based on the sensor information acquired in step S11. The situation information is information indicating the situation of the worker 20. The situation information may be, for example, information indicating the equipment of the worker 20. The situation information may be, for example, information indicating that the worker 20 is equipped with a specific shirt-type sensor 22. The situation information may be, for example, information indicating that the worker 20 is wearing heavy equipment or that the worker 20's surface is covered with a mask or the like. The situation information may also be information indicating the work the worker 20 is performing while wearing specific equipment. The extraction unit 12 extracts situation information from information on the sensor 5 that measured the vital signs of the worker 20 indicated by the sensor information. In this case, the extraction unit 12 refers to a database in which situation information is stored linked to each sensor 5, and extracts situation information from identification information of the sensor 5 that measured the vital signs of the worker 20 indicated by the sensor information.

[0042] The database stores sensor information and situation information in association with each other. The database stores, for example, identification information such as the name, ID, model number, and type of the sensor 5 in association with situation information. The database may store, for example, the shirt-type sensor 22 in association with situation information indicating that the worker 20 is wearing heavy equipment. The database may also store, for example, multiple pieces of sensor information in association with one piece of situation information. The database may also store, for example, information on the position of the worker 20 indicated by the sensor information in association with situation information.

[0043] The extraction unit 12 refers to the database and extracts situation information from the information of the sensor 5 that measures the vital signs of the worker 20 indicated by the sensor information. This allows the extraction unit 12 to refer to the database and acquire situation information indicating that the worker 20 is wearing heavy equipment, for example, from the sensor information detected by the shirt-type sensor 22 acquired in step S11. Furthermore, the extraction unit 12 refers to the database and acquires situation information indicating that the worker 20 is working at a height, for example, from the sensor information detected by the shirt-type sensor 22 with a safety belt or full harness acquired in step S11. This makes it possible to automatically extract information indicating the status of the equipment, etc. of the worker 20 from the sensor information.

[0044] The extraction unit 12 may also extract one piece of situation information from, for example, multiple pieces of sensor information. In this case, the extraction unit 12 may refer to a database from sensor information including, for example, identification information indicating the shirt-type sensor 22 attached to the dustproof suit and the helmet-type sensor 21 attached to the protective mask, and extract situation information indicating that the worker 20 is performing precision work. The extraction unit 12 may also refer to a database from sensor information including, for example, identification information indicating the shirt-type sensor 22 with a tool and the helmet-type sensor 21 with a breathing apparatus, and extract situation information indicating that the worker 20 is performing maintenance or dismantling work on heavy machinery with a tool while receiving a supply of exhalation gas.

[0045] Next, in step S13, the output unit 13 refers to a prediction model trained using input data including sensor information and output data including physical condition prediction information as training data, and outputs physical condition prediction information based on the sensor information acquired in step S11. The physical condition prediction information is information indicating the physical condition of the worker 20. The physical condition prediction information may be information indicating, for example, good, normal, or poor physical condition. The physical condition prediction information may also be information indicating the degree of physical condition as a percentage. The physical condition prediction information may also be information indicating the signs of a specific disease such as decompression sickness, heat stroke, or myocardial infarction. The physical condition prediction information may also be information indicating stress caused by the work of the worker 20.

[0046] The prediction model is a trained model that has been trained using, for example, sensor information and health condition prediction information as training data.

[0047] As a method for generating a predictive model, for example, machine learning based on a neural network model may be used to generate the predictive model. The predictive model may also be a multimodal model. The predictive model may be trained using machine learning based on a neural network model such as a convolution neural network (CNN), or any other model may be used. Furthermore, as a method for generating a predictive model, for example, retrieval-augmented generation (RAG), sequence-to-sequence (seq2seq) linear discriminant analysis, support vector machines, k-nearest neighbors, random forests, deep learning, etc. may be used to generate the predictive model.

[0048] In such a case, the prediction model stores learning data consisting of sensor information as input data and health condition prediction information as output data, as shown in FIG. 5. In such a case, morphemes or words contained in the sensor information may be used as input data, and morphemes or words contained in the health condition prediction information may be used as output data. The prediction model uses the learning data to learn the relationships between the learning data, the degree of connection, weight variables or propagation coefficients of neural network layers, etc. Furthermore, the sensor information and health condition prediction information used for learning may be, for example, sensor information and health condition prediction information for use in previously acquired learning data, but are not limited to this, and information acquired at any time may also be used.

[0049] For example, the predictive model is updated as appropriate during the machine learning process, and a classifier is trained using a function optimized based on, for example, multiple input data and multiple output data. This makes it possible to select output data from multiple perspectives for the input data. Furthermore, the input data and output data are not limited to these, and any other type of information may also be used. Furthermore, the input data may be, for example, text information and / or image information included in sensor information.

[0050] The predictive model may also be composed of neural network nodes in artificial intelligence. That is, the neural network nodes learn weighting coefficients for the outputs. Furthermore, the predictive model is not limited to a neural network, and may be composed of any decision-making factors that make up artificial intelligence.

[0051] Furthermore, the prediction model may be machine-learned by providing at least one hidden layer between the input data and the output data. Weighting is set for either the input data or the hidden layer data, or for both, and output selection is based on this weighting. If this weighting exceeds a certain threshold, the output may be selected.

[0052] Such learning data is learned in advance, and in step S13, the output unit 13 actually references the prediction model learned using the sensor information and the physical condition prediction information as learning data, and outputs physical condition prediction information based on the sensor information acquired in step S11. When outputting, for example, the previously acquired weighting is referenced. For example, if the newly acquired sensor information is identical to or similar to "sensor information A," it is associated with "physical condition prediction information A" via weighting, with weight AA being "73%," and with weight AB being "12%" with physical condition prediction information B. In this case, "sensor information A," which has the highest weighting, is selected as the optimal solution. However, it is not essential to select the information with the highest weighting as the optimal solution; "physical condition prediction information B" may also be selected as the optimal solution.

[0053] By referring to such a prediction model, it is possible to quantitatively select output data that is suitable for input data, not only when the sensor information is identical to or similar to the input data, but also when the sensor information is dissimilar.

[0054] In step S13, the output unit 13 may refer to a prediction model trained using the sensor information and information including at least one of worker information and environmental information, the input data, and output data including health condition prediction information as training data, and output health condition prediction information based on the sensor information acquired in step S11 and information including at least one of worker information and environmental information. In this case, the output unit 13 may refer to a prediction model generated using training data in which the input data is information including at least one of sensor information and worker information and environmental information, as shown in Fig. 6, and the output data is health condition prediction information.

[0055] Furthermore, in step S13, the output unit 13 may refer to a prediction model and output health condition prediction information based on two or more types of sensor information acquired in step S11. In such a case, the output unit 13 may refer to a prediction model generated using input data including two or more types of sensor information and output data including health condition prediction information as training data.

[0056] Furthermore, in step S13, the output unit 13 may refer to a prediction model trained using input data including sensor information and situation information and output data including health condition prediction information as training data, and output health condition prediction information based on the sensor information and situation information acquired in step S11. In this case, the output unit 13 may refer to a prediction model generated using training data in which the input data is the sensor information and situation information and the output data is the health condition prediction information, as shown in Fig. 7.

[0057] Next, in step S14, the calculation unit 14 calculates risk information indicating the risk of the worker 20's physical condition based on the physical condition prediction information output in step S13. The risk information is information indicating the possibility of developing a specific disease, such as decompression sickness, heat stroke, or myocardial infarction. The risk information may be information indicating the possibility of developing a specific disease, for example, as a percentage. In step S14, the calculation unit 14 may calculate the risk information based on the physical condition prediction information output in step S13, for example, by using a correspondence relationship that associates previously acquired physical condition prediction information with risk information and that is stored. Furthermore, the calculation unit 14 may calculate the risk information based on the physical condition prediction information output in step S13, for example, by referring to a risk model that has been trained using input data including the physical condition prediction information and output data including the risk information as training data.

[0058] This completes the operation of the worker physical condition prediction system 100 in this embodiment. The worker physical condition prediction system 100 of the present invention references a prediction model and outputs physical condition prediction information based on sensor information. This makes it possible to flexibly predict the physical condition of the worker 20 using the trained model according to the situation of the worker 20.

[0059] Although the embodiments of the present invention have been described, these embodiments are presented as examples and are not intended to limit the scope of the invention. These novel embodiments can be embodied in various other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their modifications are included within the scope and spirit of the invention, and are also included in the inventions and their equivalents as defined in the claims. [Explanation of symbols]

[0060] 1. Worker health condition prediction device 2. User terminal 3 Server 4 Public communication network 5 sensors 10. Cabinet 11 Acquisition Department 12 Extraction part 13 Output section 14 Calculation section 15 Storage section 20 workers 21 Helmet-type sensor 22 Shirt-type sensor 23 Ring-type sensor 24 Wristwatch-type sensor 100 Worker health prediction system 101 CPU 102 ROM 103 RAM 104 Preservation Department 105 I / F 106 Interface 107 Interface 108 Input section 109 Display section 110 Internal Bus

Claims

1. an acquisition means for acquiring sensor information detected by a sensor attached to a worker at a construction site; and an output means for outputting health condition prediction information based on the sensor information acquired by the acquisition means, by referring to a prediction model trained using input data including sensor information and output data including health condition prediction information indicating the worker's health condition as training data. A worker health condition prediction system characterized by the above.

2. The acquisition means acquires at least one of worker information related to the worker or environmental information related to the environment of the construction site, The output means refers to the prediction model trained using information including sensor information, and at least one of worker information and environmental information, input data, and the output data as training data, and outputs the health condition prediction information based on the sensor information acquired by the acquisition means, and at least one of worker information and environmental information. The worker physical condition prediction system according to claim 1,

3. the acquiring means acquires two or more types of sensor information detected by two or more types of sensors attached to the worker, The output means outputs the health condition prediction information based on two or more types of sensor information acquired by the acquisition means. The worker physical condition prediction system according to claim 1,

4. the acquiring means acquires, in accordance with the acquired two or more types of sensor information, status information indicating a status of the worker that is preset for each of the sensor information; The output means refers to the prediction model trained using input data including sensor information and situation information and the output data as training data, and outputs the physical condition prediction information based on two or more types of sensor information and situation information acquired by the acquisition means. The worker physical condition prediction system according to claim 3,

5. Further, a calculation means is provided for calculating risk information indicating the risk of the worker's physical condition based on the physical condition prediction information output by the output means. The worker physical condition prediction system according to claim 1,

6. an acquisition step of acquiring sensor information detected by a sensor attached to a worker at a construction site; and an output step of outputting health condition prediction information based on the sensor information acquired in the acquisition step, by referring to a prediction model trained using input data including sensor information and output data including health condition prediction information indicating the worker's health condition as training data. A worker health prediction program that features:

Images