JP7777832B2 - Biometric information processing method and biological information processing system - Google Patents

Description

This application relates to a biometric information processing method and a biometric information processing system that evaluates the physical condition of a person being evaluated based on biometric information obtained from the person being evaluated.

In recent years, with the development of internet connection environments such as wireless LAN, the development of means that allow for short-range information transmission such as Bluetooth (registered trademark), and the widespread use of high-performance mobile devices such as smartphones and small sensor devices that can measure physical data such as body temperature, heart rate, and sweat rate, evaluation systems that evaluate the physical condition of the person being evaluated based on the biometric information obtained by sensor devices, and health management systems that manage the health of the person being evaluated based on the evaluation results and reduce the risk of developing heatstroke, which has become a problem in recent years, have been put into practical use.

As an example of a biometric information processing system that performs such physical condition evaluation and management, a heatstroke risk management system has been proposed that uses a wearable biometric signal detection device equipped with a three-dimensional acceleration sensor that captures the subject's physical movements and a biometric information acquisition unit that detects heart rate to constantly evaluate the subject's risk of developing heatstroke and enable measures to be taken to reduce that risk (see Patent Document 1).

JP 2018-130531 A

In the above-mentioned conventional heatstroke risk management system, a biometric information acquisition unit is placed on the chest of the undershirt. The biometric information acquired by the biometric information acquisition unit is equipped with an electrometer that detects the wearer's heart rate, a three-dimensional acceleration sensor that can detect the wearer's body movements, and a temperature sensor that can detect the temperature inside the clothing. The biometric information acquired by the biometric information acquisition unit is sent to an information processing unit on a cloud server on the Internet via a communication device such as a smartphone carried by the person being measured. The information processing unit assesses the risk of heatstroke for each individual, or for a group of workers working in the same environment, and reduces the risk of heatstroke by instructing workers at high risk to take a break.

Biometric information acquisition units equipped with sensors that detect the subject's pulse, body movements, body temperature, etc. include wearable clothing types, such as those used in the above-mentioned conventional heatstroke risk management systems, in which an electrometer or the like is attached to the subject's body using a shirt, etc. In addition, various types have been proposed that differ in the method of acquiring biometric information, device shape, and wearing method, such as wristwatch-type types that detect pulse using vibration or optical methods, which are easy to put on and take off and reduce the discomfort of wearing a biometric information acquisition unit, and types that are worn on the earlobe or fingertip and optically detect the subject's pulse.

Due to the constraints of their configuration, these various biometric information acquisition units differ in the type of biometric information acquired and its accuracy. Furthermore, each of these biometric information acquisition units is incorporated into a unique biometric information processing system, with different settings for the frequency with which biometric data is detected and the frequency with which data is transmitted. Furthermore, due to differences in the shape and attachment location of each biometric information acquisition unit, the data transfer capabilities within the system employed, and the data processing functions incorporated within the biometric information acquisition unit, the data processing specifications, such as the extent to which data is processed before the acquired data is sent from the biometric information acquisition unit, also differ. For this reason, in conventional biometric information processing systems that use biometric information acquisition units, the biometric information acquisition unit worn by the person being evaluated was limited to one specific to that system.

However, in a biometric information processing system, collecting biometric information from more subjects improves the accuracy of evaluation results, such as physical condition assessments based on the biometric information. Increasing the amount of acquired biometric information in this way improves the accuracy of evaluation results, allows for the construction of a more cost-effective biometric information processing system, and allows for acquisition of biometric information in a more appropriate form depending on the subject's environment and the type and intensity of the subject's movements, such as the type of exercise or work, and furthermore aims to acquire biometric information under less stressful conditions according to the preferences of each subject. Therefore, it is preferable to have a biometric information processing system that can use biometric information from a variety of different types of biometric information acquisition units.

The present application aims to solve the problems associated with the above-mentioned prior art, and to provide a biometric information processing method capable of processing data from multiple forms as a biometric information acquisition unit that acquires biometric information of the subject, as well as a biometric information processing system that can employ multiple forms of biometric information acquisition units.

In order to solve the above problem, the biometric information processing method disclosed in this application comprises the steps of acquiring biometric information of the person being evaluated using a biometric information acquisition unit, and calculating an index representing the state of the person being evaluated based on the acquired biometric information, and is characterized in that the biometric information acquired by two or more different types of biometric information acquisition units is converted into a common parameter to calculate the index.

The biometric information processing system disclosed in this application also includes two or more different types of biometric information acquisition units that acquire biometric information of the person being evaluated, and a data processing unit that calculates an index representing the state of the person being evaluated based on the acquired biometric information, and the data processing unit converts the biometric information acquired by the biometric information acquisition units into common parameters to calculate the index.

With the above configuration, the biometric information processing method disclosed in this application uses the biometric information of the person being evaluated acquired by two or more types of biometric information acquisition units to calculate various indices that represent the state of the person being evaluated. In other words, by converting the biometric information acquired from different types of biometric information acquisition units into common parameters, differences in data accuracy and data specifications can be absorbed. As a result, it is possible to calculate indices that represent the state of the person being evaluated with high accuracy using the biometric information of many people being evaluated, and it is also possible for the person being evaluated to select the type of biometric information acquisition unit that is more preferred.

Furthermore, with the above configuration, the biometric information processing system disclosed in this application has a wider range of options for the biometric information processing units that can be adopted in the system, making it possible to realize a biometric information processing system that is lower cost, has high accuracy, and is more easily accepted by the person being evaluated.

FIG. 1 is a block diagram showing the configuration of each part of a heatstroke risk management system described as an embodiment. FIG. 2 is a diagram illustrating the configuration of a first biological information acquisition unit used in the heatstroke risk management system described in this embodiment. FIG. 3 is a diagram illustrating a second biological information acquisition unit used in the heatstroke risk management system described in this embodiment. FIG. 4 is a diagram illustrating a third biological information acquisition unit used in the heatstroke risk management system described in this embodiment. FIG. 5 is a flowchart illustrating a method for assessing the risk of developing heatstroke in the heatstroke risk management system described in this embodiment. 6A and 6B show the results of measuring standard heart rate responses, which indicate heart rate responses to acceleration deviations. Fig. 6A shows a plot of all approximately 3 million data points, and Fig. 6B shows the standard heart rate response obtained using the median value of the heart rate responses to acceleration data. FIG. 7 is a diagram for explaining a method for estimating a heart rate index and a physical fitness index in the biological information processing method according to this embodiment. FIG. 8 is a diagram illustrating a first correction map for determining a work strain index in the work strain estimation method described in this embodiment. FIG. 9 is a diagram illustrating a second correction map for determining a work strain index in the work strain estimation method described in this embodiment.

The biometric information processing method disclosed in this application comprises the steps of acquiring biometric information of the person being evaluated using a biometric information acquisition unit, and calculating an index representing the state of the person being evaluated based on the acquired biometric information, and converting the biometric information acquired by two or more different types of biometric information acquisition units into common parameters to calculate the index.

By having the above configuration, the biometric information processing method disclosed in this application allows for a wider range of options for the biometric information acquisition unit worn by the person being evaluated, making it possible to acquire more biometric information from the person being evaluated in a manner that places less burden on the person being evaluated, and to calculate accurate indicators using the acquired biometric information.

In the above biometric information processing method, it is preferable that the index is at least one of a work burden index indicating the degree of impact of the workload on the person being evaluated, a physical condition assessment index indicating the degree of change from the normal state of the person being evaluated's physical condition, and a heatstroke risk index indicating the degree of risk of the person being evaluated developing heatstroke. In this way, it is possible to obtain practical indexes that represent the condition of many people being evaluated.

It is also preferable that the biometric information is at least one of heart rate data, acceleration data, and METs (estimated metabolic equivalents). This makes it possible to correct for differences in individual characteristics and differences in the specifications of the biometric information acquisition unit, obtain the biometric information necessary to accurately assess the state of the person being evaluated, and accurately calculate an index representing the state of the person being evaluated. Note that heart rate data includes heart rate and heartbeat interval. The biometric information may also include an estimated value of energy expenditure.

Furthermore, it is preferable that the common parameter used for the heart rate data is the median heart rate. By doing so, accurate heart rate data for the subject can be obtained regardless of the heart rate data acquisition method or data processing specifications used by the biometric information acquisition unit.

Furthermore, it is preferable that the common parameter used in the acceleration data or the METs (estimated metabolic equivalents) is acceleration deviation. This allows the subject's movements to be accurately understood regardless of the method used to understand the subject's body movements or the data processing method used by the biometric information acquisition unit.

The biometric information processing system disclosed in this application comprises two or more different types of biometric information acquisition units that acquire biometric information of the person being evaluated, and a data processing unit that calculates an index representing the state of the person being evaluated based on the acquired biometric information, and is characterized in that the data processing unit converts the biometric information acquired by the biometric information acquisition units into common parameters and calculates the index.

By having the above configuration, the biometric information processing system disclosed in this application has a wider range of options for the biometric information acquisition unit, allowing the person being evaluated to select a biometric information acquisition unit that suits their preferences and the physical condition of their body during measurement. It is also possible to easily calculate indices representing the condition of a larger number of people being evaluated using their biometric information, thereby realizing a biometric information management system that can obtain accurate evaluation results at low cost.

Embodiments of the biometric information processing method and biometric information processing system disclosed in this application are described below with reference to the drawings.

(Embodiment)
[Overall system configuration]
First, the overall configuration of an example of a biological information processing system disclosed in the present application will be described.

This embodiment illustrates a heatstroke risk management system that evaluates a worker's physical condition and assesses and manages the risk of heatstroke based on a work strain index indicating the level of strain caused by work, based on the worker's movements, environmental temperature, heart rate, etc., and a thermal load index indicating the level of thermal load on the worker. The heatstroke risk management system of this embodiment is suitable for use with, for example, multiple workers working at a construction site as subjects of evaluation. It assesses the risk of heatstroke based on biometric information obtained from each worker and aims to reduce the risk of heatstroke at construction sites by issuing warnings to workers at increased risk of heatstroke and taking measures such as encouraging them to take appropriate breaks.

Figure 1 is a block diagram showing an example configuration of each component of a heatstroke risk management system according to this embodiment.

As shown in FIG. 1, the heatstroke risk management system of this embodiment is composed of a worker 10 who is being assessed; a cloud server 21 on the Internet 20 that evaluates the worker's 10 physical condition based on his or her biometric information and assesses the risk of developing heatstroke; a site supervisor 30 who is the manager who oversees the worker 10 who is being assessed and a work group that includes a certain number of workers 10; and an office 40 that manages multiple site supervisors 30, provides overall visibility, and operates and maintains the heatstroke risk assessment system. Note that the above is a generalized example assuming a typical construction site. Needless to say, the heatstroke risk management system of this embodiment can take various forms depending on the actual configuration of the site where it is being implemented, such as when one site supervisor 30 manages one worker 10, when the site supervisor and the office are inseparable, or when multiple offices are involved and the entire construction site is managed on a larger scale.

In the heatstroke risk management system described in this embodiment, each worker 10 wears a measuring device 11, which is a biometric information acquisition unit, capable of detecting at least one of the following biometric information: their own heart rate data, acceleration data indicating body movement, and clothing temperature data representing the ambient temperature. At least two types of measuring devices 11 are used.

For example, some workers wear undershirts with a biosensor 11a attached to the chest as a first measuring device, which includes a temperature sensor that detects the temperature inside the clothing, an electrometer that detects heart rate data, and a three-dimensional acceleration sensor that detects body movement. Another worker wears a wireless pulse meter 11b as a second measuring device, which includes a measuring unit that is attached to the earlobe to optically measure pulse, and a main unit that is attached to the clothing near the center of the body and has a built-in three-dimensional acceleration sensor. Another worker wears a wristwatch-type sensing unit 11c as a third measuring device, which includes a pulse sensor that can detect pulse as minute vibrations, a three-dimensional acceleration sensor, and sensors that detect temperature and humidity.

Note that the biometric information acquisition units that can be used in the heatstroke risk management system of this embodiment are not limited to those exemplified above, and various types of biometric information acquisition units can be used, including those equipped with heart rate data of the person being evaluated, an acceleration sensor that indicates body movement, a temperature sensor that measures environmental temperature, etc.

In addition, various means can be used to attach the biosensor 11a, which is the first biometric information acquisition unit used in the heatstroke risk management system of this embodiment, to the chest or center of the body of the worker 10. These include a belt method in which the biosensor 11a is secured using a belt, and a measurement patch method in which the biosensor 11a is placed on an adhesive sheet. The specific configuration of the measurement device 11, which is the biometric information acquisition unit, and details of the data processing of the biometric information data of the worker 10 acquired by each measurement device 11 will be described in detail later.

In the heatstroke risk management system according to this embodiment, each worker 10 carries a smartphone 12 as a mobile terminal. The measurement device 11 and the smartphone 12 carried by the worker 10 are constantly connected via short-range communication such as Bluetooth (registered trademark), and the biological information acquired by the measurement device 11 is sent to the smartphone 12 as needed.

The smartphone 12 is equipped with a data receiving unit 15 and a data transmitting unit 16, and is constantly connected to the Internet 20 as a network environment via a wireless LAN or a mobile phone information carrier. In this embodiment of the heatstroke risk management system, the smartphone 12 is linked to the identification data of each worker 10, and the smartphone 12 also has an assessee information transmitting unit 13, which uses the data transmission function of the smartphone 12 to transmit biometric information linked to the worker's identification information to a cloud server 21 located on the Internet 20.

Measuring devices worn by workers can be linked to IDs that identify the work itself in a variety of ways, including the worker entering the name and control number of the measuring device into their smartphone, using the smartphone's image recognition function to read a two-dimensional or three-dimensional identification code attached to the measuring device, using an identification code in short-range communication between the smartphone and the measuring device, or other methods selected by the worker using an application on their smartphone. Furthermore, if the smartphone itself is not the worker's personal property but is loaned out as part of system use, worker recognition can be achieved by entering data using the smartphone, reading an identification code, using a facial recognition system, or other methods.

In addition, since the smartphone 12 is capable of receiving data, emitting sound, and displaying images, the heatstroke risk management system of this embodiment is equipped with a warning notification unit 14 that notifies the worker 10 of the risk of developing heatstroke and urges them to take a break, and an image display unit 17 that displays in an easy-to-read manner the results of the heatstroke risk assessment for each worker 10 and the entire group to which that worker 10 belongs.

The cloud server 21 has an internal data receiving unit 23 and data transmitting unit 26, and transmits and receives information via the Internet 20. The cloud server 21 also has an evaluation and determination unit 22 as a data processing unit, which acquires biometric information data of all workers 10 who are the targets of the heatstroke risk management system, calculates a work strain index indicating the degree of impact of the workload on each worker 10, and a physical condition evaluation index indicating the extent to which the worker's physical condition has changed from their usual state, and calculates a heatstroke risk index indicating the degree of risk of each worker 10 developing heatstroke based on these indices. The evaluation and determination unit 22 can also manage the risk of heatstroke for groups of workers 10 formed based on commonalities in work content and working environment.

The heatstroke risk management system according to this embodiment manages the risk of heatstroke for each worker 10, and if the risk of heatstroke is determined to be particularly high, communicates this information to encourage the worker to take measures to reduce the risk of heatstroke. To this end, the cloud server 21 evaluates and determines the risk of heatstroke, and if the risk of heatstroke is increasing, creates warning information to warn the worker of this fact.

The cloud server 21 also has a weather information acquisition unit 25, which acquires weather information from information sites that provide weather information via the Internet 20, and can obtain current weather conditions such as temperature, humidity, and amount of sunlight in the area where the worker 10 is working, as well as a weather forecast that predicts changes over the next few hours, allowing weather conditions to be taken into account when assessing the risk of developing heatstroke.

Furthermore, the cloud server 21 is equipped with a data recording unit 24, which can chronologically record measurement data from each worker 10 registered in the heatstroke risk management system, the history of warning information creation, and other information. This makes it possible, for example, to manage the risk of heatstroke based on the results of each worker's 10 physical condition assessment up to the current day and the previous day, or to perform a more accurate risk assessment of heatstroke based on past heatstroke risk assessment results under similar weather conditions.

The cloud server 21 is connected via the Internet 20 to a personal computer 31 that serves as a manager's information terminal used by a site supervisor 30, who is the manager who supervises the work of the worker 10 being evaluated at the construction site. Therefore, the site supervisor 30, who is at the work site where the worker 10 is working, can use the data receiving unit 33 of the personal computer 31 to grasp the biometric information data of the worker 10 that is sent from the cloud server 21 at any time, as well as whether or not warning information has been generated by the evaluation determination unit 22.

The evaluation and judgment unit 22 of the cloud server 21 evaluates the physical condition of the worker 10 based on the heart rate data, acceleration data, and temperature data inside the clothing obtained from the measuring device 11 worn by the worker 10, and further calculates a work strain index and calculates a heatstroke risk index for the worker 10 by taking into account the temperature information inside the clothing and environmental temperature information at the work location obtained via the Internet.

The specific details of the estimation of the worker's 10 workload and the heatstroke risk assessment performed by the evaluation and determination unit 22 will be explained later.

The cloud server 21 corrects the assessment results for the individual worker's 10 risk of developing heatstroke based on historical data representing the past history information of the worker 10 being assessed, recorded in the data recording unit 24, weather information for the work area acquired by the weather information acquisition unit 25, and environmental information such as changes in various information acquired from workers other than the worker being assessed who work at the same site as the worker being assessed, thereby enabling more realistic management of the risk of developing heatstroke.

In the heatstroke risk management system exemplified in this embodiment, the evaluation and determination unit 22 is not limited to being provided on the cloud server 21. For example, the various functions of the cloud server 21 may be implemented on an administrator information terminal or a business management computer, and as long as the functions can be realized, the location or device on which the evaluation and determination unit is implemented does not matter.

The site supervisor's 30's computer 31 is equipped with an information management unit 32 that manages various information obtained by the measuring devices 11 regarding the workers 10 (including the workers 10) who belong to the work site supervised by the site supervisor 30, as well as whether warning information has been generated. Based on information transmitted from the cloud server 21, the information management unit 32 constantly keeps track of the latest information obtained from each worker 10 and information that serves as the basis for assessing the risk of heatstroke, such as whether warning information has been generated. The information management unit 32 also outputs the acquired assessment results of the heatstroke risk for each worker 10 and other environmental information to the display image processing unit 35, which then adjusts the screen content displayed on a display device 36, such as an LCD monitor.

In this way, the site supervisor 30 can grasp information about the workers 10 working at the work site he or she supervises, such as the risk of developing heatstroke, in a unified manner as a whole, or as detailed information about each individual worker, on an easy-to-read screen. Note that the specific screen content displayed on the display device 36 after processing by the display image processing unit 35 is not described in detail in this specification, as it is sufficient to display the information required by the appropriately configured system in an easy-to-read manner.

Furthermore, the site supervisor's 30's computer 31 can check whether the worker 10 has taken measures to prevent the onset of heatstroke by checking changes in the biometric information obtained from the worker 10 after notifying the warning information and by receiving confirmation of receipt of the warning information from the worker 10.If the worker 10 has not taken measures to prevent the onset of heatstroke, the site supervisor's 30 can further warn the worker 10 by repeatedly transmitting the warning information to the worker 10 in question.

In the above explanation, an example was given in which the evaluation and judgment unit 22 of the cloud server 21 generates warning information informing the worker 10 that the risk of developing heatstroke is increasing. However, the warning information can also be generated by the information management unit 32 installed on the computer 31 of the site supervisor 30. It is also possible to configure both the evaluation and judgment unit 22 and the information management unit 32 to generate warning information. In this way, the computer 31 of the site supervisor 30 who actually supervises the work site can generate warning information prior to the judgment results of the evaluation and judgment unit 22 and transmit it to the target worker 10, which may further reduce the risk of developing heatstroke depending on the actual conditions at the work site.

Warning information generated by the evaluation/determination unit 22 of the cloud server 21 or the computer 31 of the site supervisor 30 is transmitted from the data transmission unit 34 of the computer 31 of the site supervisor 30 to the smartphone 12 carried by the worker 10 via a network including a local network such as a wireless LAN and a mobile phone information carrier. The warning notification unit 14 of the smartphone 12 that receives the warning information uses various communication methods such as voice, screen display, illuminating or flashing lights, and vibration to notify the worker 10 that their risk of developing heatstroke is increasing. Upon confirming the warning information, the worker 10 reports receipt of the warning information via the touch panel or operation buttons of the smartphone 12, and takes measures to prevent heatstroke, such as stopping work and taking a rest.

The worker's 10 smartphone 12 sends a message to the supervisor's 30's computer 31 that the worker 10 has confirmed the warning information and stopped working, allowing the supervisor 30 to confirm that the worker 10 has taken measures to prevent the onset of heatstroke.

Furthermore, in the heatstroke risk management system described in this embodiment, heatstroke risk data for the work site, as determined by the site supervisor 30, is sent to the smartphone 12 of the worker 10, allowing the worker 10 to check the current heatstroke risk at the work site where he or she is working. For example, if it is confirmed that the risk of heatstroke for workers other than the worker 10 is increasing, each worker can take proactive measures to prevent the onset of heatstroke. Furthermore, if a worker learns that other workers have stopped working after receiving warning information about the risk of heatstroke, it is expected that they will respond more readily to the warning information sent to them by the site supervisor 30.

Furthermore, the smartphone 12 owned by the worker 10 can display on the screen related information such as changes in the worker's 10's current risk of developing heatstroke, changes in the worker's own heart rate acquired by the biosensor 11, changes in the physical condition assessment index calculated from acceleration data, and calories burned, for the worker 10 to refer to. Since the display screen on the smartphone owned by the worker 10 only needs to be able to clearly display the necessary information according to each purpose, a detailed description will be omitted in this specification.

The cloud server 21 is also connected via the Internet 20 to a management computer 41 within the company or business 40 to which the worker 10 belongs, and transmits, in real time, to the management computer 41 of the business 40, the measurement result information for the worker 10 sent to the computer 31 of the site supervisor 30, as well as various information used by the cloud server 21 to determine the risk of heatstroke. The management computer 41 of the business 40 is equipped with its own data receiving unit 42 and data transmitting unit 43, and is also connected via the Internet to the computer 31 of the site supervisor 30, and can confirm information such as whether warning information from the site supervisor 30 was correctly transmitted to the worker 10 and whether the worker 10 has taken preventative measures against heatstroke, and can issue specified instructions as necessary. This effectively supports efforts to prevent the risk of heatstroke for the worker 10.

In addition, because the cloud server 21, the site supervisor's 30's personal computer 31, and the business establishment's 40 management computer 40 are connected via the Internet 20 environment, the cloud server 21 can be accessed from the personal computer 31 or management computer 40, allowing the cloud server 21 to be controlled, the data processing content in the cloud server 21 to be updated, the evaluation program in the evaluation and judgment unit 22 to be updated, and information necessary for heatstroke prevention and management to be retrieved from the cloud server 21 as appropriate.

In the above explanation, a smartphone was used as an example of a mobile device carried by a worker, but the mobile device used by the worker is not limited to a smartphone. It can also be a mobile phone, tablet, or even a small, dedicated terminal device capable of sending and receiving information specifically designed for heatstroke risk management systems. Furthermore, the manager's information terminal operated by the site supervisor can be any type of information device capable of sending and receiving information over a network, displaying and recording data, such as a desktop computer, laptop computer, tablet computer, or small server device, in addition to the computer mentioned above.

Furthermore, in the above explanation, we have described a form in which warning information is sent from the site supervisor's manager information terminal to the worker's mobile device, but if the warning information is generated by the evaluation and judgment unit of the cloud server, the system can also be configured to send the warning information directly from the cloud server to the worker's mobile device.

Furthermore, the means of communication between workers, site supervisors, and the management department within the workplace are not limited to those exemplified above, and it goes without saying that various information communication means for sending and receiving data can be used.

[Biometric information acquisition unit]
(First measuring device)
First, a first embodiment of the measuring device 11 as a biometric information acquisition unit will be described, in which a biometric sensor is placed in close contact with the chest of the worker to be evaluated to acquire biometric information.

Figure 2 shows an example configuration of an undershirt worn by a worker in a heatstroke risk management system according to this embodiment, equipped with a biosensor, which is the first biometric information acquisition unit. Figure 2(a) shows the front side of the undershirt equipped with the biosensor, and Figure 2(b) shows the back side of the undershirt, i.e., the side that faces and comes into contact with the surface of the worker's body.

As shown in Figure 2, a biosensor 11a is placed on the chest of the undershirt 18 worn by the worker 10. More specifically, the biosensor 11a is composed of a data acquisition and transmission unit 11a1 placed in the center of the chest on the front surface 18a of the undershirt 18, and an electrode portion 11a2 connected to this data acquisition and transmission unit 11a1 and positioned extending in the left-right direction on the back surface 18b of the undershirt 18, i.e., the side that comes into contact with the skin.

In the heatstroke risk management system of this embodiment, a worker 10 wearing a first biometric information acquisition unit uses a biometric sensor 11a to detect the worker's 10 heart rate, temperature inside the clothing, and movement. An electrode serving as a heart rate detection means, located on the undershirt 18's back surface, contacts the worker's chest, enabling the worker's 10 heart rate to be detected from changes in surface potential. A temperature sensor (not shown) for detecting the temperature inside the clothing and an acceleration sensor chip (not shown) for detecting acceleration in three dimensions are housed within the data acquisition and transmission unit 11a1. As described above, in the heatstroke risk management system of this embodiment, the smartphone 12 carried by each worker is used as a repeater for the measured biometric information. Therefore, the data acquisition and transmission unit 11a1 need only include a minimum amount of data processing circuitry, a data transmission unit for short-range communication, and a power source for driving these electronic circuits, including the sensors. This allows the data acquisition and transmission unit 11a1 to be made smaller and lighter, thereby reducing the discomfort felt by the worker 10 wearing the undershirt 18 attached to their chest.

As mentioned above, in the heatstroke risk management system described in this embodiment, various methods are known for attaching the biosensor 11a, which contacts the chest or other area of the worker 10 to acquire heart rate data, temperature data inside the clothing, and acceleration data generated by movement. However, compared to methods such as attaching a measurement patch directly to the chest or using an elastic attachment belt, the method of attaching the biosensor 11a to the undershirt 18 worn by the worker 10 as shown in FIG. 2 allows the worker 10 to acquire necessary information without having to worry about wearing the biosensor 11a. Furthermore, even if the worker 10 sweats or twists his or her body while working, the biosensor 11a attached to the undershirt 18 will not ultimately detach from the worker 10's body surface, and its attachment position can be maintained substantially unchanged. Therefore, although some heart rate data may not be acquired the moment the biosensor 11a leaves the worker 10's chest, a situation in which no heart rate data is acquired for a prolonged period can be avoided.

The biosensor 11a for acquiring the worker's 10 heart rate data can be placed on the worker's waist, back, upper arms, or legs, in addition to the chest as described above. However, it goes without saying that it is preferable to limit the location to a range where the acceleration sensor and temperature sensor built into the biosensor 11a can acquire good measurement data. Furthermore, rather than using a system to manage the risk of heatstroke with workers working at a construction site as the subject of assessment, as described in this embodiment, in cases where the risk of heatstroke is assessed, for example, as part of an assessment of the physical condition of an athlete undergoing training, it is conceivable that the subject of assessment will be wearing sportswear, and in this case too, it is most reasonable to place the biosensor on the chest of the clothing worn on the upper body.

(Second Measuring Device)
Next, as an example of a second form of the measuring device 11 as a biometric information acquisition unit, we will explain a type of sensor that optically detects blood vessel contraction to detect the worker's pulse as heart rate data, and also detects the worker's body movements using a three-dimensional acceleration sensor.

Figure 3 shows a state in which a worker is wearing a wireless pulse monitor, which is a second measuring device that acquires the worker's biological information in the heatstroke risk management system according to this embodiment.

As shown in Figure 3, the wireless pulse meter 11b consists of a pulse detection part that is worn by pinching the earlobe with a pair of measuring pieces 11b1 and 11b2 biased by a spring or the like, and a main body part 11b3 that is connected to the pulse detection part by wire.

The pulse detection section has a pair of measuring pieces 11b1, 11b2, one of which has a light source such as an LED and the other a light-receiving element such as a CCD, arranged facing each other, and is secured to the earlobe by a spring or other biasing means. The pulse of the worker 11 is detected from the contraction of the blood vessels in the ear photographed by the CCD. The main body 11b3 houses a control circuit that outputs pulse rate data after appropriately filtering out noise components from the image of the contracted blood vessels detected by the pulse detection section, as well as a power supply for the entire pulse meter, a three-dimensional acceleration sensor, a data transmission section, and more.

In the wireless pulse monitor 11b of the type shown in Figure 3, in order to more accurately grasp the physical movements of the worker 11 being evaluated, it is preferable to attach the main body 11b3, on which the three-dimensional acceleration sensor is located, to the upper body of the worker 10, close to the torso, and as shown in the figure, it is considered preferable to attach it to the back of the collar, where movements that differ from the physical movements of the worker 10 are less likely to occur.

The commercially available wireless pulse monitor 11b shown in Figure 3 does not have a function for measuring environmental temperature. However, if the main body 11b3 is attached to the back of the collar as shown, the discomfort felt by the worker 10 (the person being evaluated) when wearing the main body 11b3 is less than when the biosensor 11a shown in Figure 2 is attached to the chest. Therefore, various sensors for monitoring environmental conditions, such as a thermometer or hygrometer, can be built into the main body 11b3. For the same reason, the main body 11b3 can be equipped with a transmission unit capable of directly transmitting biometric information to an Internet environment. In this case, the worker 10 does not need to carry the smartphone 12 equipped with the evaluation subject information transmission unit 13. If the worker 10 does not have a smartphone 12, an audio warning function can be built into the main body 11b3 of the wireless pulse monitor 11b to fulfill the function of the warning notification unit 14 previously performed by the smartphone 12.

In the wireless pulse meter shown in Figure 3 as the second measuring device, heart rate data is obtained by measuring the contraction of blood vessels in the earlobe, but pulse can also be detected optically at locations other than the earlobe, such as the fingertip. Also, while the wireless pulse meter shown in Figure 3 shows a wired connection between the pulse detection unit and the main unit, it is also possible to connect the pulse measurement unit and the main unit via a short-distance wireless connection if the operating power source for the pulse measurement unit can be easily secured.

(Third Measuring Device)
Next, we will explain the third form of the measuring device 11 as a biometric information acquisition unit, which is a wristwatch-type sensor that is worn on the wrist using a belt and has a device main body that internally contains a vibration sensor that detects pulse, an acceleration sensor that detects movement, and sensors that measure environmental conditions such as temperature and humidity.

Figure 4 shows a wristwatch-type sensing unit, which is a third measuring device used to acquire a worker's biometric information in the heatstroke risk management system of this embodiment.

As shown in Figure 4, the sensing unit 11c is composed of a device main body 11c1 that is worn in contact with the wrist and a watchband-type fixed belt 11c2, and has an appearance similar to that of a wristwatch. Note that the sensing unit 11c illustrated in Figure 4 is designed so that the device main body 11c1 is detachable from the fixed belt 11c2 for purposes such as easy maintenance.

The wristwatch-type sensing unit 11c is worn with the device body 11c1 on the outside of the wrist (the back of the hand), and the side that comes into contact with the wrist (the inside of the fixing strap 11c2) is equipped with a pulse detection unit (not shown) that is pressed against the outside of the wrist of the worker 10 being evaluated to detect the pulse. Inside the device body 11c1 are sensors that measure environmental information such as temperature, humidity, and air pressure, as well as a three-dimensional acceleration sensor that detects the movements of the worker 10. As shown in the figure, small openings are formed on the outer surface of the device body 11c1 in areas corresponding to these sensors. Furthermore, connection electrodes (not shown) are located on the side of the device body that supply power to a secondary battery located inside the device body and enable data exchange with the internal memory element. The device can be placed in a dedicated cradle to charge the operating battery of the sensing unit 11c and exchange data with a computer or the like.

Wristwatch-type measuring devices also include so-called smartwatches, which have a clock function that displays the current time on an image display device located on the surface of the device body. Some smartwatches are equipped with an optical heart rate sensor that measures heart rate based on infrared absorption and also have carrier communication capabilities. Using such smartwatches, biometric information such as heart rate information acquired by the measuring device can be sent directly to a cloud server via the Internet.

A wristwatch-type measuring device, such as the sensing unit 11c illustrated in FIG. 4, used in the heatstroke risk management system of this embodiment has a transmission unit within the device main body 11c1 that transmits measured biometric information to the smartphone 12 carried by the worker 10. Note that the third measuring device is also worn on the outside of the wrist of the person being evaluated. Therefore, similar to the second measuring device, the wireless pulse meter, the discomfort felt by the worker 10 when wearing the device main body 11c1 is less than when the biometric sensor 11a shown in FIG. 2 is worn on the chest. Therefore, the sensing unit 11c shown in FIG. 4 can also be provided with a transmission unit within the device main body 11c1 that can transmit biometric information directly to an Internet environment. Furthermore, if the worker 10 does not have a smartphone 12 with the subject information transmission unit 14, the device main body 11c1 will perform the function of the warning notification unit 14. However, it is easy to adopt a warning notification function for the worker 10 and a configuration that displays heatstroke risk data as an image, etc., in the case of a smartwatch, as well as other devices.

[Heatstroke risk assessment method]
Next, the specific contents of the heatstroke risk assessment for an individual worker in the hot environment assessment in the heatstroke risk management system according to this embodiment will be described.

The heatstroke risk assessment method according to this embodiment calculates a work strain index indicating the intensity of the work performed by the person being assessed based on the person's heart rate data detected by a heart rate detection means provided in a measuring device worn by the person being assessed and acceleration data acquired by a three-dimensional acceleration sensor. The method also calculates a heat stress index for the person being assessed based on the temperature inside the person's clothes obtained from the measuring device and the environmental temperature of the site where the person is working. Then, based on the calculated work strain index and heat stress index, a heatstroke risk index indicating the risk of developing heatstroke is calculated.

In the following explanation of the heatstroke risk assessment method, the components of the heatstroke risk management system according to this embodiment, as explained using Figure 1, will be illustrated as appropriate.

Figure 5 is a flowchart showing the flow of heatstroke risk assessment in the assessment unit in the heatstroke risk management system described in this embodiment.

In the heatstroke risk assessment system according to this embodiment, the assessment and determination unit 22, which is a control means provided in the cloud server 21 on the Internet, calculates the worker's 10 work strain index and heat load index based on data obtained from the biosensor 11a, which is a first measuring device serving as a biometric information acquisition unit worn by the worker 10 being assessed, and data obtained from each component within the cloud server 21, thereby calculating a heatstroke risk index.

As shown in Figure 5, when the evaluation by the evaluation assessment unit 22 begins (START), the evaluation assessment unit 22 begins calculating the workload index of the person being evaluated.

The start (START) of the evaluation by the evaluation and determination unit 22 can be set in various ways, such as by the worker 10 himself or the site supervisor 30, who is the manager, turning on the power switch of the biosensor 11a, which is a measuring device; by a timer being set so that the biosensor 11 automatically starts operating when the work start time arrives; or by the biosensor 11a itself detecting that the worker has put on an undershirt 18 equipped with the biosensor 11a and starting operation.

To calculate the work strain index, the evaluation and determination unit 22 first checks whether historical data, which is heart rate data and acceleration data, of the worker 10 being evaluated has been recorded in the data recording unit 24 (step S101).

If the worker 10 has previously been evaluated by the heatstroke risk management system described in this embodiment and historical data for the worker 10 is recorded in the data recording unit 24 (if "Yes" in step S101), a linear section in which heart rate changes linearly with acceleration is determined from the collection of historical data, and a regression line is determined for the historical data included in that linear section. The regression line determined from this historical data represents the characteristics (individuality) of the worker's heart rate response. The slope of this regression line is defined as the heart rate response coefficient αr, and the intercept is defined as the intercept heart rate βr, and these are calculated (step S102).

Then, the evaluation and determination unit 22 detects the heart rate data of the worker 10 measured by the biosensor 11a (step S103).

On the other hand, if no historical data exists for the worker 10, or if data exists but a certain period of time (for example, one month) has passed since the last data was recorded, the evaluation and determination unit 22 determines that it is not possible to calculate the correct standardized heart rate for the worker 10, and calculates the work strain index for the worker 10 from the acceleration data alone, not based on the heart rate data. The evaluation and determination unit 22 calculates the acceleration deviation, which is a numerical value indicating the movement of the worker 10 (step S110), and calculates the work strain index for the worker 10 based only on the acceleration deviation (step S111). In this case, the acceleration deviation can be converted to a work strain index using a predetermined formula, such as by multiplying it by an appropriate coefficient.

If the data recording unit 24 has recorded the worker's 10 history data ("Yes" in step S101), the evaluation and determination unit 22 ensures the reliability of the heart rate data when detecting the heart rate data. In the heatstroke risk management system of this embodiment, as described above, the electrode portion 11a2 of the biosensor 11a is located on the back surface 18b of the undershirt 18 worn by the worker 10 to better acquire the heart rate data of the worker being evaluated. However, the worker's body movements, sweating on the body surface, and other factors can sometimes prevent the heart rate from being detected correctly. For this reason, in the heatstroke risk management system of this embodiment, if the heart rate data of the worker being evaluated 10 cannot be measured correctly, the system checks whether the heart rate data has been acquired correctly to avoid calculating the work strain index based on incorrect data and resulting in an incorrect heatstroke risk assessment.

First, the evaluation and determination unit 22 calculates the heartbeat waveform detection rate to evaluate the reliability of the heartbeat data (step S104).

Raw data sampled from biosensors may contain a certain percentage of noise (abnormal heart rate data) due to factors such as poor contact between the subject's skin and the electrodes. Therefore, for each beat of data (heart rate interval), for example, data with a heart rate interval between 0.33 seconds and 1.33 seconds, and a difference from the previous data (differential heart rate interval) of 0.15 seconds or less, is determined to be normal and labeled.

The threshold for determining normality/abnormality can be set arbitrarily, but it is sufficient to set an appropriate value so that data with heartbeat intervals that are impossible from a physiological perspective can be removed. Then, the measurement data is divided into k subintervals with a specified time width, and the heartbeat waveform detection rate Q is calculated as the percentage of data labeled as normal for each subinterval out of the total data for that interval.

Next, the evaluation/determination unit 22 determines the heartbeat waveform detection rate for each subsection (step S105).

If the heart rate waveform detection rate is equal to or greater than a reference value (e.g., 50%), the section is determined to be reliable ("Yes" in step S105), and a work strain index is calculated using the heart rate data (step S106).

On the other hand, if the heart rate waveform detection rate is less than the reference value (50%), it is determined to be unreliable ("No" in step S105), and the process proceeds to step S110, where the acceleration data for that section is used to calculate the work strain index.

Note that the reference values in the above explanation are merely examples, and may be adjusted as appropriate depending on the performance of the biosensor, the subject's occupation, etc. For example, the threshold may be set lower for occupations that require intense movement, and higher for occupations that require less movement.

If the heart rate waveform detection rate is 50% or higher ("Yes" in step S105), the evaluation unit 22 calculates the median heart rate from the obtained heart rate data (step S106). Here, the representative value (median) for each sub-interval set in step S104 is used as the median heart rate data. The representative value may be the average value of the interval, but is preferably the median value of the interval. This is because even if the data obtained from the measurement device contains a small number of irregular values, the influence of these can be eliminated.

Furthermore, the evaluation and determination unit 22 simultaneously calculates the acceleration deviation, which is a numerical value indicating the operating status of the worker 10, from the acceleration data obtained from the biosensor 11a (step S107).

Next, the median heart rate is corrected based on the historical data and a previously created standard heart rate response model to determine the standardized heart rate (step S108).

Specifically, the median heart rate data is converted to a standardized heart rate using the heart rate response coefficient and intercept heart rate, as well as the standard heart rate response coefficient and standard intercept heart rate, which are parameters of the standard heart rate response model, using the following formula (Equation 1):

HR _S [k] = (αs/αr) (HR[k] - βr) + βs (Formula 1)
where:
Central heart rate data: HR [k]
Standardized heart rate: HR _S [k]
Heart rate response coefficient: αr
・Intercept heart rate: βr
Standard heart rate response coefficient: αs
Standard intercept heart rate: βs
Here, k represents the number of the subinterval.

The standard heart rate response model is a heart rate response model created based on large-scale data obtained by measuring a large number of people. It represents the standard human heart rate response to acceleration (body movement) and can be expressed using various parameters and specified mathematical formulas.

Figure 6 shows an example of the measurement results used to derive the standard heart rate response model.

Figure 6(a) plots all of the large-scale data (approximately 3 million points), with the shading representing data density. Line 51 in Figure 6(a) represents the 5% data line, line 52 represents the 25% data line, line 53 represents the 50% data line, line 54 represents the 75% data line, and line 55 represents the 95% data line.

The median values of each section are calculated and are indicated by the cross marks in FIG. 6(b), and the approximate curve F _HR (reference numeral 56) fitted to each median value represents the standard heart rate response model.

The approximate curve _FHR can be found using various curve fitting methods and can be expressed as a function _FHR (A _RMS ) that gives the estimated standardized heart rate _FHR when the acceleration deviation A _RMS is given. The slope of the linearly changing portion of _FHR (the section between around 0.05 and 0.45 acceleration) corresponds to the standard heart rate response coefficient, and the intercept of the approximate curve corresponds to the standard intercept heart rate.

The large-scale data may be data from the past few days for multiple workers at the site in question, or accumulated data sampled in advance at another site. Preferably, a standard heart rate response model is created based on large-scale data obtained by measuring a large number of workers who perform the same tasks as the worker in question. This is a heart rate response model optimized for the task in question, and is thought to represent the typical heart rate response of workers performing that task. There are no particular rules regarding the number of people on which the large-scale data is based, but the larger the sampling number, the more accurately the heart rate response can be approximated. Preferably, it is five or more people, more preferably fifty or more people. There are also no particular rules regarding the accumulation period, but it is preferable to obtain data for two or more days, more preferably five or more days, at the same site.

Based on the standardized heart rate and acceleration deviation obtained in this manner, the evaluation and determination unit 22 calculates the work strain index of the worker 10 (step S109).

At this time, the evaluation and determination unit 22 determines whether to use the standardized heart rate or the estimated standardized heart rate using a correction map created based on the standardized heart rate response model. Specific methods for calculating the median heart rate and acceleration deviation from the heart rate data obtained from the biosensor 11 (a measurement device) and the values of the three-dimensional acceleration sensor, as well as criteria for selecting heart rate data using the correction map, will be described in detail later.

On the other hand, if the heartbeat waveform detection rate Q is not equal to or greater than the reference value (50% in the above example) ("No" in step S105), the evaluation and judgment unit 22 determines that the reliability of the obtained heartbeat data is low and does not calculate the median heart rate, but instead calculates the acceleration deviation, which is a numerical value indicating the operating status of the worker 10, from the acceleration data obtained from the biosensor 11a (step S110). In this case, the evaluation and judgment unit 22 calculates the work strain index of the worker 10 based only on the acceleration deviation (step S111).

The work strain index obtained in step S111 is thought to be less accurate than the work strain value obtained in step S109 because it does not reflect heart rate data. However, since the worker 10's movements are continuous, it is preferable to obtain a work strain index continuously rather than not calculating the work strain index during that time because heart rate data is not available.

Furthermore, the evaluation and determination unit 22 calculates the heat load index of the worker 10 based on the temperature data inside the worker's 10 clothes obtained from the biosensor 11a and the environmental temperature data at the work site where the worker 10 is working (step S112). The environmental temperature data at the work site can be obtained based on the ambient temperature data at the work site obtained by the weather information acquisition unit 25 of the cloud server 21, or, if the worker is working indoors, based on temperature information obtained from a temperature sensor installed at the work site.

The specific procedure for calculating the heat load index based on the temperature data inside the worker's 10 clothing and the environmental temperature data will be described in detail later.

Then, the evaluation and determination unit 22 calculates the risk of heatstroke for the worker 10 as a heatstroke risk index based on the obtained work strain index and heat load index (step S113).

The heatstroke risk index in the heatstroke risk assessment system shown in this embodiment can be determined as the linear sum of the work strain index and the heat load index. Therefore, the higher the value of the heatstroke risk index, the higher the risk that the worker will develop heatstroke. By defining the magnitude of the heatstroke risk index as a range, the risk of developing heatstroke can be ranked as high (= dangerous), somewhat high (= caution), or low (= safe). Therefore, depending on the rank of the heatstroke risk index calculated by the evaluation/determination unit 22, the worker 10 or the site supervisor 30 can take measures such as stopping work and taking a break, reducing the work load, or reducing the heat load by lowering the temperature inside the clothing, thereby effectively avoiding the onset of heatstroke.

[Method for assessing the risk of heat stroke]
Here, an algorithm for calculating a heatstroke risk index, which is a heatstroke risk assessment index for each worker, performed in the heatstroke risk assessment system according to this embodiment will be described.

(Estimation of workload)
<a. Pretreatment>
First, pre-processing is performed on the heart rate data and acceleration data to calculate a work strain index.

As explained using the flowchart in Figure 5, the heart rate data is preprocessed by calculating the central heart rate from the heart rate data detected by the biosensor 11a worn by the worker 10 (step S106 in Figure 5).

More specifically, if the heartbeat waveform detection rate for a subsection is 50% or higher, the heartbeat data acquisition interval included in the subsection is converted into the heartbeat rate per subsection (e.g., per past minute) to obtain the median heart rate HR.

On the other hand, for acceleration data obtained by the acceleration sensor, the average value ΔA for the past minute is calculated using the following procedure.

1) Exponential moving average of data at unequal time intervals For the acceleration data {Ax(t)}, {Ay(t)}, and {Az(t)} in the x-, y-, and z-axis directions, the exponential moving average of the acceleration data in each axial direction is calculated using the exponential moving average method, a statistical technique, with a time constant of 10 seconds. The time constant is not particularly limited, but may be set appropriately in the range of 5 to 10 seconds, for example, depending on the performance of the acceleration sensor.

Here, the exponential moving averages for the x-axis, y-axis, and z-axis directions are {Sx(t)}, {Sy(t)}, and {Sz(t)}, respectively.

2) Removal of Exponential Moving Averages The exponential moving averages described above are removed from the acceleration data for each axis to obtain detrended time series acceleration. For example, in the case of the x-axis, the result is "Ax(t) - Sx(t)".

3) Calculation of the sum of squares For the trend-removed time series acceleration, calculate the square at each time using the following formula (Formula 2) to find the sum.

4) Minute-by-minute average of the square sum "ΔA ² (t)" calculated above is calculated as the average value "ΔA ² _ave " for each minute. Here, the average value is calculated by dividing by the number of data points. Also, the square root "ΔA _ave " of the square mean of the acceleration "ΔA ² _ave " is calculated. Here, ΔA _ave is the acceleration deviation A _RMS .

<b. Removal of outliers>
Regarding the heart rate data obtained from the heart rate data, non-numeric data and heart rate values of 40 or less and 180 or more are removed as abnormal values.

Furthermore, for acceleration data, non-numeric data and data with a ΔA of 0.05 or less or 0.55 or more are excluded as abnormal values.

<c. Calculation of intercept heart rate and heart rate response coefficient>
For a set of historical data, a linear response interval is set in which the median heart rate changes linearly with respect to the acceleration deviation, and a regression line is found for the data included in that linear response interval. The slope of this regression line is the heart rate response coefficient αr, and the intercept is the intercept heart rate βr.

There are no particular limitations on how the regression line is applied. For example, the range of acceleration deviations from 0.05 to 0.4 is set as the linear response range, and this range is divided into m sub-ranges (m is, for example, 3 to 7). Next, the median heart rate and the median acceleration deviation are calculated for each sub-range. A regression line is then applied to the calculated median coordinates of the m points (see Figure 7).

<d. Calculation of Workload Index>
5, if historical data of the worker 10 being measured is available, a standardized heart rate HR _S is calculated based on this historical data and the standard heart rate response model, as shown in step S108. By calculating the standardized heart rate HR _S , it is possible to correct for individual differences in calculating the workload index from the heart rate data due to the characteristics of each individual worker.

On the other hand, the estimated standardized heart rate is calculated by estimating the standard heart rate from the acceleration deviation A _RMS of the subject using the approximate formula F _HR (A _RMS ) of the standard heart rate response model. The key to calculating the actual workload index is determining whether to trust the standardized heart rate or the estimated standardized heart rate. In this embodiment, a correction map created based on the standard heart rate response model is used to determine which heart rate to select, and the corrected heart rate HR _s is obtained.

Figure 8 is a first example of a correction map.

As shown in Figure 8, the correction map shows an approximate curve _FHR 71, which has the standard intercept heart rate βs as its y-axis intercept and the slope of its linear portion is the standard heart rate response coefficient αs. Here, the approximate curve _FHR 71 is referred to as the judgment line. It has been found that, once the acceleration deviation exceeds 0.45, the judgment line 71 is no longer a straight line, as shown in Figure 7, and the degree of increase in median heart rate relative to acceleration deviation decreases.

In the correction map shown in Figure 8, a boundary line 72 is set in the area where the acceleration deviation, which indicates the movement of the worker being measured, is 0.2. This is because in areas where the acceleration deviation is smaller than 0.2, the influence of emotions is greater than the change in heart rate due to movement, and in areas where the acceleration deviation is larger than 0.2, the fluctuation in heart rate due to body movement is thought to be greater.

In the correction map shown in FIG. 8 , the relationship between the heart rate and the acceleration deviation is corrected to fall within the hatched regions 73 and 77 in the map. For example, in the range of acceleration deviation up to 0.2, when the standardized heart rate is large and located above the judgment line 71, the value of the judgment line 71, i.e., the estimated standardized heart rate, is used as the corrected heart rate _HRs , as indicated by arrow 74 in the figure. When the standardized heart rate is smaller than the standard intercept heart rate βs, the value of the standard intercept heart rate βs is adopted as the corrected heart rate _HRs . Furthermore, when the standardized heart rate is equal to or smaller than the judgment line 71 but equal to or larger than the standard intercept heart rate βs, the standardized heart rate is adopted as the corrected heart rate _HRs . In this way, when a standardized heart rate greater than the estimated standardized heart rate is detected in the region where the acceleration deviation is smaller than 0.2, this can be eliminated as an effect of emotion.

On the other hand, in the region where the acceleration deviation is greater than 0.2, a parallel line 76 is drawn with the same slope as the aforementioned standard heart rate response coefficient αs, i.e., parallel to the linear portion of the approximated curve F _HR , to define the region where the heart rate value is determined to be too high. A region 77 sandwiched between this parallel line 76 and the judgment line 71 is determined to be a region where the correct heart rate has been detected. If the heart rate falls within this region 77, the workload index is calculated using the standardized heart rate as is as the corrected heart rate _HRs . If the standardized heart rate is greater than the upper limit of the parallel line 76, the value on the parallel line 76 is adopted as the corrected heart rate _HRs , as indicated by arrow 78 in the figure, thereby eliminating the influence of errors. Furthermore, for values appearing in the region below the judgment line 71, the value on the judgment line 71, i.e., the estimated standardized heart rate, is adopted as the corrected heart rate _HRs , as indicated by arrow 79 in the figure. This prevents a heart rate value that is too low from being used to calculate the workload index even when the subject is performing more than a certain level of movement.

The correction map shown in Figure 9 is used when the detected heart rate data is deemed to be more reliable.

Heart rate data is considered to be highly reliable when the detection rate of the heart rate data acquired by the biosensor 11 is higher than the judgment criterion (for example, 50%), and continues to exceed, for example, 80%.

The correction map shown in Fig. 9 is basically the same as the correction map shown in Fig. 7 , but differs in that the acceleration deviation is 0.2 or greater and the standardized heart rate is in a region lower than the judgment line 81. When the reliability of the heart rate data shown in Fig. 9 is high, a boundary line 88 is drawn below the judgment line 81 from the position of the standard intercept heart rate βs at an acceleration deviation of 0.2 to define a lower limit parallel to the judgment line 81, and the standardized heart rate in a region 89 between the boundary line 88 and the judgment line 81 is used as is as the corrected heart rate _HRs . When the standardized heart rate is smaller than the boundary line 88, the value on the boundary line 88 is adopted as the corrected heart rate _HRs , as shown by arrow 91 in the figure.

By doing this, a wider range of standardized heart rates can be used to calculate a more accurate workload index.

<e. Workload evaluation>
Based on the corrected heart rate HR _c obtained using the correction map, the workload index W is calculated as follows:

First, the corrected heart rate _HRc is converted into metabolic equivalents (METs) using the following formula (Formula 3).

METs=a _METs ×HR _c +b _METs (Formula 3)
Here, a _METs and b _METs are predetermined parameters that can be determined based on respiration measurement experiments.

Next, the metabolic equivalents METs are converted to a workload index W using the following formula (Equation 4):

W=a _W ×METs+b _W (Formula 4)
Here, a _W and b _W are predetermined parameters.

For example, when a _W = 0.2 and b _W = -0.2 are set, the work burden can be evaluated as follows: if the work burden index W is 0.6 or more, it is work that requires a high metabolic rate, i.e., work that places a heavy burden on the worker; and if the value of W is 1 or more, it is work that places an extremely high metabolic rate, i.e., work that places a very heavy burden on the worker.

(Evaluation of heat load)
Using the clothing temperature Ti obtained by the measuring device 11 and the outside air temperature To obtained as the environmental temperature, the heat load index H is calculated by the following formula (Formula 5).

If the heat load index H is less than 0, H = 0.

When the heat load index H is 0.6 or higher, the heat load is considered to be relatively high, and when the heat load index H is 1 or higher, the heat load is considered to be extremely high.

(Assessment of the risk of heat stroke)
Using the work strain index W and heat load index H obtained by the above calculation, a heat stroke risk assessment index R for the worker 10 to be evaluated is calculated as shown in the following formula (Formula 6).

Here, a is a numerical value defined according to the heat acclimatization of the worker being evaluated, with a = -1.8 if heat acclimatized and a = -1.3 if not.

The heatstroke risk assessment value R calculated in the above manner can be judged as low risk if R is less than 0.6, low risk if R is 0.6 or greater but less than 1.0, requiring caution and a warning level, and high risk if R is 1.0 or greater, indicating a danger level of heatstroke.

However, since it is not possible to verify the actual occurrence of heatstroke, when determining the criteria for judging the risk of heatstroke, they should be determined so that the risk of heatstroke can be judged as strictly as possible, i.e., on the safe side.

(Continuous assessment of heatstroke risk)
When continuously assessing the risk of heatstroke of a worker 10 based on measurement results obtained from a biosensor 11, which is a measuring device worn by the worker 10, the moving average values of the heat load index H and the work strain index W are calculated using the following equations (Equation 7) and (Equation 8) with a sampling interval of one minute.

Here, w ₁ =2/31 and w ₂ =2/11.

Furthermore, the exponential moving average value of the heatstroke risk index R can be calculated using the following equation (Equation 9):

For example, if the exponential moving average value of the heatstroke risk index R remains at or above 1 for 30 minutes or more, it is determined that the risk of developing heatstroke is extremely high, and measures are taken to prevent the onset of heatstroke, such as encouraging workers to take breaks.

(Displayed on a 2D map)
As can be seen from the above equation (Equation 6), in the heatstroke risk management system of this embodiment, the index R representing the risk of heatstroke is expressed as a linear sum of the heat load index H and the work burden index W for the worker 10.

Using this, the heatstroke risk index can be displayed as a heatstroke risk index on a two-dimensional map with the heat load index and work strain index as axes. For example, by displaying symbols on a two-dimensional map corresponding to the current heatstroke risk index for each of multiple workers 10 managed by a site supervisor 30, who is also a manager, the site supervisor 30 can grasp the overall risk index of the workers under his/her management at a glance. Specific explanations of the display image that shows the degree of heatstroke risk for each worker 10 will be omitted.

[Handling of biological information obtained by a device other than the first measuring device]
As described above, in the heatstroke risk management system according to this embodiment, the physical condition of each worker is evaluated and the risk of developing heatstroke is assessed based on the biological information acquired by the biological sensor 11a, which is the first measuring device, attached to the chest area of the undershirt shown in Fig. 2. Therefore, by converting the biological information acquired by the wireless pulse meter 11b as the second measuring device or the sensing unit 11c as the third measuring device described above into biological information acquired by the first measuring device, workers wearing different types of biological information acquisition units can be managed in common by the heatstroke risk management system.

In this embodiment of the heatstroke risk management system, biometric information obtained from different measurement devices is integrated by using median heart rate as a common parameter for heart rate data and acceleration deviation as a common parameter for acceleration data.

(Method for processing biological information acquired by second measurement device)
The wireless pulse rate monitor 11b outputs the pulse wave interval measured as heart rate data every second, which differs from the biosensor 11a in that the heart rate interval is output for each heart rate.

As described above, the biosensor 11a judged the reliability of the heart rate data by estimating the heart rate waveform detection rate, and calculated the central heart rate HR using only heart rate data that was judged to be normal. Correspondingly, the wireless pulse meter 11b also discarded unreliable pulse data output each second, and only used values considered to be highly reliable.

More specifically, data with a value of 0, or data below 45 bpm or above 180 bpm was discarded as being unreliable. Furthermore, because it is physiologically unnatural for the same pulse to occur consecutively, if the same pulse data occurred five or more times in a row, the data from the fifth beat onwards was discarded as defective data.

In this way, using only reliable data, the one-minute median of the central heart rate over a 60-second window every 30 seconds was used to determine the central heart rate (HR).

As for the acceleration data, since the wireless pulse rate monitor 11b measures the area around the head or neck, such as the back of the collar, or the upper body as shown in Figure 3, only the high-frequency components of the acceleration data were extracted. As an example, by passing the data through a high-pass filter with a cutoff frequency of low-frequency components below 0.1 Hz, it is believed that noise components that are not related to the movements of the worker 10 being evaluated can be removed.

Then, the squared deviation over one minute is calculated for the acceleration data that has passed through the high-pass filter in the x, y, and z directions.

The values of the median heart rate and acceleration deviation obtained in this way are considered to be the same as the median heart rate and acceleration deviation obtained by the first measuring device, the biosensor 11a, so the work strain index can be calculated as described above.

As mentioned above, some commercially available wireless pulse rate monitors 11b do not have sensors to measure environmental conditions such as temperature and humidity. For this reason, when calculating the heat load index, it is preferable to estimate the environmental conditions in which the worker 10 is located using data acquired by the weather information acquisition unit 25 of the cloud server 21. In this case, research materials comparing data from clothing temperature sensors with WBGT (such as experimental data from Shinshu University) are publicly available, and appropriate reference to such materials is expected to lead to more accurate calculations of the heat load index.

(Method for processing biological information acquired by a third measurement device)
The sensing unit 11c outputs pulse wave intervals measured as heart rate data at regular intervals (e.g., every minute), and also outputs METs estimates for the subject's activity status at regular intervals (e.g., every minute).

In this case, the one-minute pulse interval data is multiplied by a predetermined coefficient, for example 0.984, to correlate it with the heart rate output by the biosensor 11a and determine the median heart rate value. This coefficient is used to correct for the asymmetry in the statistical distribution of heart rate intervals and the difference between the median and average heart rate intervals over one minute due to the characteristics of the pulse wave measuring device.

Furthermore, taking advantage of the fact that METs data and acceleration data correspond linearly during walking, light jogging, and the like, it is possible to calculate a correlation coefficient by actually acquiring data from two measuring devices simultaneously, and then convert the output METs data into acceleration data. From the resulting converted acceleration data, acceleration displacement can be calculated in the same way as when it is calculated using the biosensor 11a.

In addition, because the sensing unit 11c is worn on the arm of the worker 10 being evaluated, it detects not only the movements of the worker's 10 entire body but also the fine movements of the worn arm as acceleration data. For this reason, for example, even at a work site, the acceleration data measured will differ between work such as carrying and assembling large components and work requiring the use of hands to handle small parts. In such cases, the reliability of the measurement data can be improved by taking measures such as appropriately grouping workers 10 performing similar work and setting different evaluation criteria for each group.

In this way, by using the heart rate data and METs data acquired by the sensing unit 11c to determine the central heart rate and acceleration deviation, the work strain index can be calculated in the same way as with the data obtained by the biosensor 11a.

The sensing unit 11c has a sensor that measures environmental conditions such as temperature and humidity. However, the measurement location is the wrist, which is naturally different from the temperature inside the clothing measured by the biosensor 11a. For this reason, when calculating the heat load index, as in the case of the wireless pulse meter 11b described above, it is preferable to estimate the environmental conditions in which the worker 10 is placed using data acquired by the weather information acquisition unit 25 of the cloud server 21, or to separately calculate the correlation between the temperature inside the clothing and the temperature of exposed body surfaces, and then correct the temperature data as appropriate to calculate the heat load index.

As explained above, the biometric information processing method and biometric information processing system disclosed in this application use a common index to convert the biometric information of the person being evaluated acquired by different biometric information acquisition units into each other, allowing evaluation and management of the person being evaluated within a common system without being restricted by the type of biometric information acquisition unit.

As a result, it is possible to build a biometric information processing system that can evaluate more subjects at a lower cost, tailored to customer needs, for example by allowing the subject to select and use the type of biometric information that they prefer, or by using a biometric information acquisition unit that they already own. As the number of subjects increases, the number of data processing examples in the biometric information processing system increases, making it possible to make corrections using past data and further improving the accuracy of the evaluation results obtained from the biometric information.

In addition, by referencing a large amount of data, it is possible to understand trends in the accuracy of acquired data due to differences in measurement devices, thereby improving the accuracy of conversion of data acquired by different biometric information acquisition units.

In the above embodiment, the median heart rate was used as a common parameter for heart rate data, but parameters other than the median heart rate, such as the average heart rate, can also be used.

In addition, in the above embodiment, acceleration deviation was used as a common parameter for acceleration data, but parameters other than acceleration deviation, such as the square root of the resultant acceleration deviation, can also be used.

In addition, in the above embodiment, the biometric information processing system disclosed in the present application is exemplified as a heatstroke risk management system in which workers working at construction sites, etc., are evaluated. However, the system is not limited to the above example and can acquire biometric information from multiple subjects, and can be used as a biometric information processing system that evaluates the work strain index, physical condition assessment index, heat load index, exercise load index, and other indices of each subject based on the acquired biometric information.

For example, it can be used in systems that process a wide range of biometric information, with different subjects, biometric information measured, and evaluation purposes, such as managing the physical condition of athletes during training or managing the physical condition of residents in elderly care facilities.

The biometric information processing method and biometric information processing system disclosed in this application are not limited to the biometric information acquisition unit worn by the person being evaluated, making it possible to realize a method for processing biometric information for a larger number of people being evaluated, and it is possible to construct a highly versatile biometric information processing system at low cost, making it extremely useful.

10. Worker (evaluated person)
11a Biosensor (first measuring device, bioinformation acquisition unit)
11b Wireless pulse rate monitor (second measuring device, biological information acquisition unit)
11c Sensing unit (third measurement device, biological information acquisition unit)
22 Evaluation and judgment unit (data processing unit)

Claims (6)

a step of acquiring biometric information of the person to be evaluated by a biometric information acquisition unit;
and calculating an index representing the state of the subject based on the acquired biometric information,
There are multiple assessees,
The biometric information of each of the plurality of subjects is acquired by one of a plurality of types of biometric information acquisition units, each of which has a different frequency of detecting biometric information data or a different frequency of data transmission;
In a data processing unit located away from the subject, if the biological information is heart rate data, it is converted into a central heart rate as a common parameter, and if the biological information is acceleration data or METs, it is converted into an acceleration deviation or a square root of the squared deviation of a composite acceleration as a common parameter,
A biometric information processing method, characterized in that the biometric information acquired by the biometric information acquisition units in different forms is integrated into a common parameter to calculate the index. The biometric information processing method according to claim 1 , wherein the plurality of forms include at least three types of forms: a clothing type, a wristwatch type, and a type worn on an earlobe or a fingertip. A biometric information processing method as described in claim 1 or 2, wherein the index is at least one of a workload index indicating the degree of impact of the workload on the person being evaluated, a physical condition assessment index indicating the degree of change in the person's physical condition from its normal state, and a heatstroke risk index indicating the degree of risk of the person being evaluated developing heatstroke. 4. The biological information processing method according to claim 1 , wherein the biological information is at least one of heart rate data, acceleration data, and METs. The biological information processing method according to claim 4 , wherein the common parameter used for the heart rate data is a median heart rate. Two or more different types of biometric information acquisition units, each with a different frequency of detecting biometric information data and a different frequency of transmitting data , acquire biometric information of each of a plurality of subjects;
a data processing unit located away from the subject and calculating an index representing the subject's condition based on the acquired biometric information;
The data processing unit converts the biometric information acquired by the biometric information acquisition unit into a central heart rate as a common parameter if the biometric information is heart rate data, or into acceleration data or METs if the biometric information is acceleration deviation or the square root of the squared deviation of resultant acceleration as a common parameter, and integrates the biometric information acquired by the biometric information acquisition unit in different forms into these common parameters to calculate the index.