Work monitoring assistance device, work monitoring assistance system, work monitoring assistance method, and program

Abstract

A work monitoring assistance device (1) of the present disclosure comprises: an unsafe determination unit (17) that, using sensor information acquired by a sensor (2) which detects the conditions of a work site, performs first detection processing for detecting whether or not an environment in which a worker working at the work site is present is an unsafe environment, and second detection processing for detecting whether or not the worker is in an unsafe state, that determines an occurrence of an unsafe event if the result of the first detection processing and the result of the second detection processing indicate that the worker is present in an unsafe environment and the worker is in an unsafe state, and that generates unsafe information indicating the determination result; and a transmission / reception unit (11) that outputs the unsafe information.

Description

Work monitoring support device, work monitoring support system, work monitoring support method and program

[0001] The present disclosure relates to a work monitoring support device, a work monitoring support system, a work monitoring support method, and a program that support monitoring of a work site.

[0002] At construction sites and other work sites, it is generally necessary to monitor whether any unsafe events involving workers have occurred. Traditionally, supervisors, such as site supervisors and work chiefs, have visually monitored the work site, but visual monitoring makes it difficult to constantly monitor all workers. For this reason, the introduction of tools to support work site monitoring is desirable. Patent Document 1 discloses a system that uses image data to monitor unsafe behaviors, such as when a worker fails to fasten the hook of their safety belt to the pole to which it is attached. When the system described in Patent Document 1 detects unsafe behavior, it notifies an administrator terminal of the location information and date and time at which the unsafe behavior was detected, and the administrator terminal displays the notified information. The administrator can recognize the occurrence of unsafe behavior by checking the information displayed on the administrator terminal.

[0003] International Publication No. 2021 / 177183

[0004] The system described in Patent Document 1 notifies an administrator terminal of an unsafe behavior when it detects an unsafe behavior. However, for example, even if a worker does not fasten the hook of his / her safety belt to the pole to which it is attached, the worker's condition is not actually unsafe as long as the worker is in a safe environment. In this way, even if a worker performs an unsafe behavior, it may not actually be unsafe depending on the environment in which the worker is present. With the technology described in Patent Document 1, the administrator is notified that an unsafe event has occurred even when it is not actually an unsafe event.

[0005] The present disclosure has been made in view of the above, and aims to provide a work monitoring support device that can provide more accurate information regarding unsafe conditions.

[0006] In order to solve the above-mentioned problems and achieve the objectives, the work monitoring support device disclosed herein uses sensor information acquired by a sensor that detects the situation at the work site to perform a first detection process to detect whether the environment in which a worker working at the work site is located is an unsafe environment and a second detection process to detect whether the worker is in an unsafe state, and is equipped with an unsafety determination unit that determines that an unsafe event has occurred and generates unsafety information indicating the determination result, and an information output unit that outputs the unsafety information.

[0007] The work monitoring support device of the present disclosure has the effect of being able to provide more accurate information regarding unsafe conditions.

[0008] FIG. 1 is a diagram showing an example of the configuration of a work monitoring support system according to an embodiment; FIG. 2 is a diagram for explaining sensors according to an embodiment; FIG. 3 is a flowchart showing an example of a work monitoring support processing procedure in a work monitoring support device according to an embodiment; FIG. 4 is a diagram showing an example of a method for associating sensors with workers according to an embodiment; FIG. 5 is a diagram showing an example of worker information according to an embodiment; FIG. 6 is a diagram showing an example of the configuration of a first detection unit and a second detection unit according to an embodiment;

[0009] Hereinafter, a work monitoring support device, a work monitoring support system, a work monitoring support method, and a program according to embodiments will be described in detail with reference to the accompanying drawings.

[0010] 1 is a diagram showing an example of the configuration of a work monitoring support system according to an embodiment. The work monitoring support system of this embodiment includes a work monitoring support device 1 that supports monitoring of a work site, and a sensor 2 that detects the state of the work site. The work site of this embodiment is, for example, a work site that involves work at height, but is not limited to this. Below, an example will be described in which the work site is a construction site, but the work site may also be a machine installation site in a factory, a machine repair site in a factory, a production line, etc., and is not limited to a construction site.

[0011] The work monitoring support device 1 uses information acquired by the sensor 2 to detect unsafe environments and unsafe conditions for workers. If both an unsafe environment and an unsafe condition are detected, the work monitoring support device 1 outputs unsafe information indicating the occurrence of an unsafe event for the worker. The work monitoring support device 1 outputs the unsafe information by, for example, transmitting the unsafe information to at least one of a mobile terminal 3 carried by the worker, an alarm device 4 carried by the worker, a supervisor terminal 5 carried by a supervisor who monitors the work at the work site, such as a site supervisor or work chief, and a management device 6 installed in an office. The office may be, for example, at least one of a site office installed near the work site, the office of the business operator to which the worker belongs, and the office of a construction management company such as a general constructor (hereinafter abbreviated as "general contractor") that oversees the entire work. The unsafe information may be an alarm using audio, vibration, or text, or may include at least one of text and an image indicating the detected unsafe event. Details of the unsafe information will be described later.

[0012] The unsafe state includes, for example, at least one of a state in which a worker is engaging in unsafe behavior and a state in which a worker is in poor physical condition (hereinafter also referred to as a poor physical condition). The unsafe state includes, for example, at least one of a worker not wearing a safety harness (fall prevention equipment) and a worker jumping from one place to another, but is not limited to these. The poor physical condition includes, for example, at least one of a state in which a worker is seen staggering, a state in which a worker is suspected to have a specific symptom such as heat stroke, a state in which there is an abnormality in vital data, and the like, but is not limited to these.

[0013] "Not wearing a safety harness" refers to a state in which the safety harness hook is not attached to the equipment to which the hook is attached, such as a lifeline. "Jumping" refers to a worker jumping from one structure to another. In this embodiment, the work monitoring support device 1 outputs unsafety information indicating the occurrence of an unsafe event, for example, when it detects that the safety harness hook is not attached and the environment is an unsafe environment in which the safety harness hook should be attached. Furthermore, when changing the safety harness hook attachment point, an operation called "double hooking" must be performed, in which two hooks are attached to the two attachment points before and after the change. Therefore, the work monitoring support device 1 outputs unsafety information indicating the occurrence of an unsafe event, for example, when it detects that the safety harness hook is not attached or that only one safety harness hook is attached to a lifeline or the like and the environment is an unsafe environment in which double hooking should be performed.

[0014] An unsafe environment is an environment that is unsafe for workers, and is, for example, at least one of, but not limited to, an environment on a steel frame, an environment where equipment for attaching a hook for a safety harness such as a lifeline is provided, an environment where a hole exists, an environment where the height is equal to or greater than a threshold value, an environment where a handrail is provided, an environment where the floor is not fully tightened, etc. The height threshold is, for example, 2 m, but is not limited to this.

[0015] The sensor 2 detects the situation at the work site and transmits the detection results as sensor information to the work monitoring support device 1. The sensor 2 used to detect an unsafe state of a worker and the sensor 2 used to detect an unsafe environment may be provided separately, or the same sensor 2 may be used for detecting an unsafe state of a worker and an unsafe environment. The sensor 2 is, for example, a camera that captures visible images. However, the sensor 2 is not limited to this and may also be an infrared camera, radar, a light detection and ranging (LiDAR), a time-of-flight (ToF) sensor, or multiple types of sensors. The sensor 2 may also include a hook sensor that detects that the hook of the safety belt is attached to the lifeline, or a human presence sensor that detects the presence of a person near the attachment equipment such as the lifeline. The hook sensor is provided on the hook of the safety belt. The hook sensor may be, for example, a sensor that detects the attachment of the hook based on the presence or absence of electricity, or a contact sensor that detects contact with the attachment equipment. The detection method is not particularly limited. The sensor 2 may also include a vital sensor that detects vital data of the worker, such as heart rate, body temperature, blood pressure, etc. The sensor 2 may be worn directly or indirectly by the worker, or may be installed at the work site.

[0016] 1 , the work monitoring support device 1 includes a transmitter / receiver 11, an information storage unit 12, a first detection unit 13, a second detection unit 14, a determination unit 15, and a response management unit 16. The first detection unit 13, the second detection unit 14, and the determination unit 15 constitute an unsafety determination unit 17. The unsafety determination unit 17 uses sensor information to perform a first detection process for detecting whether the environment in which a worker exists is an unsafe environment and a second detection process for detecting whether the worker is in an unsafe state, and determines that an unsafe event has occurred when the results of the first detection process and the second detection process indicate that the worker is in an unsafe environment and is in an unsafe state, and generates unsafety information indicating the determination result.

[0017] The transmitter / receiver 11 communicates with other devices. For example, the transmitter / receiver 11 receives sensor information, which is information acquired by the sensor 2, from the sensor 2, stores the received sensor information in the information storage unit 12, and outputs the sensor information to at least one of the first detector 13 and the second detector 14. Note that the first detector 13 and the second detector 14 may read the sensor information from the information storage unit 12 instead of receiving the sensor information from the transmitter / receiver 11. The transmitter / receiver 11 also functions as an information output unit that outputs unsafety information received from the determination unit 15 by transmitting the unsafety information to at least one of the mobile terminal 3, the alarm device 4, the supervisor terminal 5, and the management device 6. The information storage unit 12 stores the sensor information, the unsafety information generated by the determination unit 15, information used in processing by the work monitoring support device 1, and the like.

[0018] The first detection unit 13 uses the sensor information to perform a detection process for an unsafe environment (first detection process) and outputs the detection result to the determination unit 15. The detection process for an unsafe environment may use progress information indicating the progress of the work. The progress information is, for example, managed by the management device 6, received from the management device 6 by the transmission / reception unit 11, and stored in the information storage unit 12. However, this is not limiting, and the progress information may be transmitted from a device other than the management device 6, or input to the work monitoring support device 1 by an input means (not shown). The second detection unit 14 uses the sensor information to perform a detection process for an unsafe state of the worker (second detection process), and outputs the detection result to the determination unit 15. The first detection unit 13 uses the sensor information to perform a detection process for an unsafe state of the worker, and outputs the detection result to the determination unit 15.

[0019] The determination unit 15 determines whether or not both an unsafe environment and an unsafe state have been detected using the detection results of the first detection unit 13 and the second detection unit 14, and if both an unsafe environment and an unsafe state have been detected, determines that an unsafe event has occurred, and outputs unsafety information indicating the determination result to the transmission / reception unit 11. The determination unit 15 also stores the unsafety information in the information storage unit 12. Details of the operations of the first detection unit 13, the second detection unit 14, and the determination unit 15 will be described later.

[0020] The mobile terminal 3 is a terminal such as a smartphone that can be carried by a worker. While FIG. 1 illustrates one mobile terminal 3, multiple mobile terminals 3 may be used. The mobile terminal 3 includes a transceiver 31, a display 32, an audio output unit 33, a vibration generator 34, and a reader 35. The transceiver 31 communicates with other devices. For example, when the transceiver 31 receives unsafety information from the work monitoring support device 1, the transceiver 31 outputs the received unsafety information to at least one of the display 32, the audio output unit 33, and the vibration generator 34 depending on the type of the unsafety information. For example, if the unsafety information includes display data indicating the content to be displayed, the transceiver 31 outputs the display data to the display 32. If the unsafety information includes audio data, the transceiver 31 outputs the audio data to the audio output unit 33. If the unsafety information includes vibration instruction data instructing the generation of vibration, the transceiver 31 outputs the vibration instruction data to the vibration generator 34.

[0021] The display unit 32 displays characters, images, etc. The display unit 32 displays, for example, display data received from the transmitting / receiving unit 31. The audio output unit 33 outputs audio. For example, the audio output unit 33 outputs audio data received from the transmitting / receiving unit 31 as audio. The vibration generating unit 34 vibrates the portable terminal 3. For example, the vibration generating unit 34 vibrates the portable terminal 3 based on vibration instruction data received from the transmitting / receiving unit 31.

[0022] The reading unit 35 reads two-dimensional codes such as a QR (registered trademark) (Quick Response) code and outputs the read result to the transceiver 31. The reading unit 35 includes, for example, a camera capable of capturing images of two-dimensional codes and an analysis unit that converts the two-dimensional code captured by the camera into information indicated by the two-dimensional code. For example, as described below, if a camera attached to a worker's helmet or the like is used as the sensor 2 and a two-dimensional code indicating the camera's identification information is affixed to the camera, the reading unit 35 reads the two-dimensional code affixed to the camera to obtain the camera's identification information and outputs the camera's identification information to the transceiver 31. The transceiver 31 transmits the camera's identification information received from the reading unit 35 to the work monitoring support device 1. Note that if the camera's two-dimensional code, which will be described later, is not used to associate the camera with the worker, the portable terminal 3 does not need to include the reading unit 35.

[0023] The alarm device 4 is a device carried by a worker, and issues an alarm by, for example, at least one of sound, light, and vibration. Although one alarm device 4 is illustrated in FIG. 1 , there may be multiple alarm devices 4. The alarm device 4 issues an alarm when it receives information from the work monitoring support device 1 instructing the issuance of an alarm as unsafe information. The alarm may be directed to the worker, or may be directed to another person such as a supervisor, or may be directed to both the worker and another person.

[0024] The supervisor terminal 5 is a terminal such as a smartphone, tablet, or personal computer carried by a supervisor such as a site supervisor or work chief who monitors the work site. Although one supervisor terminal 5 is illustrated in FIG. 1 , there may be multiple supervisor terminals 5. The configuration of the supervisor terminal 5 is similar to that of the portable terminal 3, for example. Here, an example will be described in which the configuration of the supervisor terminal 5 is similar to that of the portable terminal 3, but the configuration of the supervisor terminal 5 may be different from that of the portable terminal 3. For example, the supervisor terminal 5 does not need to include a vibration generating unit or a reading unit.

[0025] The management device 6 is a device installed in an office such as a site office, the office of the business operator to which the worker belongs, or the office of a construction management company such as a general contractor, and manages work. While FIG. 1 illustrates one office and one management device 6, a management device 6 may be installed in each of the site office, the office of the business operator to which the worker belongs, and the office of the construction management company, or multiple management devices 6 may be installed in one office. The management device 6 may manage, for example, worker information indicating the name and affiliation of the worker performing the work, progress information indicating the progress of the work, and drawing data such as construction drawings for each process of the work. The worker information may include identification information for the mobile terminal 3 carried by the worker.

[0026] In Figure 1, the destinations to which unsafe information can be sent by the work monitoring support device 1 are shown as a mobile terminal 3, an alarm device 4, a supervisor terminal 5, and a management device 6, but the destinations to which unsafe information can be sent by the work monitoring support device 1 do not have to be all of these; it can be one or more of these.

[0027] FIG. 2 is a diagram illustrating the sensor 2 according to the present embodiment. In the example shown in FIG. 2 , two cameras 21 attached to a helmet 7 worn by a worker are used as the sensor 2. The cameras 21 are, for example, 360-degree cameras capable of capturing images in all directions. Hereinafter, an example in which the cameras 21 are 360-degree cameras will be described. However, the cameras 21 may be wide-angle cameras, such as 180-degree cameras capable of capturing images in 180 degrees, or cameras other than wide-angle cameras; they are not limited to 360-degree cameras. In FIG. 2 , two cameras 21 are attached, one on each side of the helmet 7, but the positions and number of the cameras 21 are not limited to the example shown in FIG. 2 . The number and position of the cameras 21 may be one. The angle of view, number, and arrangement of the cameras 21 may be determined so as to detect an unsafe state of the detection target. For example, the cameras 21 may be wearable cameras, such as goggles or wristwatches, or may be cameras installed at the work site. In addition, both a camera 21 worn by a worker and a camera 21 set up at the work site may be used.

[0028] As shown in FIG. 2, when a 360-degree camera attached to the worker's helmet 7 is used as the camera 21, the camera 21 can capture an image of the worker regardless of the worker's position. Therefore, the work monitoring support system of this embodiment can prevent unsafe conditions of workers from being missed due to blind spots while limiting the number of cameras 21 used.

[0029] For example, when working at heights, as shown in FIG. 2 , workers are required to wear a safety belt 8, and a hook 9 of the safety belt 8 is attached to an attachment facility such as a lifeline 200. In the example shown in FIG. 2 , the work monitoring support device 1 detects, using image data acquired by a camera 21, whether the environment around the worker is an environment where the lifeline 200, which is an example of an unsafe environment, is installed. The surroundings of the worker may be, for example, a range where the distance from the worker is equal to or less than a predetermined value, but is not limited to this. The range where the distance from the worker is equal to or less than a predetermined value is determined, for example, based on sensor information. The work monitoring support device 1 also detects, using image data acquired by the camera 21, whether the worker is in an unsafe state, more specifically, whether the worker is not wearing the safety belt 8, which is an example of unsafe behavior. In the example shown in FIG. 2 , the image data acquired by the camera 21 is used for both detecting an unsafe environment and detecting an unsafe state. For example, when the work monitoring support device 1 detects that the environment around a worker is an environment in which a lifeline 200 is installed and that the worker is not wearing a safety belt 8, it determines that an unsafe event has occurred to the worker and outputs unsafety information indicating the determination result. The unsafety information may be, for example, information indicating an alarm to the worker who needs to take action to eliminate the unsafety, or may be information to notify a person other than the worker, such as a supervisor, of the need to take action to eliminate the unsafety.

[0030] As described above, the sensor 2 is not limited to the camera 21. For example, by using a vital sensor as the sensor 2, the work monitoring support device 1 may use vital data to detect a poor physical condition, which is an example of an unsafe condition, of a worker. For example, the sensor 2 may include a position detection sensor that can detect the position of the worker, and the work monitoring support device 1 may use sensor information acquired by the position detection sensor to detect that the surroundings of the worker are at a high altitude, which is an example of an unsafe environment, and when detecting that the worker is in a poor physical condition, determine that an unsafe event has occurred to the worker and output unsafety information indicating the determination result.

[0031] As described above, the unsafe environment is not limited to high places, but may also be an environment on steel beams, an environment with holes, an environment where the floor is not fully tightened, an environment where the life rope 200 is installed, or other environments. When a poor physical condition is detected, the work monitoring support device 1 may identify the worker as a worker who needs to take action to eliminate the unsafe condition, regardless of whether the environment around the worker is unsafe, and may output unsafety information based on the identification result.

[0032] Furthermore, the poor physical condition is not limited to being detected using vital data, but may also be detected using image data. The work monitoring support device 1 may detect the poor physical condition using both vital data and image data. For example, the sensor 2 may include at least one of a sensor that detects temperature and a sensor that detects humidity, and the work monitoring support device 1 may detect a worker suffering from heatstroke using the results detected by these sensors and at least one of image data and vital data acquired by the camera 21. A worker suffering from heatstroke is an example of a worker in poor physical condition.

[0033] The definition of an unsafe environment may be common regardless of the type of unsafe condition, such as not wearing the safety belt 8, jumping over, or being in poor physical condition, or the definition of the corresponding unsafe environment may be different for each type of unsafe condition.

[0034] Next, the operation of this embodiment will be described. FIG. 3 is a flowchart showing an example of a work monitoring support process procedure in the work monitoring support device 1 of this embodiment. As shown in FIG. 3, the work monitoring support device 1 associates the worker with the sensor 2 (step S1). Specifically, the correspondence management unit 16 associates the worker with the sensor 2. Note that, although an example in which the worker wears the sensor 2 will be described here, if a sensor worn by the worker is not used as the sensor 2, for example, step S1 may not be performed. Note that "worn by the worker" includes both a case in which the worker wears the sensor directly, such as a wristwatch-type sensor, and a case in which the worker wears the sensor indirectly, such as when the camera 21 is attached to the helmet 7 as shown in FIG. 2.

[0035] The association between the sensors 2 and the workers may be performed in any manner. For example, a sensor 2 to be used may be assigned to each worker in advance, and the correspondence management unit 16 of the work monitoring support device 1 may acquire the association between the worker and the identification information of the assigned sensor 2 by receiving input from an operator or the like, and store the acquired information in the information storage unit 12 as part of the worker information. Alternatively, the correspondence management unit 16 may store association information indicating the association between the worker and the identification information of the assigned sensor 2 in the information storage unit 12 separately from the worker information. Furthermore, the transmitter / receiver 11 of the work monitoring support device 1 may receive information indicating the association between the worker and the identification information of the assigned sensor 2 from another device such as the management device 6, and the correspondence management unit 16 may receive the information from the transmitter / receiver 11, thereby associating the worker with the identification information of the assigned sensor 2.

[0036] Alternatively, the sensor 2 may be associated with the worker when the work is being performed. For example, the sensor 2 may be loaned to the worker at the work site before the work begins, and the loaned sensor 2 may be associated with the worker. The loaned sensor 2 may be associated with the worker by, for example, having the worker read a two-dimensional code affixed to the sensor 2 with the mobile terminal 3.

[0037] FIG. 4 is a diagram illustrating an example of a method for associating a sensor 2 with a worker according to the present embodiment. In the example illustrated in FIG. 4 , a camera 21 attached to a helmet 7 as the sensor 2 is loaned to a worker at a work site. A sticker 22 having a two-dimensional code printed thereon is affixed to the camera 21. Note that the two-dimensional code need only be configured to be displayed on the exterior of the camera 21, and is not limited to being affixed by the sticker 22. For example, the two-dimensional code may be printed on the camera 21 itself. The two-dimensional code includes identification information of the camera 21, such as the number of the camera 21. In the following, an example will be described in which the identification information of the camera 21 is the number of the camera 21 (hereinafter also referred to as the camera number), but the identification information of the camera 21 is not limited thereto. Note that while FIG. 4 illustrates one camera 21, when multiple cameras 21 are used, as in the example illustrated in FIG. 2 , two-dimensional codes corresponding to different camera numbers may be displayed for each camera 21, or the same camera number may be displayed for multiple cameras 21 as a group.

[0038] The worker has the mobile device 3 read the two-dimensional code displayed on the provided camera 21, i.e., the two-dimensional code displayed on the sticker 22. The reader 35 of the mobile device 3 reads the two-dimensional code and outputs the read result to the transmitter / receiver 31, which then transmits it to the work monitoring support device 1. When the correspondence management unit 16 of the work monitoring support device 1 receives the read result from the mobile device 3 via the transmitter / receiver 11, it associates the worker with the camera 21 using the camera 21 number included in the read result. For example, if identification information of the mobile device 3 corresponding to the worker is stored in the information storage unit 12 as worker information, the correspondence management unit 16 may associate the camera 21 with the worker based on the identification information of the mobile device 3 added to the read result received from the mobile device 3. The identification information of the mobile device 3 may be, for example, a Media Access Control (MAC) address.

[0039] Alternatively, application software (hereinafter also referred to as an app) for using the camera 21 may be installed on the mobile terminal 3, the reading unit 35 may be implemented by the app, and the worker may log in to the app and then have the reading unit 35 read the two-dimensional code. In this case, the work monitoring support device 1 stores user identification information (hereinafter also referred to as a user ID (ID)) used for logging in as worker information in the information storage unit 12, and the mobile terminal 3 transmits the read result together with the user ID to the work monitoring support device 1. This allows the correspondence management unit 16 to associate the worker with the camera number using the user ID. Alternatively, instead of using an app, the worker may access a web page for using the camera 21 and log in on the web page, thereby performing association using the user ID.

[0040] Alternatively, for example, if an authentication terminal (not shown) for performing facial recognition is installed at the work site, the results of the facial recognition may be used to associate camera numbers with workers. For example, facial image data associated with workers may be stored in advance as facial image information in the information storage unit 12 of the work monitoring support device 1, and the correspondence management unit 16 of the work monitoring support device 1 may also function as an authentication unit that authenticates workers. In this case, the authentication terminal photographs the worker's face and transmits the image data obtained by photographing to the work monitoring support device 1. The correspondence management unit 16 then compares the image data received from the authentication terminal via the transmission / reception unit 11 with the facial image information stored in the information storage unit 12 to identify the worker corresponding to the image data.

[0041] If the correspondence management unit 16 can associate the image data received from the authentication terminal with the facial image information stored in the information storage unit 12, i.e., if the worker corresponding to the image data can be identified, the correspondence management unit 16 determines that the authentication is successful. However, if the image data received from the authentication terminal does not match any of the facial image information stored in the information storage unit 12, the correspondence management unit 16 determines that the authentication is unsuccessful. The correspondence management unit 16 transmits the authentication result to the authentication terminal via the transceiver unit 11. For example, assume that the camera number is displayed on the camera 21 and the authentication terminal has an input means such as a keyboard or a touch panel. Upon receiving the authentication result indicating successful authentication, the authentication terminal accepts input of the camera number from the worker and transmits the accepted camera number to the work monitoring support device 1. Upon receiving the camera number via the transceiver unit 11, the correspondence management unit 16 associates the camera number with the successfully authenticated worker and stores the association result in the information storage unit 12. The association result may be stored as correspondence information as described above, or may be stored in the information storage unit 12 as worker information. Note that the facial image data used for authentication may be included in the worker information.

[0042] Furthermore, the camera number is not limited to being input to the authentication terminal, but may be transmitted from the authentication terminal to the work monitoring support device 1 by reading a two-dimensional code, as in the above-described example. For example, the authentication terminal may have the same function as the reading unit 35 of the mobile terminal 3, and the worker may have the authentication terminal read the two-dimensional code displayed on the loaned camera 21, and the authentication terminal may transmit the read result to the work monitoring support device 1.

[0043] Although an example in which the work monitoring support device 1 performs the authentication process has been described above, the authentication terminal may store facial image information and perform the authentication process. In this case, the authentication terminal transmits identification information indicating the worker identified by facial recognition to the work monitoring support device 1. The authentication process may also be performed by the management device 6 or another device not shown. In this case, the device that performed the authentication process also transmits identification information indicating the worker identified by facial recognition to the work monitoring support device 1.

[0044] FIG. 5 is a diagram illustrating an example of worker information according to the present embodiment. In the example illustrated in FIG. 5 , the worker information also includes correspondence information associating the worker with the camera 21. The worker information includes terminal identification information, which is identification information for the mobile terminal 3 carried by the worker, the worker (the worker's identification information), an alarm device number, which is identification information for the alarm device 4 carried by the worker, and a camera number. The worker identification information may be the worker's name, a worker number assigned to each worker, the user ID described above, or other information. In addition to the information illustrated in FIG. 5 , the worker information may also include the worker's affiliation. Furthermore, while FIG. 5 illustrates an example in which the worker carries an alarm device 4, if the worker does not carry an alarm device 4, the worker information does not need to include the alarm device number. Below, an example is described in which the worker information includes correspondence information associating the worker with the camera 21. However, as described above, the correspondence information may be stored separately from the worker information in the information storage unit 12.

[0045] The association between the camera 21 and the worker described above is an example, and the method of associating the camera 21 and the worker is not limited to the above example. Also, when a sensor other than the camera 21 is used as the sensor 2, the sensor 2 and the worker are similarly associated.

[0046] Returning to the explanation of Fig. 3, the work monitoring support device 1 acquires sensor information (step S2). In detail, for example, the transmitter / receiver 11 receives the sensor information from the sensor 2 and stores the received sensor information in the information storage unit 12. The sensor information is stored in the information storage unit 12 together with identification information of the sensor 2 that is the device source of the sensor information. For example, if the sensor 2 is the above-mentioned camera 21, image data that is the sensor information is stored in the information storage unit 12 in association with the camera number.

[0047] Next, the work monitoring support device 1 performs an unsafe environment detection process (step S3). Specifically, the first detection unit 13 detects an unsafe environment by using the sensor information stored in the information storage unit 12 to determine whether the environment around the worker corresponding to the sensor information is an unsafe environment, and outputs the detection result to the determination unit 15. For example, the unsafe environment may be determined by identifying, from the image data of the sensor information, a sticker or two-dimensional code that serves as a marker attached to an area previously determined by a person to be an unsafe environment. Alternatively, if no such marker is present, the unsafe environment may be determined by identifying, from the image data, a hole or other factor that could make the environment unsafe for the worker. Alternatively, the unsafe environment may be determined from the image data using a trained model or the like without identifying markers or factors. Alternatively, the unsafe environment may be determined by identifying markers and factors from the image data using a trained model or the like. Furthermore, the first detection unit 13 identifies the worker corresponding to the sensor information using a sensor number corresponding to the sensor information and the worker information, and includes the identification information of the identified worker in the detection result.

[0048] As described above, the unsafe environment is at least one of an environment where the main rope 200 is installed, an environment where the height is equal to or greater than a threshold value, an environment where a handrail is installed, an environment where the floor is not fully tightened, etc., but is not limited to these. Furthermore, the definition of the unsafe environment may differ depending on the type of unsafe condition to be detected in the unsafe condition detection process of step S4 described below. If there are multiple types of unsafe conditions to be detected in the unsafe condition detection process of step S4 and the definition of the unsafe environment differs depending on the type of unsafe condition, in step S3, the first detection unit 13 performs processing to detect unsafe environments corresponding to all types of unsafe conditions. In this case, the first detection unit 13 includes an environment indicating the type of the detected unsafe environment in the detection result. The type of unsafe environment is, for example, an environment where the main rope 200 is installed, an environment where the height is equal to or greater than a threshold value, an environment where a handrail is installed, an environment where the floor is not fully tightened, etc.

[0049] The work monitoring support device 1 performs a process for detecting an unsafe state (step S4). Specifically, the second detection unit 14 detects an unsafe state by performing a process for determining whether the worker is in an unsafe state and outputs the detection result to the determination unit 15. The second detection unit 14 also identifies the worker corresponding to the sensor information using the sensor number corresponding to the sensor information and the worker information, and includes identification information of the identified worker in the detection result. The order of steps S3 and S4 may be reversed, or steps S3 and S4 may be performed simultaneously. Furthermore, step S4 may not be performed for a worker whose surroundings are determined not to be in an unsafe environment in step S3. Furthermore, if the order of steps S3 and S4 is reversed, step S3 may not be performed for a worker whose surroundings are determined not to be in an unsafe state in step S4. If either step S3 or step S4 has not been performed, the determination unit 15 determines that an unsafe event has not occurred in step S5, which will be described later.

[0050] If a sensor 2 is installed at the work site, in steps S3 and S4, the first detection unit 13 and the second detection unit 14 may identify the worker using sensor information acquired by the sensor 2. For example, if the sensor information is image data, the first detection unit 13 and the second detection unit 14 may identify the worker by face recognition processing using the image data and facial image data. Alternatively, if the portable terminal 3 has a location identification function for identifying its own location, the portable terminal 3 may transmit location information indicating its location to the work monitoring support device 1, and the first detection unit 13 and the second detection unit 14 may identify the worker using the location information and the image data. In this case, the portable terminal 3 also functions as the sensor 2, acquiring location information as sensor information. Alternatively, a location detection device that identifies its own location and transmits location information may be used as the sensor 2, separate from the portable terminal 3, and the worker may carry the location detection device.

[0051] The detection process in the first detection unit 13 and the second detection unit 14 may be performed by machine learning, for example, when the sensor information is image data. Fig. 6 is a diagram showing an example of the configuration of the first detection unit 13 and the second detection unit 14 according to this embodiment. Fig. 6 shows an example of the configuration of the first detection unit 13 and the second detection unit 14 when the detection process in the first detection unit 13 and the second detection unit 14 is performed by machine learning.

[0052] 6 , the first detection unit 13 includes a model generation unit 131 that generates a first trained model for inferring whether or not an environment is unsafe using features based on sensor information, and an inference unit 132 that infers whether or not an environment is unsafe by inputting the features based on the sensor information into the first trained model and obtaining an output of the first trained model. The features may be the sensor information itself, or may be features extracted from the sensor information by performing preprocessing determined from the sensor information.

[0053] The model generation unit 131 generates the first trained model, for example, by supervised machine learning. The model generation unit 131 generates the first trained model, for example, by using multiple data sets including feature amounts based on image data and detection results indicating whether an environment is unsafe, which are correct answer data. The first trained model is stored in the information storage unit 12. Alternatively, the model generation unit 131 may generate the first trained model using, as correct answer data, not only the detection results indicating whether an environment is unsafe, but also the type of the detected unsafe environment.

[0054] As machine learning, for example, a neural network such as deep learning, a support vector machine, etc. may be used, but is not limited to these. Furthermore, an algorithm other than supervised machine learning may also be used as machine learning.

[0055] The image data used for learning in the model generation unit 131 may be sensor information (image data) previously acquired by the sensor 2, image data acquired by a device other than the sensor 2, or both. Furthermore, the image data used for learning may include image data generated by simulation, generative AI (artificial intelligence), or the like, or may include image data generated by the generative AI. Examples of the generative AI include Gemini (registered trademark), GPT (Generative Pre-trained Transformer)-4V, GPT-3, a large-scale language model, and Stable Diffusion, an image generation AI. However, any suitable AI capable of generating image data may be used, and is not limited to these. For example, image data may be generated by inputting text or voice data describing a situation indicating an unsafe environment or an actual accident record into the generative AI.

[0056] The inference unit 132 obtains a detection result of whether or not the environment is unsafe as an output of the first trained model by inputting features based on the sensor information stored in the information storage unit 12 to the first trained model. When learning is performed by including the type of unsafe environment in the correct answer data, the inference unit 132 can also obtain the type of unsafe environment by inputting features based on the sensor information stored in the information storage unit 12.

[0057] 6 , the second detection unit 14 includes a model generation unit 141 that generates a second trained model for inferring whether or not an unsafe state exists using features based on the sensor information, and an inference unit 142 that infers whether or not an unsafe state exists by inputting the features based on the sensor information into the second trained model and obtaining an output of the second trained model. The features may be the sensor information itself, or may be features extracted from the sensor information by performing preprocessing determined from the sensor information.

[0058] The model generation unit 141 generates the second trained model, for example, by supervised machine learning. The model generation unit 141 generates the second trained model, for example, by using multiple data sets including feature quantities based on sensor information and detection results indicating whether or not an unsafe state exists, which are correct data. Note that the sensor information used for training may be sensor information previously acquired by the sensor 2, or may be image data acquired by a device other than the sensor 2. The generated second trained model is stored in the information storage unit 12. Alternatively, the model generation unit 141 may generate the second trained model using, as correct data, not only the detection results indicating whether or not an unsafe state exists, but also the type of the detected unsafe state.

[0059] For example, but not limited to, a neural network such as deep learning, a support vector machine, etc. may be used as the machine learning in the model generation unit 141. Furthermore, an algorithm other than supervised machine learning may also be used as the machine learning.

[0060] Note that, like the image data used for learning in the model generation unit 131, the image data used for learning in the model generation unit 141 may be sensor information (image data) previously acquired by the sensor 2, image data acquired by a device other than the sensor 2, or may include both of these, or may include image data generated by simulation, generation AI, or the like. For learning purposes, unsafe behavior may be intentionally performed, and image data of the unsafe behavior captured may be used as learning data. Furthermore, because unsafe situations do not occur frequently, for example, image data of actual unsafe behaviors that have occurred in the past is limited. However, the amount of image data for learning can be increased by generating image data using the generation AI based on, for example, written records of the situation or audio descriptions of people who heard the situation through hearsay.

[0061] The inference unit 142 obtains a detection result of whether or not an unsafe state exists as an output of the second trained model by inputting, into the second trained model, feature quantities based on the sensor information stored in the information storage unit 12. When learning is performed by including the type of unsafe state in the ground truth data, the inference unit 142 can also obtain the type of unsafe state by inputting feature quantities based on the sensor information stored in the information storage unit 12.

[0062] Furthermore, when the same sensor information is used in the detection by the first detection unit 13 and the detection by the second detection unit 14, a detection unit having the functions of both the detection by the first detection unit 13 and the detection by the second detection unit 14 may be provided, and the detection unit may detect both the unsafe environment and the unsafe state. In this case, for example, a first trained model for detecting the unsafe environment and a second trained model for detecting the unsafe state may be generated separately, or both the detection results of the unsafe environment and the detection results of the unsafe state may be included as correct data during training, so that one trained model outputs both the detection results of the unsafe environment and the detection results of the unsafe state.

[0063] 6 shows an example in which the model generation unit 131 and the model generation unit 141 are provided within the work monitoring support device 1, but this is not limiting, and at least one of the model generation unit 131 and the model generation unit 141 may be provided in a learning device (not shown) outside the work monitoring support device 1. In this case, at least one of the first trained model and the second trained model generated by the learning device is stored in the information storage unit 12.

[0064] Note that the detection processes in the first detection unit 13 and the second detection unit 14 may be performed in any specific manner as long as they are capable of detecting the respective detection targets according to the type of sensor 2. Even when image data is used as sensor information, at least one of the detection processes in the first detection unit 13 and the second detection unit 14 may be performed by pattern recognition or the like, not limited to machine learning. Furthermore, for example, one of the detection processes in the first detection unit 13 and the second detection unit 14 may be performed by machine learning, and the other may be performed by a method other than machine learning. For example, the detection of an unsafe environment may be performed by machine learning using image data, and the detection of an unsafe state may be performed using information indicating whether or not the safety belt 8 is worn, as determined by a hook sensor.

[0065] Returning to the description of FIG. 3 , the work monitoring support device 1 determines whether both an unsafe environment and an unsafe state have been detected (step S5). Specifically, the determination unit 15 determines whether both an unsafe environment and an unsafe state have been detected for the same worker using the detection results of the first detection unit 13 and the second detection unit 14. As illustrated in FIG. 2 , when a sensor 2 worn by a worker is used to detect both an unsafe environment and an unsafe state, the same sensor information is used for the detection by the first detection unit 13 and the detection by the second detection unit 14. Therefore, the determination unit 15 can determine whether both an unsafe environment and an unsafe state have been detected using the detection results using the same sensor information. When different sensors 2 are used for the detection by the first detection unit 13 and the detection by the second detection unit 14, the determination unit 15 determines whether both an unsafe environment and an unsafe state have been detected for each worker using the detection results associated with the same worker.

[0066] In addition, when there are multiple types of unsafe states to be detected in the detection of an unsafe state and the definition of an unsafe environment differs depending on the type of unsafe state, in step S5, the determination unit 15 determines whether or not a type of unsafe state corresponding to the type of unsafe state detected has been detected. That is, the determination unit 15 determines whether or not an unsafe state has been detected and an unsafe environment corresponding to the unsafe state (an unsafe environment that will cause an unsafe event in combination with the unsafe state) has been detected.

[0067] Furthermore, when detecting an unsafe environment and an unsafe state through machine learning using image data as sensor information, the determination result itself, indicating whether both an unsafe environment and an unsafe state have been detected, may be learned. In this case, the first detection unit 13, the second detection unit 14, and the determination unit 15 are not provided separately, and the unsafety determination unit 17 includes a model generation unit and an inference unit. The model generation unit generates a trained model using multiple data sets including feature values ​​based on the sensor information and determination results (determination results indicating whether both an unsafe environment and an unsafe state have been detected) as correct data. The inference unit obtains the determination result by inputting the feature values ​​based on the sensor information stored in the information storage unit 12 into the trained model. This model generation unit may also be provided in a learning device separate from the work monitoring support device 1.

[0068] Furthermore, for example, a human presence sensor and a hook sensor may be used as the sensor 2. For example, the human presence sensor may be installed on the main rope 200 or in the vicinity of the main rope 200. In this case, if a person is detected by the human presence sensor, the detected person is in an unsafe environment. Therefore, the detection result of the human presence sensor is used in the detection process of the first detection unit 13. Furthermore, the detection result of the hook sensor is used in the detection process of the second detection unit 14. Note that the detection result of the human presence sensor alone cannot determine which workers are in the environment where the main rope 200 is installed. Therefore, for example, the determination unit 15 may determine that an unsafe event has occurred when, even though a person is detected by the human presence sensor, the detection result indicated by the hook sensor indicates that no one is wearing a safety belt 8. Alternatively, if the sensor 2 includes a sensor for detecting the position of a worker, the first detection unit 13 may identify the worker detected by the human presence sensor based on the position of the worker and determine that the area around the identified worker is in an unsafe environment. When a hook sensor is used, the hook sensor is associated with a worker, and information indicating the association between the hook sensor and the worker is stored in the information storage unit 12, for example, as part of the worker information.

[0069] If the work monitoring support device 1 determines No in step S5, it repeats the process from step S2. If the work monitoring support device 1 determines Yes in step S5, it outputs the determination result (step S6) and repeats the process from step S2. In detail, in step S6, for example, the determination unit 15 generates unsafety information (hereinafter also referred to as an alarm) including information indicating an alarm to a worker in which both an unsafe environment and an unsafe state have been detected, and transmits the generated unsafety information via the transceiver unit 11 to at least one of the alarm device 4 and the mobile terminal 3 corresponding to the worker. When the alarm is transmitted to the mobile terminal 3, the mobile terminal 3 functions as an alarm device. In other words, the mobile terminal 3 is also an example of an alarm device that issues an alarm to a worker. As described above, the unsafety information includes, for example, at least one of display data, audio data, and vibration instruction data. Furthermore, the determination unit 15 may further generate unsafety information including information indicating a worker for whom both an unsafe environment and an unsafe state have been detected, and transmit the generated unsafety information to at least one of the monitor terminal 5 and the management device 6 via the transmission / reception unit 11. Hereinafter, the transmission of the unsafety information by the determination unit 15 via the transmission / reception unit 11 will also be referred to as the determination unit 15 transmitting the unsafety information.

[0070] The determination unit 15 may generate the display data and audio data for the unsafety information using a generation AI. For example, a trained model generated by machine learning that generates sentences using input image data may be used as the model for the generation AI, and the determination unit 15 may input image data including an event determined to be an unsafe event into the trained model and instruct it to create sentences indicating a warning. In this way, the determination unit 15 may use the generation AI to generate sentences indicating a warning according to the situation.

[0071] The content of the unsafety information may differ depending on the device to which the unsafety information is sent. The transmission of the unsafety information to at least one of the alarm device 4 and the mobile terminal 3 and the transmission of the unsafety information to the monitor terminal 5 and the management device 6 may be performed simultaneously or at different times. For example, if the unsafe event is resolved within a certain time after the transmission of the unsafety information to at least one of the alarm device 4 and the mobile terminal 3, the determination unit 15 may not transmit the unsafety information to the monitor terminal 5 and the management device 6. The determination unit 15 may transmit the unsafety information to the monitor terminal 5 and the management device 6 when the unsafe event is not resolved even after transmitting the unsafety information to at least one of the alarm device 4 and the mobile terminal 3. Alternatively, the determination unit 15 may transmit the unsafety information to the monitor terminal 5 when the unsafe event is not resolved even after transmitting the unsafety information to at least one of the alarm device 4 and the mobile terminal 3, and further transmit the unsafety information to the management device 6 when the unsafe event is not resolved within a specified time.

[0072] In addition, when the judgment unit 15 determines that an unsafe event has occurred, if the unsafe condition corresponding to the unsafe event that has occurred is a state of poor health, it may send unsafe information informing the worker of the unsafe condition to at least one of the mobile terminals 3 and the alarm devices 4 of workers other than the worker determined to be in a state of poor health.

[0073] FIG. 7 is a diagram showing an example of an alarm displayed on the mobile device 3 according to the present embodiment. In step S6, the determination unit 15 may generate, as the unsafety information to be sent to the mobile device 3, unsafety information including, for example, display data for displaying the display screen shown in FIG. 7 and audio data indicating that the safety belt 8 is not being worn. As a result, the mobile device 3 displays the alarm screen shown in FIG. 7 and outputs audio indicating that the safety belt 8 is not being worn. Alternatively, in step S6, the determination unit 15 may generate, as the unsafety information to be sent to the mobile device 3, unsafety information including display data for displaying the display screen shown in FIG. 7 and vibration instruction data for vibrating the mobile device 3. As a result, the mobile device 3 displays the alarm screen shown in FIG. 7 and vibrates the mobile device 3 to notify the worker that an alarm has been issued. FIG. 7 is an example, and the display screen showing the alarm is not limited to the example shown in FIG. 7. When the work monitoring support device 1 detects multiple types of unsafe conditions, the determination unit 15 may generate the content of the alarm according to the type of the detected unsafe condition.

[0074] While FIG. 7 illustrates an example in which unsafety information indicating an alarm is transmitted to the mobile device 3, as described above, the unsafety information may also be transmitted to the alarm device 4. Upon receiving the unsafety information, the alarm device 4 issues an alarm by at least one of sound, vibration, and display. When the mobile device 3 and the alarm device 4 issue an alarm by sound, if the worker is startled, the worker may make a sudden movement, which may put the worker in an unsafe state or affect the work. Therefore, the determination unit 15 may generate sound data so that the mobile device 3 and the alarm device 4 output a sound that does not startle the worker as much as possible. For example, the determination unit 15 may instruct the generation AI to generate sound data that does not startle the worker, thereby generating sound data indicating an alarm using the generation AI. In this case, the generation AI may be trained in advance to learn information such as what situations will startle the worker and what situations will not startle the worker. Furthermore, the generation AI may learn information for each worker such as under what circumstances the worker will be startled and under what circumstances the worker will not be startled, and the determination unit 15 may specify a worker and generate the voice data using the generation AI when generating the voice data. Similarly, for the display data and vibration instruction data, the generation AI may learn information so that the content of the alarm will be appropriate for each worker, thereby making the content of the alarm appropriate for each worker.

[0075] 8 is a diagram showing an example of an alarm displayed on the monitor terminal 5 of this embodiment. In step S6, the determination unit 15 may generate, as the unsafety information addressed to the monitor terminal 5, for example, display data for displaying the display screen shown in FIG. 8 and audio data indicating that there is a worker who is not wearing the safety belt 8. In the example shown in FIG. 7, since the alarm is notified to the worker himself, it is not necessary to indicate which worker is not wearing the safety belt 8. However, the display screen displayed on the monitor terminal 5 displays information indicating the worker who is not wearing the safety belt 8 (Mr. A in the example shown in FIG. 8). This allows the monitor to identify the worker who is not wearing the safety belt 8.

[0076] FIG. 9 is a diagram illustrating an example of an alarm displayed on the management device 6 according to the present embodiment. In step S6, the determination unit 15 may generate unsafety information addressed to the management device 6, for example, including display data for displaying the display screen shown in FIG. 9 and audio data indicating that a worker is not wearing a safety belt 8. In the example shown in FIG. 9, similar to the example shown in FIG. 8, information indicating the worker not wearing the safety belt 8 (Mr. A in the example shown in FIG. 9) is also displayed. In the example shown in FIG. 9, in addition to information indicating the worker, information such as the time, site name, work area, director, director's contact information, construction management company representative, construction management company representative contact information, work supervisor, and work supervisor contact information is also displayed. For example, this information may be stored in the information storage unit 12 as worker information, and the determination unit 15 may use the worker information to include this information in the unsafety information. By displaying detailed information on the management device 6, a manager who has checked the management device 6 can take various measures. Furthermore, image data captured when an unsafe event occurred may be displayed as a still image or a video in the image display area 301.

[0077] Note that FIG. 3 illustrates an example in which the work monitoring support device 1 outputs unsafety information indicating the determination result when an unsafe event occurs, i.e., when it is determined that both an unsafe environment and an unsafe state have been detected. That is, FIG. 3 illustrates an example in which the unsafety information is output as an alarm, but this is not limited to this. The unsafety information may also be output periodically. For example, information indicating whether or not an unsafe event has occurred for each worker may be periodically transmitted as unsafety information to at least one of the supervisor terminal 5 and the management device 6. For example, the determination unit 15 may generate display data so that information indicating whether or not an unsafe event has occurred for each worker is displayed in a table format, and transmit unsafety information including the generated display data to at least one of the supervisor terminal 5 and the management device 6. Alternatively, the determination unit 15 may generate display data so that information indicating whether or not an unsafe event has occurred for each worker is displayed as an image, and transmit unsafety information including the generated display data to at least one of the supervisor terminal 5 and the management device 6.

[0078] FIG. 10 is a diagram illustrating an example of the display of unsafety information according to this embodiment. In the example illustrated in FIG. 10 , the determination unit 15 generates display data so that information indicating whether or not an unsafe event has occurred for each worker is displayed as an image, and transmits the unsafety information including the generated display data to the supervisor terminal 5. As a result, the supervisor terminal 5 displays information indicating whether or not an unsafe event has occurred for each worker as an image, as illustrated in FIG. 10 . In the example illustrated in FIG. 10 , position information indicating the positions of the workers is acquired as sensor information, the work monitoring support device 1 grasps the positions of each worker, and drawing data of the work site is stored in the information storage unit 12. The determination unit 15 generates an image in which circles indicating the positions of each worker are superimposed on a drawing represented by the drawing data of the work site, and changes the color of the circle corresponding to each worker depending on whether or not an unsafe event has occurred. In the example illustrated in FIG. 10 , the unhatched circles and the hatched circles are different colors, and the unhatched circles indicate workers who have not experienced an unsafe event, while the hatched circles indicate workers who have experienced an unsafe event. In the example shown in Fig. 10, the name of the worker in whom the unsafe event occurred is also displayed. Fig. 10 is just an example, and the specific display mode and displayed information are not limited to the example shown in Fig. 10. The names of all workers may also be displayed. Furthermore, the type of unsafe condition corresponding to the unsafe event (such as not wearing a safety belt, jumping over, or being in poor physical condition) may also be displayed.

[0079] 10, information indicating whether or not an unsafe event has occurred is displayed on the monitor terminal 5. However, this is not limiting. For example, a light-emitting alarm device may be attached to the worker's helmet 7 as the alarm device 4, and the color of the light emitted by the alarm device 4 may indicate whether or not an unsafe event has occurred. For example, the determination unit 15 may transmit unsafety information indicating that the alarm device 4 of a worker who has not experienced an unsafe event will emit a green light, and transmit unsafety information indicating that the alarm device 4 of a worker who has experienced an unsafe event will emit a red light. This makes it easier for the monitor to monitor whether or not an unsafe event has occurred at the work site. Furthermore, the color emitted by the alarm device 4 may be changed depending on the type of unsafe state corresponding to the unsafe event.

[0080] Furthermore, in the above-described example, the first detection unit 13 detects an unsafe environment using sensor information. However, this is not limited to this. For example, an area determined to have resolved an unsafe state may be excluded from the detection target for an unsafe environment based on an instruction from a supervisor, manager, or the like. For example, the supervisor terminal 5 or the management device 6 includes an input means and receives an input from a supervisor, manager, or the like specifying a release area, which is an area to be excluded from the detection target for an unsafe environment. The supervisor terminal 5 or the management device 6 that receives the designation transmits information indicating the specified release area to the work monitoring support device 1. The first detection unit 13 of the work monitoring support device 1 excludes the specified release area from the detection target for an unsafe environment. Specifically, for example, the first detection unit 13 receives information indicating the release area via the transceiver unit 11. If the first detection unit 13 determines that the area around the worker is a release area using the sensor information, it outputs a detection result indicating that the area is not an unsafe environment to the determination unit 15. Furthermore, the release area is not limited to being input by a supervisor, manager, or the like, and may be determined based on progress information. For example, the first detection unit 13 may identify the release area using progress information, and if it determines that the area around the worker is a release area using sensor information, it may output a detection result to the judgment unit 15 that the environment is not unsafe.

[0081] Furthermore, the work monitoring support device 1 may accumulate data including a determination result when it is determined that an unsafe event has occurred and sensor information corresponding to the determination result as performance data. For example, the work monitoring support device 1 may accumulate performance data including sensor information when an unsafe event has occurred, detection results by the first detection unit 13 and the second detection unit 14, and determination results by the determination unit 15, by storing the data in the information storage unit 12 or another device. For example, by accumulating image data captured when an unsafe event has occurred as sensor information, the image data can be used for worker evaluation, improvement guidance, training of young workers, and the like.

[0082] For example, the determination unit 15 may use the performance data to generate presentation data for providing improvement guidance to a worker or for evaluating the worker, and the transmission / reception unit 11 may transmit the presentation data to the supervisor terminal 5 or another device (not shown) to output the presentation data. For example, image data captured when an unsafe event occurs may be included in the presentation data as sensor information, and the supervisor terminal 5 or another device (not shown) may display the image data as a video. This allows the performance data to be used for providing guidance to the worker. Furthermore, the determination unit 15 may use the performance data to count (count) the number of occurrences of unsafe events for each worker. Furthermore, the determination unit 15 may count the number of occurrences of unsafe events for each type of unsafe state for each worker. These counting results may be used for evaluating the worker, evaluating the organization to which the worker belongs, providing improvement guidance, etc.

[0083] In addition, in the unlikely event of an accident, the performance data can be used in reports to various organizations. The determination unit 15 may input the performance data to the generation AI and instruct it to generate sentences to be used in worker evaluations and reports to various organizations, thereby generating sentences to be used in worker evaluations and reports to various organizations.

[0084] Furthermore, the likelihood of an unsafe event occurring may differ depending on the environment of the work site. Therefore, analyzing performance data may enable understanding of the relationship between the environment of the work site and the likelihood of an unsafe event occurring. For example, the determination unit 15 may generate a trained model that infers the number of unsafe events occurring for each work site by performing machine learning using multiple data sets, each including image data of the work site and the ground truth data representing the number of unsafe events occurring in that area. The determination unit 15 may then infer the number of unsafe events occurring by inputting image data of the work site to be inferred into the trained model. For a work site where the inferred number of unsafe events occurring is equal to or greater than a threshold, the determination unit 15 may transmit, to the mobile device 3 of each worker, warning information that calls attention to the unsafe event before work begins. By displaying the warning information on the mobile device 3, workers can be warned in advance. Instead of performing machine learning for each work site, the work site may be divided into areas, and similar machine learning may be performed for each area to warn the worker.

[0085] Next, the hardware configuration of the work monitoring support device 1 of this embodiment will be described. In the work monitoring support device 1 of this embodiment shown in FIG. 1, a computer system executes a program, which is a computer program describing the processing to be performed by the work monitoring support device 1, causing the computer system to function as the work monitoring support device 1. FIG. 11 is a diagram showing an example configuration of a computer system that realizes the work monitoring support device 1 of this embodiment. As shown in FIG. 11, this computer system includes a control unit 101, an input unit 102, a memory unit 103, a display unit 104, a communication unit 105, and an output unit 106, which are connected via a system bus 107. The control unit 101 and the memory unit 103 form a processing circuit.

[0086] In FIG. 11 , the control unit 101 is a processor such as a CPU (Central Processing Unit) that executes a program describing the processing of the work monitoring support device 1 of this embodiment. Note that a portion of the control unit 101 may be realized by dedicated hardware such as a GPU (Graphics Processing Unit) or an FPGA (Field-Programmable Gate Array). The input unit 102 includes buttons, a keyboard, a mouse, a touchpad, and the like. The storage unit 103 includes various types of memory, such as a RAM (Random Access Memory) and a ROM (Read Only Memory), and a storage device, such as a hard disk, and stores programs to be executed by the control unit 101, necessary data obtained during processing, and the like. The storage unit 103 is also used as a temporary storage area for programs. The display unit 104 is, for example, a display, such as an LCD (Liquid Crystal Display). Note that the display unit 104 and the input unit 102 may be integrated and realized as a touch panel, for example. The communication unit 105 is a receiver and transmitter that perform communication processing. The output unit 106 is a speaker or the like. Note that Fig. 11 is just an example, and the configuration of the computer system is not limited to the example of Fig. 11. For example, the computer system that realizes the work monitoring support device 1 does not need to include the display unit 104 and the output unit 106.

[0087] Here, an example of the operation of the computer system until the program of this embodiment is ready to be executed will be described. In the computer system having the above configuration, for example, a computer program is installed into the storage unit 103 from a CD-ROM or DVD-ROM inserted in a CD (Compact Disc)-ROM drive or DVD (Digital Versatile Disc)-ROM drive (not shown). Then, when the program is executed, the program read from the storage unit 103 is stored in the main storage area of ​​the storage unit 103. In this state, the control unit 101 executes the processing as the work monitoring support device 1 of this embodiment in accordance with the program stored in the storage unit 103.

[0088] In the above description, a program describing the processing in the work monitoring support device 1 is provided using a CD-ROM or DVD-ROM as a recording medium, but this is not limited to this, and it is also possible to use a program provided via a transmission medium such as the Internet, depending on the configuration of the computer system, the capacity of the program to be provided, etc.

[0089] The program of this embodiment causes a computer system to, for example, use sensor information to perform a first detection process to detect whether the environment in which a worker working at a work site is located is an unsafe environment and a second detection process to detect whether the worker is in an unsafe state, and if the results of the first detection process and the second detection process indicate that the worker is located in an unsafe environment and is in an unsafe state, determine that an unsafe event has occurred, generate unsafe information indicating the determination result, and output the unsafe information.

[0090] The first detection unit 13, the second detection unit 14, the determination unit 15, and the response management unit 16 shown in FIG. 1 are realized by the control unit 101 shown in FIG. 11 executing a computer program stored in the storage unit 103 shown in FIG. 11. The storage unit 103 shown in FIG. 11 is also used to realize the first detection unit 13, the second detection unit 14, the determination unit 15, and the response management unit 16. The transmission / reception unit 11 shown in FIG. 1 is realized by the communication unit 105 shown in FIG. 11. The information storage unit 12 shown in FIG. 1 is part of the storage unit 103 shown in FIG. 11. The work monitoring support device 1 may be realized by multiple computer systems. Furthermore, for example, the work monitoring support device 1 may be realized by a cloud computer system.

[0091] The mobile terminal 3, the supervisor terminal 5, and the management device 6 shown in FIG. 1 are also realized by, for example, a computer system exemplified in FIG.

[0092] As described above, in this embodiment, when both an unsafe environment and an unsafe state of a worker are detected, the work monitoring support device 1 determines that an unsafe event has occurred and outputs unsafety information indicating the determination result. This allows the work monitoring support device 1 to provide more accurate information regarding unsafety than when determining the occurrence of an unsafe event by only detecting the unsafe state of a worker.

[0093] The configurations shown in the above embodiments are merely examples, and may be combined with other known technologies, or different embodiments may be combined with each other. It is also possible to omit or modify parts of the configurations as long as they do not deviate from the gist of the invention.

[0094] Various aspects of the present disclosure are summarized below as appendices.

[0095] (Supplementary Note 1) A work monitoring support device comprising: an unsafety determination unit that uses sensor information acquired by a sensor that detects the situation of a work site to perform a first detection process to detect whether an environment in which a worker working at the work site is present is an unsafe environment and a second detection process to detect whether the worker is in an unsafe state, and that determines that an unsafe event has occurred and generates unsafety information indicative of the determination result when results of the first detection process and the second detection process indicate that the worker is in an unsafe environment and is in an unsafe state; and an information output unit that outputs the unsafety information. (Supplementary Note 2) The work monitoring support device according to Supplementary Note 1, wherein the unsafe state includes a state in which the worker is engaging in unsafe behavior. (Supplementary Note 3) The work monitoring support device according to Supplementary Note 2, wherein the unsafe behavior includes the worker not wearing a safety belt. (Supplementary Note 4) The work monitoring support device according to Supplementary Note 3, characterized in that the sensors include a human presence sensor installed on equipment to which a safety belt hook is attached or in the vicinity of the equipment, and a hook sensor provided on the hook and detecting whether the hook has been attached. (Supplementary Note 5) The work monitoring support device according to any one of Supplements 2 to 4, characterized in that the unsafe behavior includes the worker jumping from one to another. (Supplementary Note 6) The work monitoring support device according to any one of Supplements 2 to 5, characterized in that the unsafe state includes the worker being in poor physical condition. (Supplementary Note 7) The work monitoring support device according to Supplementary Note 6, characterized in that the sensors include a vital sensor that acquires vital data of the worker. (Supplementary Note 8) The work monitoring support device according to any one of Supplements 1 to 7, characterized in that the unsafe environment includes at least one of an environment on steel frames, an environment where equipment for attaching a safety belt hook is provided, an environment where a hole exists, an environment where the height is equal to or greater than a threshold, an environment where a handrail is provided, and an environment where the floor is not fully tightened. (Supplementary Note 9) The work monitoring support device according to any one of Supplementary Notes 1 to 8, wherein the sensor includes a 360-degree camera attached to a helmet worn by the worker.(Supplementary Note 10) The work monitoring support device according to any one of Supplements 1 to 9, wherein the information output unit outputs the unsafety information to an alarm device carried by the worker. (Supplementary Note 11) The work monitoring support device according to any one of Supplements 1 to 9, wherein the information output unit outputs the unsafety information to at least one of a supervisor terminal carried by a supervisor monitoring the work site and a management device outside the work site. (Supplementary Note 12) The work monitoring support device according to Supplementary Note 11, wherein the unsafety information includes information indicating whether or not an unsafe event has occurred for each worker. (Supplementary Note 13) The work monitoring support device according to Supplementary Note 10 or 11, wherein the unsafety determination unit generates the unsafety information when it determines that an unsafe event has occurred, and the unsafety information includes information indicating an alarm. (Supplementary Note 14) The work monitoring support device according to Supplementary Note 13, characterized in that the unsafety determination unit generates information indicating the warning by inputting the image data including the event for which it has been determined that the unsafe event has occurred into a trained model generated by machine learning that generates sentences using the input image data. (Supplementary Note 15) The work monitoring support device according to Supplementary Note 13, characterized in that the information indicating the warning is audio data, and is generated by instructing a trained model generated by machine learning that generates audio using the input image data to generate audio data that will not startle the worker. (Supplementary Note 16) The work monitoring support device according to any one of Supplements 1 to 15, characterized in that when a release area is designated as an area at the work site that is to be excluded from the target of the first detection process, the work monitoring support device excludes the release area from the target of the first detection process. (Supplementary Note 17) The work monitoring support device according to any one of Supplements 1 to 16, characterized in that data including the determination result when it is determined that the unsafe event has occurred and the sensor information corresponding to the determination result is accumulated as performance data. (Supplementary Note 18) The work monitoring support device described in Supplementary Note 17 is characterized in that the unsafety judgment unit uses the performance data to generate presentation data for instructing the worker on improvements or evaluating the worker, and the information output unit outputs the presentation data.(Supplementary Note 19) The work monitoring support device according to Supplementary Note 17 or 18, characterized in that the unsafety determination unit counts the number of occurrences of the unsafe event for each worker using the performance data, and the information output unit outputs the counting result by the determination unit. (Supplementary Note 20) A work monitoring support system comprising: a sensor that detects a situation at a work site; and a work monitoring support device, wherein the work monitoring support device uses sensor information acquired by the sensor to perform a first detection process to detect whether an environment in which a worker working at the work site exists is an unsafe environment and a second detection process to detect whether the worker is in an unsafe state, and when the results of the first detection process and the second detection process indicate that the worker is in an unsafe environment and is in an unsafe state, determines that an unsafe event has occurred and generates unsafety information indicative of the determination result; and an information output unit that outputs the unsafety information. (Supplementary Note 21) A work monitoring support method in a work monitoring support device, comprising: a step of using sensor information acquired by a sensor that detects a situation at a work site to perform a first detection process to detect whether an environment in which a worker working at the work site is present is an unsafe environment and a second detection process to detect whether the worker is in an unsafe state, determining that an unsafe event has occurred when the results of the first detection process and the second detection process indicate that the worker is in an unsafe environment and is in an unsafe state, and generating unsafety information indicating the determination result; and a step of outputting the unsafety information.(Supplementary Note 22) A program causing a computer system to execute the following steps: using sensor information acquired by a sensor that detects the situation at a work site, to perform a first detection process to detect whether an environment in which a worker working at the work site is present is an unsafe environment, and a second detection process to detect whether the worker is in an unsafe state; if the results of the first detection process and the second detection process indicate that the worker is present in an unsafe environment and is in an unsafe state, to determine that an unsafe event has occurred and to generate unsafe information indicating the determination result; and outputting the unsafe information.

[0096] 1 Work monitoring support device, 2 Sensor, 3 Portable terminal, 4 Alarm device, 5 Supervisor terminal, 6 Management device, 7 Helmet, 8 Safety belt, 9 Hook, 11, 31 Transmitting / receiving unit, 12 Information storage unit, 13 First detection unit, 14 Second detection unit, 15 Determination unit, 16 Response management unit, 17 Unsafety determination unit, 21 Camera, 22 Sticker, 32 Display unit, 33 Audio output unit, 34 Vibration generation unit, 35 Reading unit, 131, 141 Model generation unit, 132, 142 Inference unit.

Claims

1. A work monitoring support device comprising: an unsafety determination unit that uses sensor information acquired by a sensor that detects the situation at a work site to perform a first detection process to detect whether the environment in which a worker working at the work site is located is an unsafe environment and a second detection process to detect whether the worker is in an unsafe state, and that determines that an unsafe event has occurred if the results of the first detection process and the second detection process indicate that the worker is in an unsafe environment and is in an unsafe state, and generates unsafety information indicative of the determination result; and an information output unit that outputs the unsafety information.

2. The work monitoring support device according to claim 1, wherein the unsafe state includes a state in which the worker is performing an unsafe behavior.

3. A work monitoring support device according to claim 2, characterized in that the unsafe behavior includes the worker not wearing a safety belt.

4. The work monitoring support device according to claim 3, characterized in that the sensors include a human presence sensor installed on the equipment to which the safety belt hook is attached or in the vicinity of the equipment, and a hook sensor provided on the hook to detect whether the hook is attached or not.

5. A work monitoring support device according to any one of claims 2 to 4, characterized in that the unsafe behavior includes the worker jumping from one work to another.

6. A work monitoring support device according to any one of claims 2 to 5, characterized in that the unsafe condition includes a state in which the worker is in poor physical condition.

7. The work monitoring support device according to claim 6, wherein the sensors include a vital sensor for acquiring vital data of the worker.

8. A work monitoring support device as described in any one of claims 1 to 7, characterized in that the unsafe environment includes at least one of an environment on a steel frame, an environment where equipment for attaching a safety belt hook is provided, an environment where a hole exists, an environment where the height is above a threshold value, an environment where a handrail is provided, and an environment where the floor is not fully tightened.

9. A work monitoring support device as described in any one of claims 1 to 8, wherein the sensor includes a 360-degree camera attached to a helmet worn by the worker.

10. A work monitoring support device as described in any one of claims 1 to 9, characterized in that the information output unit outputs the unsafety information to an alarm device carried by the worker.

11. A work monitoring support device as described in any one of claims 1 to 9, characterized in that the information output unit outputs the unsafety information to at least one of a supervisor terminal carried by a supervisor monitoring the work site and a management device outside the work site.

12. A work monitoring support device according to claim 11, characterized in that the unsafety information includes information indicating whether or not an unsafe event has occurred for each worker.

13. The work monitoring support device described in claim 10 or 11, characterized in that the unsafety judgment unit generates the unsafety information when it judges that an unsafe event has occurred, and the unsafety information includes information indicating an alarm.

14. The work monitoring support device described in claim 13, characterized in that the unsafety judgment unit generates information indicating the warning by inputting the image data including an event that has been judged to be the occurrence of the unsafe event into a trained model generated by machine learning that generates sentences using input image data.

15. The work monitoring support device described in claim 13, characterized in that the information indicating the alarm is audio data, and is generated by instructing a trained model generated by machine learning that generates audio using input image data to generate audio data that will not startle the worker.

16. A work monitoring support device as described in any one of claims 1 to 15, characterized in that when a release area is specified as an area to be excluded from the target of the first detection process at the work site, the release area is excluded from the target of the first detection process.

17. A work monitoring support device described in any one of claims 1 to 16, characterized in that data including the judgment result when it is judged that the unsafe event has occurred and the sensor information corresponding to the judgment result is accumulated as performance data.

18. The work monitoring support device described in claim 17, characterized in that the unsafety judgment unit uses the performance data to generate presentation data for instructing the worker on improvements or evaluating the worker, and the information output unit outputs the presentation data.

19. The work monitoring support device described in claim 17 or 18, characterized in that the unsafety judgment unit uses the performance data to count the number of occurrences of the unsafe events for each worker, and the information output unit outputs the counting results by the judgment unit.

20. A work monitoring support system comprising: a sensor that detects the situation at a work site; and a work monitoring support device, wherein the work monitoring support device uses sensor information acquired by the sensor to perform a first detection process that detects whether the environment in which a worker working at the work site is located is an unsafe environment and a second detection process that detects whether the worker is in an unsafe state, and an unsafety determination unit that determines that an unsafe event has occurred and generates unsafety information indicative of the determination result when the results of the first detection process and the second detection process indicate that the worker is in an unsafe environment and is in an unsafe state; and an information output unit that outputs the unsafety information.

21. A work monitoring support method in a work monitoring support device, comprising the steps of: using sensor information acquired by a sensor that detects the situation at a work site, performing a first detection process to detect whether the environment in which a worker working at the work site is located is an unsafe environment and a second detection process to detect whether the worker is in an unsafe state; and when the results of the first detection process and the second detection process indicate that the worker is in an unsafe environment and is in an unsafe state, determining that an unsafe event has occurred and generating unsafety information indicating the determination result; and outputting the unsafety information.

22. A program that causes a computer system to execute the following steps: using sensor information acquired by a sensor that detects the situation at a work site, to perform a first detection process to detect whether the environment in which a worker working at the work site is located is an unsafe environment, and a second detection process to detect whether the worker is in an unsafe state; if the results of the first detection process and the second detection process indicate that the worker is in an unsafe environment and is in an unsafe state, to determine that an unsafe event has occurred and to generate unsafe information indicating the determination result; and to output the unsafe information.