JP7164575B2 - Safety management system - Google Patents

Description

The present invention relates to safety management systems.

Communication equipment construction and equipment maintenance work are carried out at many construction sites every day. There are many construction sites with a small number of workers, and it is difficult to have safety confirmation personnel permanently stationed at all construction sites. For this reason, the safety checker patrols each construction site to perform safety management.

Non-Patent Document 1 discloses a safety management system for remotely managing the safety of workers.

“Worker Safety Management Service”, NTT Advanced Technology Corporation, Internet <URL: https://www.ntt-at.co.jp/product/datasol_024/>

When a safety checker patrols each construction site, there is a problem that it is difficult to provide real-time safety guidance that must be truly implemented.

The safety management system of Non-Patent Document 1 analyzes data collected from sensors attached to workers and manages the states of workers. The safety management system of Non-Patent Document 1 can detect in advance that individual workers are likely to be in a dangerous situation, but the safety manager confirms the safety of the entire construction site and cannot give guidance.

SUMMARY OF THE INVENTION The present invention has been made in view of the above, and an object of the present invention is to make it possible to provide accurate safety guidance in real time.

A safety management system according to one aspect of the present invention includes one or more imaging devices that are fixed to a construction site, captures the entire construction site and transmits the video, analyzes the video, and generates video that requires safety confirmation. a video analysis device that specifies and notifies; a confirmer terminal that displays the video and enables communication with a worker at the construction site whose video requires safety confirmation ; Equipped with a headset capable of communicating with a confirmer terminal, the video analysis device applies a learning model obtained by learning teacher data in which a tag representing the object is added to an area in the video where the object exists, and the construction site. is input, an object is detected from the image, and based on the positional relationship of the detected object, it is determined whether or not safety confirmation of the construction site is necessary, and the confirmer terminal holding a table that associates a photographing device with the headset, displaying the images from the plurality of construction sites, and the headset associated with the photographing device that photographed the images in response to an operation on the images; to start a call .

According to the present invention, accurate safety guidance can be implemented in real time.

FIG. 1 is a diagram showing an example of the configuration of the safety management system of this embodiment. FIG. 2 is a diagram showing an example of a display screen of a checker terminal. FIG. 3 is a diagram showing an example in which a video of a construction site is displayed on a confirmer terminal. FIG. 4 is a diagram showing an example of an arrangement of imaging devices. FIG. 5 is a diagram showing an example of the flow of processing of the safety management system.

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a diagram showing an example of the configuration of the safety management system of this embodiment. The safety management system shown in the figure includes a safety management server 1, a checker terminal 2, a photographing device 3 arranged at each construction site, and a headset 4 worn by a worker.

The safety management server 1 receives video of the construction site from the imaging device 3 at each construction site and analyzes the video in real time. The safety management server 1 transmits the video to the confirmer terminal 2 and notifies the result of the analysis. The analysis result includes information indicating whether or not the video requires safety confirmation by a safety confirmation person. The safety management server 1 determines whether or not the video of the construction site requires a safety check by a safety checker based on a predetermined notification determination logic. The safety confirmer confirms the video with the confirmer terminal 2, and if necessary, operates the confirmer terminal 2 to transmit to the headset 4 worn by the worker, and communicates with the worker in real time. The safety management server 1 will be described below.

The safety management server 1 shown in FIG. Each unit included in the safety management server 1 may be configured by a computer having an arithmetic processing unit, a storage device, etc., and the processing of each unit may be executed by a program. This program is stored in a storage device included in the safety management server 1, and can be recorded on a recording medium such as a magnetic disk, optical disk, or semiconductor memory, or can be provided through a network. The processing of each part of the safety management server 1 may be distributed to a plurality of computers, or the safety management server 1 may be realized by a virtual machine on the cloud.

The receiving unit 11 receives images of the construction site from the imaging device 3 at each construction site, transmits the received images to the analysis unit 12 , and stores them in the storage unit 15 . The receiving unit 11 grasps the correspondence relationship between the received image and the image capturing device 3 that captured the image, and associates the image with the name of the image capturing device 3 (or the identifier of the image capturing device 3).

The analysis unit 12 utilizes AI object detection technology to analyze images, and detects images that require safety confirmation from images of each construction site. For example, as a video that requires safety confirmation, the position of a worker can be detected to detect a video of a construction site where a worker is working at height, or the position of a triangular cone or construction vehicle can be detected. For example, it detects images of construction sites where the placement of triangular cones is inappropriate.

The notification unit 13 distributes the video of each construction site to the checker terminal 2 and notifies the checker terminal 2 of the video detected by the analysis unit 12 and requiring safety confirmation. The notification unit 13 distributes not only the image detected by the analysis unit 12 but also all received images to the checker terminal 2 . The notification unit 13 may distribute to the checker terminal 2 a video with a reduced resolution or a reduced frame rate. The notification unit 13 may deliver a high-resolution video upon request from the confirmer terminal 2 .

The learning unit 14 learns the tagged video stored in the storage unit 15 as teacher data to improve the detection accuracy of the analysis unit 12 .

The accumulation unit 15 accumulates images from the imaging device 3 at each construction site. The accumulated video is used for learning of the analysis unit 12 by the learning unit 14 .

The checker terminal 2 receives the video of the construction site from the safety management server 1 and displays it. The confirmer terminal 2 informs the safety confirmer when receiving the notification of the video requiring safety confirmation. The confirmer terminal 2 may indicate that the notification has been received by sound, or may indicate that the notification has been received on the screen of the confirmer terminal 2 .

FIG. 2 shows an example of the display screen 100 of the checker terminal 2. As shown in FIG. On the display screen 100 shown in the figure, images 110 of a plurality of construction sites are displayed side by side. Above the image 110, the name 120 of the photographing device 3 that shot the image 110 is displayed. A video 110 requiring safety confirmation is highlighted with a frame 130 or the like. Note that the display screen 100 in FIG. 2 is an example, and is not limited to this.

When the image 110 is selected from the list of images 110 on the display screen 100 of FIG. 2, the selected image is enlarged and displayed on the display screen 200 of the checker terminal 2 as shown in FIG. When enlarging and displaying a certain image, the checker terminal 2 may request the safety management server 1 for a high-resolution image. The safety confirmer confirms the points to be instructed on safety while watching the enlarged image.

When the safety inspector wishes to communicate with the workers at the construction site in the video, the safety inspector selects the telephone icon 210 within the display screen 200 . As a result, the headset 4 worn by the worker at the video construction site can be called. The photographing device 3 and the headset 4 are paired, and if the photographing device 3 that has photographed the video of the construction site is known, the headset 4 paired with the photographing device 3 can be known. For example, the safety management server 1 or the confirmation person terminal 2 holds a table that associates the imaging device 3 and the headset 4 (which may be the contact information of the operator), and when the telephone icon 210 is selected, the displayed The photographing device 3 that has photographed the video is specified, the contact information of the worker is acquired, and the worker is called.

The headset 4 automatically responds (off-hooks) when the worker does not respond within a predetermined number of calls (for example, within 3 calls) when a call is received from the confirmer terminal 2 to confirm safety. A call may be made between the person and the worker. Even if the operator is working and cannot take his hands off the machine, the headset 4 automatically answers the call, so that the safety confirmation person and the operator can talk to each other.

The photographing device 3 is a camera that photographs a place where a worker works at a construction site. The photographing device 3 has a function of photographing an image of the construction site and a function of transmitting the photographed image to the safety management server 1 . The photographing device 3 may have a function of displaying the photographed image so that the photographing range can be known. Since it may not be possible to secure a power source at a construction site, it is preferable that the photographing device 3 incorporates a battery. Moreover, since the work may be carried out in rain, wind, or snow, the photographing device 3 should preferably have a body that can withstand rain, wind, and snow.

As shown in FIG. 4, the imaging device 3 is arranged above the triangular cone 51 . In the example of FIG. 4, an adapter 52 for a triangular cone is attached to the upper part of the triangular cone 51, and a support base 53 of the photographing device 3 is placed on the adapter 52. In the example of FIG.

An adapter 52 for attaching a reflector or a warning light to the triangular cone 51 can be used. If a triangular cone 51 with a cut top is used, the adapter 52 is not required.

The support base 53 is preferably one to which the photographing device 3 can be attached and whose elevation angle of the photographing device 3 can be freely changed. The horizontal direction can be adjusted when the support base 53 is placed on the triangular cone 51 . If the lower surface of the support base 53 is provided with a projection so that the projection can be inserted into the adapter 52, the photographing device 3 can be placed on the triangular cone 51 more stably.

Since the triangular cone 51 is always brought to the construction site, the photographing device 3 is attached to a support base 53 that can be attached to the triangular cone 51 so that the photographing device 3 can be arranged on the triangular cone 51. Also, a tripod or the like for arranging the photographing device 3 is not required, and the photographing device 3 can be easily arranged at the construction site.

Next, the flow of processing of the safety management system of this embodiment will be described with reference to the flowchart of FIG.

First, when the worker arrives at the construction site, the worker takes out the triangular cone and cone bar from the work vehicle, arranges the triangular cone and cone bar on the construction site, and installs a security fence. Then, the operator attaches the photographing device 3 to the support base 53 and places it on the triangular cone 51 . The triangular cone 51 on which the photographing device 3 is arranged may be the triangular cone 51 used for the security fence, or may be the triangular cone 51 arranged separately from the security fence. The operator turns on the power of the imaging device 3 and confirms the imaging range of the imaging device 3 . The imaging range may be, for example, the entire place where the worker works. When the arrangement of the photographing device 3 is finished, the photographing device 3 starts photographing the construction site and transmitting the photographed image.

In step S<b>11 , the safety management server 1 receives the video captured by the imaging device 3 .

In step S12, the safety management server 1 analyzes the received video. When receiving images from a plurality of photographing devices 3, the safety management server 1 analyzes each image.

In step S13, the safety management server 1 determines whether or not there is a construction site requiring safety confirmation from the video analysis result.

If there is no construction site that requires safety confirmation, the safety management server 1 waits to receive the next video.

If there is a construction site that requires safety confirmation, in step S14, the safety management server 1 notifies the confirmer terminal 2 that a construction site that requires safety confirmation has been detected.

The safety confirmer confirms the notified video of the construction site. If the safety confirmer thinks that it is better to give safety guidance to the worker, in step S15, the worker's headset 4 is called from the confirmer's terminal 2 and talks with the worker.

Through the above processing, the video of the construction site that requires safety confirmation is narrowed down by the safety management server 1, and the safety checker can check the video and talk with the worker at the construction site of the video. A person can provide accurate safety guidance remotely in real time.

Next, learning by the learning unit 14 will be described.

In order to improve the detection accuracy of the analysis unit 12, the learning unit 14 learns the video stored in the storage unit 15 with a tag attached thereto as teacher data. For example, a region in which an object such as a worker, a utility pole, a triangular cone, a bucket, or a ladder exists in an image is given a tag representing the object to generate teacher data, and the learning unit 14 learns the teacher data. We aim to improve the accuracy of detecting an object from within an image. The learning unit 14 constructs notification determination logic for detecting a construction site for high-place work or a construction site where a jump-in accident is likely to occur, according to the positional relationship of objects detected from the video. For example, the learning unit 14 determines from the positional relationship between the utility pole and the worker whether or not the video is a construction site for high-place work, and from the positional relationship between the triangular cone and the bucket, the learning unit 14 determines whether the video is a construction site where a jump-in accident is likely to occur. Build a notification judgment logic that judges whether or not there is. A jump-in accident is an accident in which a high truck or the like contacts a bucket. If there is a bucket protruding from the security fence, there is a possibility of a diving accident.

As described above, the safety management system of the present embodiment includes one or more imaging devices 3 that capture images of a construction site and transmit images, and analyzes the images to identify and notify images that require safety confirmation. and a confirmation person terminal 2 that displays images and enables communication with workers at the construction site whose images require safety confirmation. This makes it possible to identify the video of a construction site that really needs confirmation from among many videos and communicate with the workers at that construction site, so even a limited number of people can work on a large number of construction sites. , it is possible to provide accurate safety guidance in real time.

In this embodiment, since the photographing device 3 is attached to the support base 53 and arranged on the triangular cone 51, it is not necessary to bring a tripod or the like for installing the photographing device 3 to the construction site.

In this embodiment, when a call is received from the checker terminal 2, if the worker does not respond within a predetermined call, the headset 4 automatically responds, so that the worker is at work and his hands are busy. Even if the operator cannot leave the operator, the safety confirmer can communicate with the operator.

DESCRIPTION OF SYMBOLS 1... Safety management server 11... Reception part 12... Analysis part 13... Notification part 14... Learning part 15... Accumulation part 2... Confirmer terminal 3... Imaging device 4... Headset

Claims (3)

one or more imaging devices fixed to the construction site to capture the entire construction site and transmit images;
a video analysis device that analyzes the video and identifies and notifies the video that requires safety confirmation;
a confirmer terminal that displays the video and enables communication with the worker at the construction site of the video that requires safety confirmation ;
A headset worn by the worker and capable of talking with the confirmer terminal,
The video analysis device detects an object from the video by inputting the video of the construction site to a learning model that learns teacher data that has a tag indicating the object in the region where the object exists in the video. , Based on the positional relationship of the detected objects, determine whether or not safety confirmation of the construction site is necessary ,
The confirmation person terminal holds a table that associates the photographing device and the headset, displays the images from the plurality of construction sites, and the photographing device that photographed the images in response to the operation on the images. Initiate a call with said associated headset
Safety management system. The safety management system according to claim 1, wherein the photographing device is attached to a support that can be attached to a triangular cone and arranged on the triangular cone. 3. The safety management system according to claim 1, wherein the headset automatically responds when the worker does not respond within a predetermined call when receiving a call from the confirmer terminal.

Images