KR20230030265A - a safety management method that provides a dynamic user interface using artificial intelligence - Google Patents

Abstract

The present invention relates to a safety managing method for providing a dynamic user interface using artificial intelligence comprising: a step of scanning safety target items according to each facility and structure using energy in target equipment; a step of providing a management setting interface to select one or more management items desired to be managed among the safety target items; a step of performing first monitoring depending on the number of management items and types when setting the management items by the management setting interface; a step of determining whether monitoring is necessary by the artificial intelligence except for the management items among management target items while performing the first monitoring, and selecting the same as an additional monitoring item; and a step of providing a monitoring interface screen depending on equipment use information for the set management item and the selected additional monitoring item. The present invention calculates an importance value of the additional monitoring item; determines a size of a display area on the monitoring interface screen for the equipment use information of the additional monitoring item depending on the importance value; and displays object information and environmental information corresponding to equipment use of the additional monitoring item on the monitoring interface screen when the importance value of the additional monitoring item is greater than or equal to a threshold value.

Description

A safety management method that provides a dynamic user interface using artificial intelligence}

The present invention relates to a safety management method for providing a dynamic user interface, and more particularly, to a method for providing a dynamic user interface using artificial intelligence.

Recently, manufacturing facilities have become large and highly functional, and various facilities such as air conditioning, sanitation, power, crime prevention, and disaster prevention are complexly combined. A system for more conveniently monitoring and controlling equipment according to these various facilities is being built. Recently, a safety management solution for each facility is being built together.

A safety management solution is basically a holistic system that provides real-time control that minimizes emergency situations by monitoring the usage of each facility, analyzing equipment conditions for quick and effective strategic decision-making.

However, due to the increase in the number of equipment to be managed in recently enlarged equipment, it is virtually difficult to monitor and control the use of various equipment by non-specialized personnel. In addition, it is difficult for a company providing a safety management solution to provide a unified management interface because the type, installation, and operation type of equipment are different for each equipment.

Based on the above discussion, a safety management method that provides a dynamic user interface using artificial intelligence will be provided below.

The technical problems to be achieved in the present invention are not limited to the above technical problems, and other technical problems not mentioned will be clearly understood by those skilled in the art from the description below.

In the safety management method for providing a dynamic user interface using artificial intelligence, which is an aspect of the present invention for solving the above problems, scanning safety target items according to each facility and structure using energy in target equipment ; providing a management setting interface for selecting one or more management items desired to be managed among the safety target items; If the management items are set by the management setting interface, performing a first monitoring according to the number and type of the management items; selecting, as additional monitoring items, those determined to require monitoring by artificial intelligence, excluding the management items, among the management target items while performing the first monitoring; And providing a monitoring interface screen according to the equipment usage status information for the set management item and the selected additional monitoring item, wherein the importance value of the additional monitoring item is calculated, and the equipment of the additional monitoring item is calculated according to the importance value. Determines the size of the display area in the monitoring interface screen for the usage status information, and if the importance value of the additional monitoring item is greater than or equal to a threshold value, object information and environment information corresponding to the equipment usage status of the additional monitoring item are displayed in the monitoring Characterized in that it is displayed on the interface screen.

Furthermore, by analyzing target equipment information including the type of use, structure, and size of the target equipment, a dynamic user using artificial intelligence selects an item that can cause a safety accident among the management target items and provides it as the management setting interface. It is characterized by providing an interface.

According to an embodiment of the present invention, it is possible to efficiently perform safety management providing a dynamic user interface using artificial intelligence.

The effects that can be obtained in the present invention are not limited to the effects mentioned above, and other effects not mentioned can be clearly understood by those skilled in the art from the description below. will be.

The accompanying drawings, which are included as part of the detailed description to aid understanding of the present invention, provide examples of the present invention and explain the technical idea of the present invention together with the detailed description.
1 is a configuration diagram of a facility safety management system according to an embodiment of the present invention.
2 is a reference diagram for explaining a signal processing aspect of a facility safety management system according to an embodiment of the present invention.
3 is a reference diagram for explaining the operation of a facility safety management system using a wearable device according to an embodiment of the present invention.
4 is a reference diagram for explaining an operation of an image sensing module according to an embodiment of the present invention.
5 is a reference diagram for explaining a dynamic interface according to an embodiment of the present invention.

The present invention is not limited to the description of the embodiments described below, and it is obvious that various modifications may be applied within a range not departing from the technical gist of the present invention. And, in describing the embodiments, descriptions of technical contents that are widely known in the technical field to which the present invention pertains and are not directly related to the technical subject matter of the present invention will be omitted.

Meanwhile, in the accompanying drawings, like components are represented by like reference numerals.

And in the accompanying drawings, some components may be exaggerated, omitted, or schematically illustrated. This is to clearly explain the gist of the present invention by omitting unnecessary descriptions unrelated to the gist of the present invention.

1 is a configuration diagram of a facility safety management system according to an embodiment of the present invention.

The facility safety management system according to the present invention may be composed of the operator confirmation control device 110 to which the present invention is applied and the work target facility 130 into which the operator confirmation control device is combined/inserted.

According to the present invention, the operator confirmation control device 110 is configured to include an operator confirmation module 111 and a lock control signal output module 113.

The worker identification module 111 recognizes whether the worker is wearing/unwearing or displays the work contents to the worker, thereby providing a wearable device capable of receiving confirmation of work contents and a module capable of recognizing the wearable device, a camera, and a CCTV Workers present in the working radius of the facility to be worked from various modules, such as an image capturing module through a lamp, an image-based detection module that can detect the presence of a worker by receiving images from the image capturing module, and a detection sensor module using infrared light It is possible to obtain detection information about. That is, when the detection information is obtained from at least one detection module, the operator identification module may determine whether or not a worker is present in the facility to be worked on or determine the existence thereof.

Furthermore, the detection module connected to the operator identification module may be activated with priority. For example, only when the wearable device is activated, the wearable device recognition module is activated, or when the image-based detection module is activated, the wearable Activation of the device may be interrupted. Conversely, in a situation in which an error occurs with respect to operator recognition in an image-based sensing module, etc., the wearable device or human body sensor may be activated. That is, even if a plurality of sensing modules are used, the operator confirmation module may instruct activation/deactivation of some sensing modules in consideration of sensor status, operator status, battery consumption, and the like.

The lock control signal output module 113, based on the control signal input from the existing control module 131 that controls the work target facility 130 or the worker detection signal input from the worker confirmation module 111, the work target A lock/unlock signal may be output to the lock target facility 133 among facilities.

In particular, the lock control signal output module in the present invention is included in the work target facility from the control module 131 included in the work target facility or existing outside the work target facility (eg, central control server, etc.) It is characterized in that the control signal transmitted to the device to be locked 133 is received in the middle, and only the lock signal among the received control signals is controlled. During the maintenance of work, a worker may transmit a test signal to the work target equipment in order to check whether or not the work target equipment is being properly maintained. Therefore, blocking all signals transmitted from the control module in the middle may reduce the efficiency of the worker's inspection of the equipment to be worked on, and may cause a delay in the work.

Therefore, the lock control signal output module 113 in the present invention checks whether the control signal received from the control module is a lock/unlock signal of the equipment to be worked on, and then selectively transmits a control signal. It can be determined whether to transmit the signal to the equipment to be operated or to additionally output a lock control signal instructing the equipment to be locked/unlocked.

In addition, the lock control signal output module 113 in the present invention checks whether the operator has been identified by determining not only the control signal input from the control module for the work target facility, but also the operator detection signal of the operator confirmation module 111. and output a lock control signal. Accordingly, the lock control signal output module 113 may individually determine whether to lock and unlock the work target equipment.

The lock control signal output module 113 can stop the operation of the work target equipment even if only one of the cases where the lock signal of the work target equipment is input or the worker is confirmed by the worker confirmation module among the control signals can This is to prevent an unexpected accident by a worker by preventing a malfunction of a work target facility.

However, in the case of unlocking the work target equipment, it is preferable to start the operation of the work target equipment after checking all detection signals and control signals. This is to prevent an accident that may occur by inputting a lock release signal in a situation where it is not confirmed whether a worker exists in the work target equipment from the outside.

Therefore, the lock control signal output module receives the unlock signal of the work target equipment among the control signals received from the control module, and through the worker confirmation module, it is determined that the worker does not exist in the work target equipment (ie, the worker's absent), it is possible to output a lock control signal that starts the operation of the work target equipment.

Furthermore, the lock control signal output module 113 according to the present invention, in addition to the operation of outputting the lock control signal, when the work target facility is in an abnormal state (for example, the work target facility is open even though the door is normally closed) state), it is possible to perform an operation to emergency stop the equipment, sound a warning light installed outside the equipment using a notification module, or share an emergency between workers through a wearable device. That is, the notification module may display a notification composed of tactile, auditory, and visual effects by which workers having a specific code can recognize the locked state of the manufacturing apparatus to workers. For example, the notification module may notify that a worker exists in the facility to be worked or that the facility to be worked is in an abnormal state through at least one of a light emitting body, a lamp, a display unit, a speaker, and a wearable device associated with manufacturing equipment.

In the present invention, the sensing module that is connected to the operator confirmation module and can transmit the operator detection signal may have various forms. Hereinafter, for convenience of explanation, wearable devices, wearable device surveying modules, image-based detection modules, and human body detection sensors are mainly described, but IoT sensors capable of transmitting detection signals indicating the presence of workers with respect to the worker confirmation module The present invention can also be applied to additional devices, such as the like. In the present invention, the wearable device 121 is connected to the operator confirmation control device 110 and can transmit a worker detection signal to the operator confirmation module 111 .

For example, when a worker is recognized through a wearable device having a unique identifier (eg, smart band, work helmet, wireless earphone, etc.), whether the worker works on the equipment to be locked using the wearable device Recognizable.

At this time, the wearable device 121 may be set to have each unique identifier, and each unique identifier may be set to match a specific operator. Accordingly, information such as the contents of work scheduled for a specific worker and work target facilities can be displayed, or information such as the specific worker's work activity, work hours, work target facilities, etc. can be grasped by the manager, and work target facilities It is possible to identify an abnormal situation in which an unrelated worker approaches the equipment to be worked on.

The wearable device 121 in the present invention can be activated by recognizing what a specific worker is wearing, and when the worker is recognized through the activated wearable device, the unique identifier of the activated wearable device must be checked to lock the work target facility can be configured to release

This is because, when a target facility is locked by a specific worker, the worker confirmation module can reliably determine that the target facility is working. For example, when a wearable device receives a lock code related to a facility to be worked on from a worker, the worker confirmation module confirms the unique identifier of the wearable device and the unlock code corresponding to the lock code. , it can be determined that no worker exists in the equipment to be worked on. Here, the locking code may simply be a code for locking/unlocking, but it may be a code related to a specific work target facility, a code based on a specific maintenance schedule, or a code related to a worker group to which a specific worker belongs. , and the lock/unlock code may be shared by multiple workers.

For example, if a worker wears the wearable device 121 and the wearable device is activated, and the start of the maintenance work of the facility to be worked is notified using this, the worker confirmation device is a unique identifier of the wearable device notifying the start of the maintenance work , and operate to unlock the work target equipment only when the end of the maintenance work of the work target equipment is instructed through the wearable device matching the corresponding unique identifier. By using a wearable device, the operator can easily stop the equipment right before starting work, and only the person who stopped the drive is given the right to restart, preventing safety accidents caused by unexpected driving by external workers. .

In addition, since the locking/unlocking of the lock target facility 133 in the present invention is possible only through a specific wearable device, the operation of the lock target facility is prevented from being unnecessarily stopped because the worker who started the work does not unlock the lock. To this end, a notification may be periodically sent to a worker wearing the wearable device. For example, by periodically notifying workers that the facility to be locked is locked through vibration or notification through a wearable device, the psychological stability of the worker is promoted, and even if the work on the target facility is terminated due to the worker's carelessness It is possible to prevent the equipment to be operated from being stopped. Furthermore, when a worker notifies the start of work on a target facility through a wearable device, such as an operator confirmation control device, the start of the work can be communicated to fellow workers involved in maintenance of the target facility. If the human body detection sensor/image capturing module (camera) related to the equipment to be locked is activated, the wearable device may not be activated. This is because, as described above, when one sensing module normally recognizes the operator, there is no need for the other sensing module to recognize the operator redundantly, and this is to reduce the prevention of battery consumption of the wearable device. However, even in this case, when the worker identification module recognizes the start of work by the wearable device, it is preferable to unlock the lock target equipment only when the wearable device is activated and instructs the end of the work.

Furthermore, in the present invention, when the wearable device 121 is set to connect with the operator confirmation control device 110, the wearable device measurement module 123 may be additionally activated to determine the exact location of the operator.

The wearable device measurement module 123 in the present invention may be configured to measure a position of a worker through at least one antenna. That is, by measuring the distance between the antenna and the wearable device from the individual antenna constituting the wearable device measurement module, and using a plurality of antennas to check the location of the wearable device through triangulation, the operator exists inside the facility to be locked You can check if it does or not. A detailed description of this will be described in more detail in the description related to FIG. 3 below.

In the present invention, as one of the sensing modules connected to the operator identification module, the image-based sensing module 125 may be used. The image-based detection module 125 analyzes the extracted image within a specific image area and determines that an operator exists when the image is similar to a predetermined image, and when a different image other than the predetermined image is detected within the specific image area. In this case, it may be determined that an abnormal situation has occurred and the operation of the equipment to be locked may be stopped. A detailed description of this will be described in more detail in the description related to FIG. 4 below.

In addition, in the present invention, the entry of a worker may be confirmed through a human body detection sensor module installed inside/outside of the equipment to be locked. Various sensors such as an infrared sensor, an ultrasonic sensor, an ultraviolet sensor, and a motion recognition sensor may be used as the human body detection sensor, and a worker detection signal may be transmitted to the worker identification module by configuring an IoT-based network. Furthermore, since the present invention recognizes a worker using a plurality of detection modules, priorities can be set for the efficiency of the work sensor, for example, the human body detection sensor deactivates the wearable device and image processing of the image detection module It can also be set to be activated only when is difficult. Therefore, when a specific detection module becomes abnormal, it tries to activate it by talking to other detection modules connected to the operator verification module, and when all detection modules connected to the operator verification module are deactivated, the control module of the facility to be operated is unlocked. A lock control signal for stopping the operation of the work target equipment may be output to the work target equipment even when the signal is transmitted and the worker confirmation module indicates that the worker is unidentified. This is to prevent the operator verification module from making an inaccurate determination due to an error in the detection module.

Therefore, according to the present invention, while minimizing the downtime (i.e., downtime) of the equipment to be worked on, workers can be guaranteed a safe environment in which the equipment is not operated until the necessary work is completed, and the manager or fellow workers can The operation status and work history of the system are shared in real time, and the effect of enabling easy understanding of the site can be promoted.

2 is a reference diagram for explaining a signal processing aspect of signal processing of a wearable device and a facility safety management system according to an embodiment of the present invention. In the description related to FIG. 2, the description will be made using the name of a transceiver for convenience of description, focusing on signal processing.

When the operator's transceiver 220 (for example, a wearable device) is activated through an operation such as being worn by the operator, the operator's transceiver can transmit a signal related to equipment entry/exit to a control module, operator confirmation control device, etc. and can receive event information related to equipment to be locked (eg, start/end of work by other workers, status of equipment to be locked). However, for example, in the case of a wearable device, when entering the inside of a facility to be worked on, the signal may be weakened or the strength of the signal may be relatively weak due to a problem with the battery, etc. Alternatively, a problem in which it is difficult to receive event-type information may occur.

Therefore, the facility safety management system according to the present invention may additionally include a repeater 220 . The repeater 220 may amplify the signal received from the operator transceiver and transmit it to the equipment transceiver, or transfer the amplified signal from the equipment transceiver to the operator transceiver. That is, each of the repeaters is a device for a wearable device, a sensor constituting an IoT network, and a signal transmitter/receiver for equipment for work target facilities, using a protocol set in advance for each individual repeater, and amplifying the signal transmitted to the repeater to improve reception sensitivity. And by increasing the distance, it is possible to send and receive signals to the operator's transceiver, work target equipment, or other repeater.

The equipment transceiver 230 may convert the signal received from the repeater 220 and transmit it to the lock control signal output module, or convert the signal received from the lock control signal output module and transmit the signal to the repeater/operator transceiver. The transceiver for the equipment may be included in the operator confirmation module, and in the case of a transceiver worn by the operator (eg, a wearable device), various wireless communication protocols may be used. For these wireless communication protocols, a lock control signal output module This needs to be converted into a recognizable signal. Therefore, in the present invention, as the transceiver for equipment is used, the wearable device can use various protocols such as LTE, WIFI, BLE, and LORA, and the communication method is not limited. Furthermore, signal transmission from equipment using various communication protocols to the lock control signal output module 240 can be stably performed only by performing a hardware/software update of the operator transceiver.

The locking control signal output module 240 may transmit a drive/stop command to the work target facility based on at least one of the signals transmitted through the control module and the operator transceiver, and when an event occurs from the work target device, the operator can be transmitted to the receiver. Furthermore, based on the signal recognized or transmitted through the equipment transceiver 230, the control module for the worker's entry/exit of the work target facility and the state of the work target facility (operating state, door open, failure, etc.) By sending it to , it can help the manager to understand information related to the target equipment.

At this time, the lock control signal output module 240 in the present invention transmits the signal to the equipment to be operated using power line communication (PLC), and in the case of manufacturing equipment, there are cases in which the size is larger than the signal radius of wireless communication, , Since the signal strength may be weakened or interference may occur depending on the mechanical structure of the manufacturing equipment, control commands can be stably transmitted to the work target equipment by configuring to control the work target equipment using power line communication.

The control module 260 is an operator confirmation control device including a lock control signal output module, and can receive information such as operation/stop status and operation history of the facility to be worked on, store it in a database, etc., and display it to the manager. It can also provide manager functions such as forced restart of target equipment or data initialization.

3 is a reference diagram for explaining the operation of a facility safety management system using a wearable device and a wearable device measurement module according to an embodiment of the present invention.

According to one embodiment of the present invention, a worker may input a lock code through an activated wearable device (S310). If a worker displays his/her own work start status through a wearable device upon entering a facility to be worked on, the work can be displayed on the wearable devices of other workers sharing the same lock code, and if the facility to be worked is in operation, the work can be stopped. or it can be changed to a locked state that cannot be restarted. Furthermore, workers who share a lock code associated with a work target facility in a locked state may periodically receive a signal related to the state of the work target facility.

Accordingly, when the wearable device is activated, the facility safety management system determines whether the wearable device measurement module is additionally activated (S320), and when the wearable device measurement module is activated, the operator's position can be measured using this. . The wearable device measurement module may include at least one antenna, and in the wearable device measurement module, the type of information to be measured may vary according to the number of antennas recognizing the wearable device. For example, in the wearable device measurement module, when one antenna is used, only the wearable device can be recognized, and when two antennas are used, the wearable device is measured in consideration of the recognized distance of the wearable device from each antenna. A location candidate of the device can be calculated. Furthermore, in the case of using three or more antennas, the position of a worker may be measured through triangulation based on distances from individual antennas. In particular, in a method of recognizing a location by triangulating a distance of a wearable device through individual antennas, communication protocols such as radio frequency, Bluetooth, and WIFI may be utilized.

Hereinafter, a case in which a position of a worker is measured using three or more antennas will be mainly described. When the wearable device measurement module of the facility safety management system is activated, the position of the operator is measured using three or more antennas (S325).

For example, according to triangulation based on distances from each of three or more antennas to the wearable device, the operator's location recognized from the wearable device may be configured to compare the area of the work target facility. Here, the individual antenna may be installed at a location determined in advance by a manager, but may be configured to include a GPS sensor so that its location can be automatically identified when coupled to a facility to be worked.

The operator confirmation module of the safety management system may determine whether or not the operator is present in the work target facility (S330). Even if the worker does not display his or her work start status through the wearable device upon entry, the image-based detection module or the human body detection module coupled to the equipment may check whether the worker has entered. In this case, if all or part of the human body is detected in the facility to be operated, or if normal worker detection is impossible through a worker detection module (eg, camera, CCTV, human body detection sensor, etc.) If it is covered), it is determined that a worker has entered and then an operation to stop the equipment to be worked may be performed.

In addition, when the location of the wearable device is measured, the control server may map the location of the wearable device onto the layout of the work target facility to determine whether or not the user has entered the locking device and the current location of the operator.

At this time, the operator confirmation control device receives signals from the control module associated with the facility to be worked in addition to the operator detection module (S340).

Accordingly, a locking control signal instructing to start or stop the operation of the facility to be locked may be output based on the locking signal among the control signals received from the control module or the worker detection signal received from the worker checking module (S350).

When the locking control signal instructs locking, the facility safety management system stops the locking target facility among the work target facilities (S360).

When the operator finishes the work of the target equipment (S370), the driving of the target equipment is possible. For example, when there is an external drive attempt, there is no work notification through the wearable device, and the equipment can be operated in a situation where human body detection within the equipment determines that it is normally safe. In addition, since the equipment in the locked state cannot be restarted unless the worker releases the work through the wearable device, if the unlock code is not confirmed through the wearable device, the workers are periodically informed of the locked state. inform If the operator does not indicate the completion of the work through the wearable device when leaving by mistake, the equipment will not change the state in the locked state, making it impossible to restart. It is desirable to inform

Accordingly, the facility safety management system determines the unique identifier of the activated wearable device and the unlocking code corresponding to the locking code (S380). In other words, if a worker indicates the completion of his or her work through a wearable device when leaving, the work is displayed on the wearable devices of other workers sharing the same code, and the absence of the worker is confirmed in the worker confirmation module of the facility to be worked on. do.

If the operator confirmation module cannot normally perform the detection, it is assumed that the operator exists, the state of the facility to be locked is determined to be in an abnormal state, and the locked state is maintained.

When it is determined that there is no worker and an unlock signal is input from the control module, the lock control signal output module outputs a lock control signal instructing unlock to unlock the equipment to be worked on (S390). That is, when a worker tries to drive from the outside after exiting, if work completion is confirmed through a wearable device, the fact of work completion is displayed to other workers, and after the detection module in the equipment confirms the normal situation, the facility to be locked will be restarted

4 is a reference diagram for explaining an operation of an image sensing module according to an embodiment of the present invention.

The image-based detection module according to the present invention stops the operation of the equipment to be worked upon when the presence of an object image stored in advance on successive frames within the image area is detected or an image detecting a change in a predetermined image is detected. can do. Here, the pre-stored object image may be learned by applying artificial intelligence, and for example, a detection engine for an image-based detection module may be learned using a convolutional neural network (CNN), which is a deep learning algorithm.

For example, a dataset serving as learning data for body recognition in the present invention includes not only a person's whole body image, but also includes data of a body part that a worker mainly uses, so that the worker stretches out his hand to work. It can be learned so that it can be detected even if it enters the device or only part of it. In particular, the image classification method applied in the present invention, when the driving unit (motor, robot, etc.) of the equipment to be worked is recognized in the image, does not track the object having dynamic movement, but the extracted image is pre-stored object image ( eg, human), it is possible to show a high recognition rate for a worker, and the recognition rate can be steadily increased by learning false positive/no positive situation data using CNN. Accordingly, the detection engine learned through deep learning can read the image transmitted through the camera in real time to determine whether or not a person exists inside the facility to be worked on, and can determine whether or not a person exists inside the facility to be operated. Even if it is only a part of the human body, it can be recognized by the image detection module. That is, the detection engine according to the present invention can classify individual body parts and improve the accuracy of determining that a person has entered by performing learning on each part of the human body, such as hands, feet, and head.

Therefore, in the present invention, a real-time image is determined through a detection engine learned through artificial intelligence for a captured image, and when a body is recognized over a predetermined number or more consecutive frames, it can be determined that there is a person entering the interior. can Here, the number of frames that are the basis for determining the operator and the level of similarity for determining the extracted image may be set in advance by the administrator, but may be gradually improved according to the learning data of the detection engine. .

That is, if the operator's action (entry, exit) is determined through the camera, there is a problem in that the possibility of false positives (classifying entry as non-entry) and false positives (classifying non-entry as entry) may increase structurally. The image-based detection module according to the present invention may perform an emergency stop determination as soon as the body is observed, and a specific area of an image captured using a camera or CCTV may be set as an area where a worker should not be present.

4(a) is a reference diagram for explaining a method for determining whether a worker exists in an image area set to determine similarity through an extracted image.

In FIG. 4 (a), extraction of a similarity with an image database learned based on a data set consisting of predetermined object images within a specific image area on consecutive frames of a predetermined number or more is a predetermined level or higher. When the image exists, it may be determined that the operator exists.

The image capture module (eg, camera, CCTV, etc...) continuously captures an area where a worker can be located within the work target facility in real time and transmits it to the image-based detection module. Accordingly, the image-based sensing module may cluster data for image classification by extracting a single frame from an image transmitted from the image capturing module. For example, a single frame may be analyzed in an image-based sensing module, and the analyzed result values may be output as images and data, respectively. However, since the frame related to one image corresponds to less than 0.1 second in time unit, it is not desirable to stop the equipment by looking at the result of a single frame, and it is not desirable to convert it into accumulated data to determine whether a person actually entered the equipment. it is desirable to

Therefore, for example in FIG. 4(a), when analyzing an image of 15 fps (Frame Per Second), the result (image, data) is “Person 1 92%, Person 2 72%, table 100%” , and since “Person 1 and Person 2” are two entities classified as a body on a single frame, the image-based detection module can determine that a person has been detected if it is greater than the threshold set by the administrator. . For example, if the threshold is set to 75%, the “Person 1” object can be judged as a body on a single frame, and “Person 2”, the other object, does not show a similarity above the threshold, so it can be analyzed using an additional frame. In this case, by applying a method of determining the similarity of an object by applying a high threshold on a single frame and determining the similarity by applying a relatively low threshold on successive frames, the presence or absence is determined more efficiently. You may.

That is, if the facility safety management system lacks judgment data or determines that it is not a human body, it can determine whether the operation target facility can be operated by determining an additional frame. If a person is detected, an emergency stop command can be sent to the facility to be locked. In addition, images judged to be entered by a person may be stored separately for still recording of work target equipment and performance improvement of an image-based sensing module.

4(b) is a reference diagram for explaining a case in which a specific area of an image is set as an area where a worker is not supposed to be present. When an image different from a predetermined image is detected within a specific image, it may be determined that an operator exists. For example, even though only equipment should exist in the image area, when an image different from the image of the equipment is detected, an accident due to an unexpected entry of a worker may be prevented by temporarily stopping the operation. In other words, performing worker detection using artificial intelligence for all areas on the image may cause a load on the operation of the control server. can make it

In addition, the image-based detection module determines the similarity of the extracted image for a specific area on one image through a detection engine, and at the same time sets it to detect a different image for the remaining area, thereby efficiently preventing accidents of workers. can

5 is a reference diagram for explaining a safety management method for providing a dynamic user interface using artificial intelligence.

Scanning safety target items according to each facility and structure using energy from the target equipment; providing a management setting interface for selecting one or more management items desired to be managed among the safety target items; If the management items are set by the management setting interface, performing a first monitoring according to the number and type of the management items; selecting, as additional monitoring items, those determined to require monitoring by artificial intelligence, excluding the management items, among the management target items while performing the first monitoring; And providing a monitoring interface screen according to the equipment usage status information for the set management item and the selected additional monitoring item, wherein the importance value of the additional monitoring item is calculated, and the equipment of the additional monitoring item is calculated according to the importance value. Determines the size of the display area in the monitoring interface screen for the usage status information, and if the importance value of the additional monitoring item is greater than or equal to a threshold value, object information and environment information corresponding to the equipment usage status of the additional monitoring item are displayed in the monitoring displayed on the interface screen.

In addition, by analyzing target equipment information including the type of use, structure, and size of the target equipment, an item that can cause a safety accident among the management target items is selected and provided as the management setting interface.

The embodiments of the present invention disclosed in the present specification and drawings are only presented as specific examples to easily explain the technical content of the present invention and help understanding of the present invention, and are not intended to limit the scope of the present invention. It is obvious to those skilled in the art that other modified examples based on the technical idea of the present invention can be implemented in addition to the embodiments disclosed herein.

Claims (2)

In the safety management method providing a dynamic user interface using artificial intelligence,
Scanning safety target items according to each facility and structure using energy from the target equipment;
providing a management setting interface for selecting one or more management items desired to be managed among the safety target items;
If the management items are set by the management setting interface, performing a first monitoring according to the number and type of the management items;
selecting, as additional monitoring items, those determined to require monitoring by artificial intelligence, excluding the management items, among the management target items while performing the first monitoring; and
Including the step of providing a monitoring interface screen according to the equipment usage status information for the set management item and the selected additional monitoring item,
A value of importance of the additional monitoring item is calculated, a size of a display area in the monitoring interface screen for the equipment usage status information of the additional monitoring item is determined according to the value of importance, and the value of importance of the additional monitoring item is a threshold value. If this is the case, displaying object information and environment information corresponding to the equipment usage status of the additional monitoring item on the monitoring interface screen,
A safety management method that provides a dynamic user interface using artificial intelligence. The method of claim 1,
A dynamic user interface using artificial intelligence that is provided as the management setting interface by selecting an item that can cause a safety accident among the management target items by analyzing the target equipment information including the type of use, structure, and size of the target equipment Safety management methods provided.

Images