KR102545637B1 - Business management system and method using user terminal - Google Patents

Abstract

Disclosed are a business management system and a method using a user terminal. The business management system includes: a user terminal used by a worker; and a management device which estimates worker's status information using at least one of location information, beacon information, sensor information, data traffic information, video information, and text information received from the user terminal, and manages the worker using the estimated status information. Therefore, the worker can be monitored remotely in real time.

Description

Worker management system and method using user terminal {Business management system and method using user terminal}

The present invention relates to a worker management system, and more particularly, to a worker management system and method using a user terminal that performs access control, safety management, issue management, etc. using a user terminal.

Due to the recent rapid spread of mobile terminals, efforts are being made to utilize them in various fields, particularly in worker management. Due to these efforts, the manager can monitor the worker's condition anytime, anywhere, regardless of time or place.

However, as most of the workers are managed based on GPS information, which is location information of mobile terminals, it is difficult to precisely manage workers in indoor places where GPS information is incomplete or in areas with many buildings.

Accordingly, there is a need for a method capable of improving the accuracy and precision of worker management.

Korean Registered Patent Publication No. 10-2324843 (2021.11.04.)

The problem to be solved by the present invention is to provide a worker management system and method using a user terminal that performs access management, safety management, and issue management of the worker using the user terminal used by the worker remotely.

In order to solve the above problems, the worker management system according to the present invention uses at least one of a user terminal used by a worker and location information received from the user terminal, beacon information, sensor information, data traffic information, video information, and text information. and a management device for estimating state information of the worker and managing the worker using the estimated state information.

In addition, it further includes a beacon device installed in a designated area, and the management device transmits an activation signal to the beacon device when the location information received from the user terminal is included in the designated area, so that the beacon device is sent to the user terminal. Controlling to transmit a beacon signal, receiving beacon information included in the beacon signal from the user terminal, and determining that the worker entered the designated area when the received beacon information matches pre-stored reference beacon information characterized by

In addition, when another worker has entered and exited the designated area before the worker entered the designated area, the management device further transmits an activation signal to other user terminals used by the other worker when transmitting an activation signal to the beacon device. to control the other user terminal to transmit an authentication signal to the user terminal, receive the beacon information and beacon information and authentication information included in the authentication signal from the user terminal, and the received beacon information and the authentication information It is characterized in that it is determined that the operator has entered the designated area if it matches the previously stored standard beacon information and standard authentication information.

In addition, when access registration of the user terminal to the designated area is completed, the management device receives the beacon information from the user terminal at each predetermined period, and if the beacon information is not received even after the period, the operator It is characterized in that it is determined that it has deviated from the designated area.

In addition, the management device receives sensor information and data traffic information from the user terminal, uses the sensor information to estimate a body condition of the worker, and estimates a usage condition of the user terminal using the data traffic information. And, it is characterized in that the presence or absence of a safety accident of the worker is determined using the estimated physical condition and usage condition.

In addition, the management device receives at least one of text information and image information from the user terminal, generates an embedding vector representing a characteristic of at least one of the received text information and image information, and at least one learned analysis model. It is characterized in that issue information including information to be shared with other users is detected by analyzing the embedding vector through.

In addition, when the text information is received, the management device generates a text embedding vector by applying the text information to a natural language processing pretraining language model, and the text embedding is processed through a first analysis model that processes a long word sequence. The vector is analyzed, but first, the text embedding vector is applied to a sigmoid function in the output layer of the first analysis model to determine whether or not an issue is included in the text information, and if it is determined that the issue is included, the corresponding It is characterized in that the content related to the issue is classified by applying the text embedding vector to a softmax function in the output layer of the first analysis model.

In addition, when the image information is received, the management device generates an image embedding vector using a feature detection filter related to at least one of image coefficients, color division, and texture division of the image information, and converts the generated image embedding vector to It is characterized in that a probability for the presence or absence of issue information is calculated by applying to the second analysis model for which learning between a plurality of layers has been completed, and image information including issue information is detected according to the calculated probability.

In addition, when the issue information is included in the text information and the image information, the management device estimates a commonality of the issue information included in each piece of information, and combines the text information and the image information into one information by using the estimated commonality. It is characterized by integrating into

A worker management method according to the present invention includes the steps of collecting at least one of location information, beacon information, sensor information, data traffic information, image information, and text information by a user terminal used by a worker, and the collected information by a management device. Receiving and estimating state information of the worker using the received information, and managing the worker by the management device using the estimated state information.

The management device according to the present invention uses an information communication unit that receives at least one of location information, beacon information, sensor information, data traffic information, image information, and text information from a user terminal used by a worker, and the received information, Estimating state information of, and including a device control unit for managing the operator using the estimated state information.

According to an embodiment of the present invention, a worker can be remotely monitored in real time based on location information, sensor information, data traffic information, image information, and text information collected from a user terminal used by the worker.

In addition, artificial intelligence technology can be applied to the monitoring process to accurately estimate the worker's condition.

1 is a configuration diagram for explaining a worker management system according to an embodiment of the present invention.
2 is a block diagram for explaining a user terminal according to an embodiment of the present invention.
3 is a block diagram for explaining a management device according to an embodiment of the present invention.
FIG. 4 is a block diagram for explaining the device control unit of FIG. 3 .
5, 6, 7, 8, 9, and 10 are diagrams for explaining traffic information of a user terminal according to an embodiment of the present invention.
11 is a diagram for explaining a process of detecting issue information from text information according to an embodiment of the present invention.
12 and 13 are diagrams for explaining a process of detecting issue information from image information according to an embodiment of the present invention.
14 is a flowchart illustrating a process of registering access to a designated area according to an embodiment of the present invention.
15 is a flowchart illustrating a process of registering access to a designated area according to another embodiment of the present invention.
16 is a flowchart illustrating a process of canceling access registration of a designated area according to an embodiment of the present invention.
17 is a flowchart illustrating a process of managing worker's safety according to an embodiment of the present invention.
18 is a flowchart illustrating a process of managing a worker's issue according to an embodiment of the present invention.
19 is a block diagram illustrating a computing device according to an embodiment of the present invention.

Hereinafter, with reference to the accompanying drawings, embodiments of the present invention will be described in detail so that those skilled in the art can easily carry out the present invention. However, the present invention may be embodied in many different forms and is not limited to the embodiments described herein. And in order to clearly explain the present invention in the drawings, parts irrelevant to the description are omitted, and similar reference numerals are attached to similar parts throughout the specification.

In the present specification and drawings (hereinafter referred to as 'this specification'), redundant descriptions of the same components are omitted.

In addition, in this specification, when a component is referred to as being 'connected' or 'connected' to another component, it may be directly connected or connected to the other component, but another component in the middle It should be understood that may exist. On the other hand, in this specification, when a component is referred to as 'directly connected' or 'directly connected' to another component, it should be understood that no other component exists in the middle.

In addition, terms used in this specification are only used to describe specific embodiments and are not intended to limit the present invention.

Also, in this specification, a singular expression may include a plurality of expressions unless the context clearly indicates otherwise.

In addition, in this specification, terms such as 'include' or 'having' are only intended to designate that there is a feature, number, step, operation, component, part, or combination thereof described in the specification, and one or more It should be understood that the presence or addition of other features, numbers, steps, operations, components, parts, or combinations thereof is not precluded.

Also in this specification, the term 'and/or' includes a combination of a plurality of listed items or any item among a plurality of listed items. In this specification, 'A or B' may include 'A', 'B', or 'both A and B'.

Also, in this specification, detailed descriptions of well-known functions and configurations that may obscure the subject matter of the present invention will be omitted.

1 is a configuration diagram for explaining a worker management system according to an embodiment of the present invention.

Referring to FIG. 1 , the worker management system 500 remotely performs access management, safety management, and issue management of the corresponding worker using the user terminal 100 used by the worker. The worker management system 500 includes a user terminal 100 and a management device 200, and may further include a beacon device 300 and a manager terminal 400.

The user terminal 100 is a terminal used by a worker and includes a plurality of terminals 100a, 100b, and 100c. The user terminal 100 collects at least one of location information, sensor information, data traffic information, image information, and text information, and transmits the collected information to the management device 200 . Through this, the user terminal 100 supports management of workers in the management device 200 . Here, the location information may include Global Positioning System (GPS) information, the sensor information may include tilt information, speed information, acceleration information, and biometric information, and the data traffic information may be used by the user terminal 100. It means information about the amount of data being processed.

In the following description, the user terminal 100a is described as a terminal used by a worker to manage, and the user terminals 100b and 100c are described as terminals used by workers other than the worker to manage, but are not limited thereto.

The management device 200 is a device for monitoring the current situation of a worker and may be built in the form of a server. The management device 200 estimates the state information of the operator using at least one of location information, beacon information, authentication information, sensor information, data traffic information, image information, and text information received from the user terminal 100a. Here, the status information may include an access status for a designated area, a safety status for danger, an issue status for a specific event, and the like. At this time, the management device 200 may estimate state information corresponding to the received information. For example, the management device 200 may estimate an access state using at least one of location information, beacon information, and authentication information, and may estimate a safety state using sensor information and data traffic information, and image information. And an issue state can be estimated using text information. The management device 200 may transmit a notification message to a terminal corresponding to the type of estimated state information. In addition, the management device 200 may generate monitoring information for the worker's status information and transmit the generated monitoring information to the manager terminal 400 .

The beacon device 300 is installed for each designated area and includes a plurality of devices 300a, 300b and 300c. Here, the designated area includes a worker's working area, a dangerous area, a restricted area, and the like. The beacon device 300 maintains an idle state at normal times and converts it into an active state when receiving an activation signal from the management device 200, thereby minimizing power consumption. The beacon device 300 may transmit a beacon signal at predetermined intervals or transmit a beacon signal in real time. Here, the beacon signal includes beacon information including an identification number of the corresponding beacon device.

The manager terminal 400 is a terminal used by a manager, receives monitoring information about a worker, and outputs the received monitoring information. That is, the manager can check the worker's condition in real time using the monitoring information.

Meanwhile, the worker management system 500 establishes a communication network 550 between the user terminal 100, the management device 200, the beacon device 300, and the manager terminal 400 so that communication between them can be made. The communication network 550 may be composed of a backbone network and a subscriber network. The backbone network may be composed of one or a plurality of integrated networks among an X.25 network, a Frame Relay network, an ATM network, a Multi Protocol Label Switching (MPLS) network, and a Generalized Multi Protocol Label Switching (GMPLS) network. Subscriber networks include FTTH (Fiber To The Home), ADSL (Asymmetric Digital Subscriber Line), cable network, zigbee, Bluetooth, and wireless LAN (IEEE 802.11b, IEEE 802.11a, IEEE 802.11g, IEEE 802.11n ), Wireless Hart (ISO/IEC62591-1), ISA100.11a (ISO/IEC 62734), CoAP (Constrained Application Protocol), MQTT (Message Queuing Telemetry Transport), WIBro (Wireless Broadband), Wimax, 3G, HSDPA (High Speed Downlink Packet Access), 4G and 5G. In some embodiments, the communication network 150 may be an internet network or a mobile communication network. In addition, the communication network 550 may include all other well-known wireless communication or wired communication methods to be developed in the future.

2 is a block diagram for explaining a user terminal according to an embodiment of the present invention.

1 and 2, the user terminal 100a includes a communication unit 110, a GPS unit 120, a sensor unit 130, a control unit 140, an output unit 150, and a storage unit 160. do.

The communication unit 110 performs communication with other user terminals 100b and 100c, the management device 200, and the beacon device 300. The communication unit 110 receives an authentication signal from any one of the other user terminals 100b and 100c. The communication unit 110 transmits at least one of location information, beacon information, authentication information, sensor information, data traffic information, image information, text information, and terminal information to the management device 200, and from the management device 200 Receive notification messages. Also, the communication unit 110 receives a beacon signal from the beacon device 300 .

The GPS unit 120 measures location information about the current location of the user terminal 100a. That is, the GPS unit 120 receives signals transmitted from GPS satellites, calculates current coordinates of the user terminal 100a, and measures location information.

The sensor unit 130 measures tilt information, speed information, acceleration information, etc., which are sensor information of the user terminal 100a. Also, when the user terminal 100a is a wearable device, the sensor unit 130 may further measure biometric information. To this end, the sensor unit 130 may include a tilt sensor, a speed sensor, an acceleration sensor, a gyro sensor, an electrocardiogram sensor, an electromyogram sensor, an EEG sensor, and the like.

The controller 140 performs overall control of the user terminal 100a. The control unit 140 collects at least one of location information, sensor information, data traffic information, image information, and text information, and transmits the collected information to the management device 200 . In addition, the control unit 140 transmits beacon information included in the beacon signal, authentication information included in the authentication signal, and terminal information including the identification number of the terminal to the management device 200 . When receiving a notification message from the management device 200, the control unit 140 controls the corresponding notification message to be output. Here, the notification message may be a message informing permission to enter the designated area, a warning message related to a safety accident, a report message due to the occurrence of an issue, and the like.

The output unit 150 outputs a notification message. At this time, the output unit 150 may output a notification message as video information through a display and output a notification message as audio information through a speaker.

The storage unit 160 stores a program or algorithm for driving the user terminal 100a. The storage unit 160 stores at least one of location information, sensor information, data traffic information, image information, text information, and terminal information. Also, the storage unit 160 may store a notification message. The storage unit 160 may be a flash memory type, a hard disk type, a multimedia card micro type, a card type memory (eg SD or XD memory, etc.), RAM (Random Access Memory, RAM), SRAM (Static Random Access Memory), ROM (Read-Only Memory, ROM), EEPROM (Electrically Erasable Programmable Read-Only Memory), PROM (Programmable Read-Only Memory), magnetic memory, It may include at least one storage medium of a magnetic disk and an optical disk.

Meanwhile, since the other user terminals 100b and 100c have the same configuration as the user terminal 100a, a description thereof will be omitted.

3 is a block diagram for explaining a management device according to an embodiment of the present invention, FIG. 4 is a block diagram for explaining a device controller of FIG. 3, and FIGS. 5 to 10 are user according to an embodiment of the present invention. 11 is a diagram for explaining traffic information of a terminal, FIG. 11 is a diagram for explaining a process of detecting issue information from text information according to an embodiment of the present invention, and FIGS. 12 and 13 are diagrams according to an embodiment of the present invention It is a diagram for explaining a process of detecting issue information from image information.

Referring to FIGS. 1 and 3 to 13 , the management device 200 includes a device communication unit 210 , a device control unit 220 and a device storage unit 230 .

The device communication unit 210 communicates with a plurality of user terminals 100 , a plurality of beacon devices 300 and a manager terminal 400 . The device communication unit 210 receives at least one of location information, beacon information, authentication information, sensor information, and terminal information from the user terminal 100a used by the worker, and transmits a notification message to the user terminal 100a. . The device communication unit 210 transmits an activation signal to the other user terminals 100b and 100c used by other workers, and transmits a notification message to the other user terminals 100b and 100c. The device communication unit 210 transmits an activation signal/deactivation signal to the beacon device 300a installed in a designated area where a worker has entered or exited. The device communication unit 210 transmits a notification message to the manager terminal 400 .

The device controller 220 performs overall control of the management device 200 . The device control unit 220 estimates the state information of the operator using at least one of location information, sensor information, data traffic information, image information, and text information received from the user terminal 100a, and uses the estimated state information. manage workers To this end, the device control unit 220 includes a learning unit 221, an information collection unit 223, an access management unit 225, a safety management unit 227, and an issue management unit 229.

The learning unit 221 trains an analysis model for analyzing text information and image information. The learning unit 221 may include a first learning unit 221a for learning related to text information and a second learning unit 221b for learning related to image information.

The first learning unit 221a prepares text learning data including a text embedding vector for learning and a label corresponding to the text embedding vector for learning for learning the first analysis model. Here, the first analysis model may be a long short-term memory (LSTM) neural network capable of processing long word sequences, and a label corresponding to a text embedding vector for learning may be a label corresponding to an output value of an output layer. The structure of the LSTM neural network includes a plurality of hidden layers that perform calculations using memory cells, so that word sequences can be accurately processed. The first learning unit 221a includes a sigmoid function and a softmax function in the output layer of the LSTM neural network. The first learning unit 221a performs learning to determine whether an issue is included or not included in the text information for learning using a simoid function. If it is determined that the text information for learning includes an issue, the first learning unit 221a performs learning to classify content related to the issue included in the text information for learning using a softmax function. At this time, the first learning unit 221a compares the output value of the first analysis model with a label corresponding to the text embedding vector for learning, and performs optimization learning by modifying parameters of the first analysis model to minimize loss. This optimization may be repeatedly performed using a plurality of different training text embedding vectors, and such repetition may be performed until a desired accuracy is reached through an evaluation index.

The second learning unit 221b prepares image learning data including an image embedding vector for learning and a label corresponding to the image embedding vector for learning for learning the second analysis model using the image information for learning. Here, the image information for training may be image information including at least one issue, and the label corresponding to the image embedding vector for training includes a hard label and a soft label. The hard label is a label corresponding to the hard output value of the output layer, and the soft label is a label corresponding to the soft output value of the output layer. The soft output value is a value representing the probability of whether or not an issue exists, and the hard output value represents a numerical value representing the probability as a value that determines the presence or absence of an issue. For example, the soft output value is represented by [0.852, 0.148], and the hard output value is represented by [1, 0]. The second learning unit 221b creates a hard loss function for hard loss optimization by setting hyperparameters of the basic loss function. The second learning unit 221b inputs an image embedding vector for learning to the second analysis model. Here, the second analysis model is an artificial neural network, and may be a multilayer perceptron (MLP), but is not limited thereto. The second learning unit 221b converts the calculated soft output value into a hard output value when the second analysis model calculates a soft output value through a plurality of operations in which a plurality of inter-layer weights are applied to the image embedding vector for learning. The second learning unit 221b performs hard loss optimization by modifying the parameters of the second analysis model so that the hard loss representing the difference between the hard label and the hard output value is minimized using the converted hard output value and the hard loss function. This hard loss optimization is repeatedly performed using a plurality of different training image embedding vectors, and such repetition may be performed until a desired accuracy is reached through an evaluation index.

The information collection unit 223 preprocesses location information, beacon information, authentication information, sensor information, data traffic information, image information, and text information received from the device communication unit 210 . That is, the information collection unit 223 changes the format method into a format for estimating the state information of the operator to be performed later, or filters unnecessary or erroneous information.

The access management unit 225 manages access to a predetermined designated area using at least one of location information, beacon information, and authentication information transmitted from the information collection unit 223 . Here, the designated area includes a worker's working area, a dangerous area, a restricted area, and the like. The access management unit 225 can manage access of a worker to a designated area only by having the user terminal 100a without a separate operator operation. The access control unit 225 transmits an activation signal to the beacon device 300a installed in the designated area when the location information indicating the location of the worker is included in the designated area, so that the beacon device 300a transmits the beacon signal to the user terminal 100a. control to do Here, the activation signal is a wake-up signal, which activates the beacon device 300a in an idle state. The beacon signal is a signal including beacon information, and the beacon information includes an identification number of the beacon device 300a. When the access management unit 225 receives the beacon information included in the beacon signal, if the received beacon information matches the pre-stored reference beacon information, it is determined that the worker has entered the designated area, and if it does not match, the worker does not enter the designated area. judge that it is not That is, the access management unit 225 may perform access management through a dual-structured authentication procedure through location information and beacon information.

Meanwhile, the access management unit 225 may perform access management through a three-layered authentication procedure when another worker has entered or exited the designated area before the worker entered the designated area. In detail, when the access management unit 225 transmits an activation signal to the beacon device 300a during the dual-structure authentication process, the activation signal is further transmitted to one of the other user terminals 100b and 100c used by other workers so that the corresponding other user terminal 100b transmits an authentication signal to the user terminal 100a. The authentication signal is a signal for access-related authentication, and may include OTP (One Time Password) information generated according to the time when the activation signal is sent, but is not limited thereto. When the access control unit 225 receives the beacon information included in the beacon signal and the authentication information included in the authentication signal, the operator enters and exits the designated area when the received beacon information and authentication information match the pre-stored standard beacon information and standard authentication information. If it does not match, it is determined that the worker has not entered the designated area.

In addition, when it is determined that the worker has entered or exited, the access management unit 225 registers access to the corresponding user terminal 100a and transmits an access registration notification message to the corresponding user terminal 100a. At this time, the access management unit 225 may sort according to the access time for each user terminal registered to access the designated area using received terminal information such as beacon information, and update the sorted access registration list in real time.

Preferably, the access management unit 225 receives beacon information from the user terminal 100a at predetermined intervals even after the user terminal 100a completes access registration in the designated area. At this time, when the access control unit 225 does not receive the beacon information even after a predetermined period, it determines that the worker has left the designated area and cancels the access registration of the user terminal 100a.

The safety management unit 227 manages safety related to workers using sensor information and data traffic information transmitted from the information collection unit 223 . The safety management unit 227 can manage the safety state of the worker only by having the user terminal 100a without a separate manipulation by the worker. The safety management unit 227 estimates the current state of the user by using sensor information representing the body state of the worker and data traffic information representing the use state of the user terminal 100a. That is, when the sensor information includes gyro information, the safety management unit 227 estimates the physical state of the worker based on the tilt and orientation included in the gyro information, and when the sensor information includes electrocardiogram information, which is biometric information, the safety management unit 227 estimates the physical state of the operator. Based on the information, the worker's biological state is estimated. In addition, the safety management unit 227 estimates the usage state of the user terminal 100a by checking the amount of data currently used by the worker based on the data traffic information. In addition, the safety management unit 227 transmits and receives traffic information transmitted and received by the user terminal 100a periodically or continuously (background) and while performing a specific action (use of a specific application, etc.) of the user terminal 100a. The usage state of the user terminal 100a may be estimated by dividing the traffic.

For example, the safety management unit 227 determines that the user terminal 100a is Android, and transmits/receives TCP (Transmission　Control　protocol), UDP (user datagram protocol), SSL (Secure Sockets Layer), and DNS (domain name system) traffic, etc., when there is no traffic information caused by the user using a specific application or Android function, various unstructured network packets It is assumed that it is not used (FIG. 5). The safety management unit 227 detects that the user terminal 100a is Android-based, accesses a specific server IP (internet protocol) to transmitted and received information, detects a specific GET, POST message, detects a specific TCP Push ack through data area analysis, and detects a specific TCP Push If information such as detection through analysis of ack capacity and occurrence pattern, detection of specific UDP data area, detection of domain movement path through DNS packet analysis, etc. is included, it is assumed that Android is used (FIG. 6). The safety management unit 227 detects that the user terminal 100a is Android-based, and in transmitted and received information, service server IPs dedicated to C and N companies, keywords unique to C and N companies in the data area of the TCP Push Ack packet, If information such as analysis of capacity occurrence patterns between Push Ack packets and Ack packets is included, it is assumed that Company C applications (Fig. 7(a)) and N Company applications (Fig. 7(b)) are used.

In addition, the safety management unit 227 is configured so that the user terminal 100a is an IOS family and transmits and receives various unstructured network packets such as TCP, UDP, SSL, and DNS traffic to information transmitted and received without traffic information caused by the user using a specific application or IOS function. If there is none, it is assumed that the IOS is not used (FIG. 8). The safety management unit 227 detects that the user terminal 100a is an IOS family, accesses a specific server IP, detects a specific GET, POST message, analyzes a specific TCP Push ack data area, and detects the capacity and occurrence of a specific TCP Push ack in transmitted and received information. If information such as detection through pattern analysis, detection of a specific UDP data area, detection of a domain movement path through DNS packet analysis, etc. is included, it is assumed that IOS is used (FIG. 9). The safety management unit 227 detects that the user terminal 100a is of the IOS family, and in transmitted and received information, service server IPs dedicated to companies C and N, keywords unique to companies C and N within the data area of TCP Push Ack packets, Push Ack packets and capacity generation pattern analysis between Ack packets, it is assumed that the C company application (FIG. 10(a)) and the N company application (FIG. 10(b)) are used.

On the other hand, when no traffic information is received from the user terminal 100a even after a preset period of time, the safety management unit 227 mixes the worker's biometric state estimated from the sensor information and previously estimated traffic information to expose the worker to risk. condition can be further estimated.

In this way, the safety management unit 227 determines whether the worker has a safety accident by using the current state including at least one of the worker's physical condition, the use state of the user terminal 100a, and the risk exposure state. Preferably, the safety management unit 227 determines whether there is a safety accident of the worker based on the use state of the user terminal 100a when the physical condition of the worker is not measured, and the use state of the user terminal 100a is not measured. If not, the worker's safety accident is determined based on the worker's physical condition. Here, the safety accident may include a fall, fall, cardiac arrest, and the like.

When it is determined that a safety accident has occurred, the safety management unit 227 transmits a safety accident notification message to the manager terminal 400 . At this time, the safety management unit 227 may transmit a notification message about a safety accident to the user terminal 100a, and when other user terminals 100b and 100c exist around the user terminal 100a, the other user terminal ( In steps 100b and 100c), a notification message for safety accidents may be transmitted.

The issue management unit 229 manages issues related to workers by using at least one of text information and image information delivered from the information collection unit 223 . The issue management unit 227 may manage issues of the worker only by having the user terminal 100a without a separate manipulation by the worker. That is, the issue management unit 227 generates an embedding vector representing at least one feature of text information and image information, and analyzes the embedding vector through the analysis model learned by the learning unit 221 to obtain information to be shared with other users. Detect issue information including

In detail, the issue management unit 229 generates a text embedding vector by applying the text information to a BERT model, which is a natural language processing pre-training language model for recognizing the intent of a query. The BERT model improves the accuracy of the existence of issues and identification of issues through pre-training embedding before identifying issues. The issue manager 227 analyzes the text embedding vector by applying it to the learned first analysis model. The first analysis model may be a long short-term memory (LSTM) neural network capable of processing long word sequences. The structure of the LSTM neural network includes a plurality of hidden layers that perform calculations using memory cells, so that word sequences can be accurately processed. The issue manager 229 includes a sigmoid function and a softmax function in the output layer of the LSTM neural network. The issue management unit 229 first applies the text embedding vector to the simoid function of the output layer to determine whether issue-related text is included in the text information or not. If it is determined that the text information includes issue-related text, the issue management unit 229 classifies the issue-related text included in the text information by applying the corresponding text embedding information to the softmax function in the output layer.

For example, the issue management unit 229 generates a text embedding vector by applying the text information to a BERT model when text information including text for a meeting appointment (corresponding to an issue) is transmitted, and generates a text embedding vector through an LSTM neural network. Enter as an input value for The issue management unit 229 confirms that the issue-related text is included by using the LSTM neural network with the sigmoid function as the output layer, and then sends the issue-related text to the attendees of the meeting appointment, using the LSTM neural network with the softmax function as the output layer. It can be classified by time, place, content of meeting, etc. Here, the first analysis model consists of an LSTM neural network that divides only the output layer into a simoid function and a softmax function while sharing an input layer and a plurality of hidden layers, or a simoid function and a softmax function. Consisting of each LSTM neural network as an output layer It can be.

The issue management unit 229 generates an image embedding vector using a feature detection filter related to at least one of image coefficient, color segmentation, and texture segmentation with respect to image information.

The issue management unit 229 may generate an image embedding vector using a first filter F1 that is an image coefficient filter. The issue management unit 229 divides the image information into a plurality of square sub-blocks, performs a discrete cosine transform (DCT) on each of the divided sub-blocks, converts them into frequency bands, and converts them into frequency bands. An image embedding vector may be generated by quantizing each of the transformed subblocks.

The issue management unit 229 may generate an image embedding vector using a second filter F2 that is a color segmentation filter. The issue management unit 229 divides image information into three channels including hue, saturation, and value, and performs a Fourier transform on each of the three divided channels in the frequency domain. After transforming to , an image embedding vector may be generated by quantizing a frequency domain value into a preset number of bands (eg, 10 bands).

The issue management unit 229 may generate an image embedding vector using a third filter F3 that is a texture segmentation filter. The issue management unit 229 divides image information into three channels including the Y channel, which is a component representing luminance, and the I (In-phase) channel and Q (Quadrature) channel, which are components representing color difference, and After transforming each of the three channels into a frequency domain through wavelet transform, an image embedding vector may be generated by quantizing values in the frequency domain into a predetermined number of bands (eg, 10 bands).

When the image embedding vector is generated, the issue management unit 229 analyzes the image embedding vector through a second analysis model to detect image information including an issue. At this time, the second analysis model analyzes the image embedding vector to calculate a probability of whether an issue exists, and the issue management unit 229 detects image information including an issue according to the calculated probability.

The second analysis model may be a multilayer perceptron as shown in FIG. 6, but is not limited thereto. The second analysis model includes a plurality of layers (IL, HL, OL). The plurality of layers include an input layer (IL), a plurality of hidden layers (HL: HL1 to HLk), and an output layer (OL).

In addition, each of the plurality of layers IL, HL, and OL includes a plurality of nodes. For example, the input layer IL may include n input nodes i1 to in, and the output layer OL may include two output nodes P and N. In addition, among the hidden layers (HL), the first hidden layer (HL1) includes a number of nodes (h11 to h1a), the second hidden layer (HL2) includes b number of nodes (h21 to h2b), kth hidden The layer HLk may include c nodes hk1 to hkc.

All of the plurality of nodes included in the plurality of layers have operations. In particular, a plurality of nodes of different layers are connected by a channel (indicated by a dotted line) having a weight (W). In other words, the calculation result of one node is weighted and becomes the input of the next layer node.

Meanwhile, a node h according to an embodiment of the present invention is shown in FIG. 7 . This node (h) may be any one node of the detection model (DM). Node h is input signal x = [x1, x2, ... , xn] with weights w=[w1, w2, … , wn], take the function F as the result. Here, function F is an activation function. Activation functions may include sigmoid, softmax, hyperbolic tangent (tanh), exponential linear unit (ELU), rectified linear unit (ReLU), leaky ReLU, Maxout, Minout, and the like. Any one of these activation functions can be selected and used.

The output of each node is as shown in [Equation 1] below.

의 값이 임계치 보다 작을 때 해당 노드가 활성화되지 않도록 하는 역할을 한다. Here, among the unexplained parameters, b is a threshold value, and this threshold value in [Equation 1]

It serves to prevent the corresponding node from being activated when the value of is smaller than the threshold.

According to [Equation 1], even when the input is the same, the output becomes a different value according to the weight. For example, it is assumed that there are 3 nodes in the previous layer of node h. Accordingly, there are three inputs (n=3) X1, X2, and X3 and three weights W1, W2, and W3 for the corresponding node. Node h receives a value obtained by multiplying the three inputs X1, X2, and X3 by weights W1, W2, and W3, sums them all, and substitutes the summed value into an activation function to calculate an output. Specifically, it is assumed that the inputs [X1, X2, X3] = 0.5, -0.3, 0, and the weight w = [W1, W2, W3] = 4, 5, 3. In addition, assuming that the transfer function is 'sgn()' for convenience of description, the output value is calculated as follows.

x1 × w1 = 0.5 × 4 = 2

x2 × w2 = -0.3 × 5 = -1.5

x3 × w3 = 0 × 3 = 0

2 + (-1.5) + 0 = 0.5

sgn(0.5) = 1

In this way, any one node of the second analysis model receives a value obtained by applying a weight to a node value of the previous layer as an input, sums the values, performs an operation according to an activation function, and transfers the result to the input of the next layer.

In summary, in the second analysis model, any one node of any one layer receives a value obtained by applying weights to inputs from nodes of the previous layer, sums them, takes an activation function, and passes this result to the input of the next layer. . Accordingly, when the image embedding vector is input to the input layer (IL) of the second analysis model, the second analysis model performs a plurality of operations to which weights of a plurality of layers (IL, HL, OL) are applied to the image embedding vector Then, the probability of the presence or absence of an issue is calculated. Then, the issue management unit 229 detects image information including an issue according to the calculated probability.

Referring back to FIG. 6 , when the image embedding vector is input to the plurality of input nodes i1 to in of the input layer IL of the second analysis model, the plurality of first hidden nodes h11 of the first hidden layer HL1 ~ h1a) Each receives a value to which a weight is applied to the image embedding vector of a plurality of input nodes (i1 to in) (indicated by a dotted line), sums all the input values, and then calculates the summed values according to the activation function. An operation is performed to calculate a plurality of first hidden node values. Then, each of the plurality of second hidden nodes h21 to h2b of the second hidden layer HL2 receives a value to which a weight is applied to each of the plurality of first hidden node values of the plurality of first hidden nodes h11 to h1a, (indicated by a dotted line), all input values are summed, and a plurality of second hidden node values are calculated by performing an operation according to an activation function on the summed values. In this way, in the hidden layer (HL), a previous node value is transmitted with a weight applied, and a current node value is calculated through an operation. By repeating this process, a plurality of k-th hidden node values of a plurality of k-th hidden nodes (hk1 to hkc) of the k-th hidden layer (HLk) can be calculated. Then, each of the output nodes P and N applies a weight w = [w1, w2, . . . , wc × 2] is applied (indicated by a dotted line), and after summing up all the input values, an operation is performed on the summed values according to an activation function to calculate an output value.

Each of the two output nodes P and N of the output layer OL corresponds to a case with an issue and a case without an issue. That is, the first output node P corresponds to the case where there is an issue, and the second output node N corresponds to the case where there is no issue. Accordingly, the output value of the first output node (P) is the probability that there is at least one issue corresponding to the image embedding vector input to the input layer (IL), and the output value of the second output node (N) is the input image embedding Represents the probability that there is no at least one issue corresponding to the vector. For example, if the output values of the first and second output nodes P and N are 0.852 and 0.148, respectively, the probability that there is at least one issue corresponding to the image embedding vector is 85%, and there is at least one issue corresponding to the image embedding vector. There is a 15% chance that there will be no issue of In this way, if the second analysis model outputs probabilities (0.852, 0.148), the issue manager 229 determines that there is at least one issue corresponding to the image embedding vector according to these probabilities (0.852, 0.148).

The issue management unit 229 transmits issue information regarding text information and video information to at least one of the user terminal 100a, other user terminals 100b and 100c, and the manager terminal 400. At this time, when the issue information is included in both the text information and the image information, the issue management unit 229 estimates the commonality of the issue information included in each information, and combines the text information and the image information into one information by using the estimated commonality. can be integrated into The issue manager 229 transmits the integrated issue information to at least one of the user terminal 100a, other user terminals 100b and 100c, and the manager terminal 400.

The device storage unit 230 stores a program or algorithm for driving the management device 200 . The device storage unit 230 stores at least one of location information, sensor information, data traffic information, image information, text information, and terminal information. The device storage unit 230 stores standard beacon information and standard authentication information that are comparison standards for confirming access of workers. The device storage unit 230 stores learning data for learning the first analysis model and the second analysis model. In addition, the device storage 230 may store phrases for notification messages. The device storage unit 230 is a flash memory type, hard disk type, media card micro type, card type memory (eg SD or XD memory, etc.), RAM, SRAM, ROM, EEPROM, PROM, magnetic memory, magnetic disk and an optical disk.

14 is a flowchart illustrating a process of registering access to a designated area according to an embodiment of the present invention, and FIG. 15 is a flowchart illustrating a process of registering access to a designated area according to another embodiment of the present invention. 16 is a flowchart illustrating a process of canceling access registration of a designated area according to an embodiment of the present invention.

Referring to FIGS. 14 to 16 , in the worker management method, access management of the corresponding worker is performed remotely using a user terminal used by the worker. Access management among worker management methods may include various embodiments, and three embodiments are described in the present invention.

If there is no terminal registered for access to the current designated area, which is the first embodiment, the worker management method may perform terminal registration through the following process.

In step S101, the user terminal 100a collects location information. The user terminal 100a collects current location information through GPS information.

In step S103, the user terminal 100a transmits location information to the management device 200.

In step S105, the management device 200 determines whether the received location information is a designated area. Here, the designated area includes a worker's working area, a dangerous area, a restricted area, and the like. The management device 200 checks whether the location information is included in the designated area, performs step S107 if included in the designated area, and branches to step S101 if not included.

In step S107, the management device 200 transmits an activation signal to the beacon device 300a. Here, the activation signal may be a wake-up signal, and the beacon device 300a is located within a designated area.

In step S109, the beacon device 300a activates the system. The beacon device 300a may be activated when an activation signal is received while maintaining an idle state in normal times.

In step S111, the beacon device 300a transmits a beacon signal to the user terminal 100a. The beacon device 300a may transmit a beacon signal in a broadcast manner.

In step S113, the user terminal (100a) checks the beacon information. The user terminal 100a detects and confirms beacon information included in the beacon signal. Here, the beacon information may include an identification number of the beacon device 300a.

In step S115, the user terminal (100a) transmits beacon information and terminal information to the management device (200).

In step S117, the management device 200 determines whether the operator has entered or not. If the beacon information matches the pre-stored standard beacon information, the management device 200 determines that the operator has entered and exits and performs step S119, and if it does not match, it determines that the operator has not entered and branches to step S101.

In step S119, the management device 200 registers the access terminal for the user terminal 100a. The management device 200 registers the access terminal by using the terminal identification number included in the terminal information and the access time of the user terminal 100a.

In step S121, the management device 200 transmits an access registration notification message to the user terminal 100a.

In step S122, the user terminal 100a outputs an access registration notification message.

In the second embodiment, when at least one terminal is registered for access to the current designated area, the worker management method may perform terminal registration through the following process.

In step S201, the user terminal 100a collects location information. The user terminal 100a collects current location information through GPS information.

In step S203, the user terminal 100a transmits location information to the management device 200.

In step S205, the management device 200 determines whether the received location information is a designated area. The management device 200 checks whether the location information is included in the designated area, performs step S207 if included in the designated area, and branches to step S201 if not included.

In step S207, the beacon device 300a transmits a beacon signal to the user terminal 100a. Since the beacon device 300a maintains an activated state when at least one operator exists in the designated area, a separate activation signal may not be received.

In step S209, the management device 200 transmits an activation signal to other user terminals 100b and 100c. Here, the other user terminals 100b and 100c may be terminals currently registered for access to the designated area.

In step S211, the other user terminals 100b and 100c activate the system.

In step S213, the other user terminals 100b and 100c transmit authentication signals to the user terminal 100a. At this time, the other user terminals 100b and 100c are deactivated after transmitting the authentication signal. Here, the authentication signal is a signal for access-related authentication, and may include, but is not limited to, OTP information generated according to the time when the activation signal is sent.

In step S215, the user terminal (100a) checks the beacon information and authentication information. The user terminal 100a detects and confirms beacon information included in the beacon signal and authentication information included in the authentication signal.

In step S217, the user terminal (100a) transmits beacon information, authentication information and terminal information to the management device (200).

In step S219, the management device 200 determines whether the operator has entered or not. If the beacon information and authentication information match the pre-stored standard beacon information and standard authentication information, the management device 200 determines that the worker has entered and performs step S221, and if they do not match, determines that the worker has not entered and entered step S201. branch into

In step S221, the management device 200 registers the access terminal for the user terminal 100a. The management device 200 registers the access terminal by using the terminal identification number included in the terminal information and the access time of the user terminal 100a.

In step S223, the management device 200 transmits an access registration notification message to the user terminal 100a.

In step S225, the user terminal 100a outputs an access registration notification message.

When the access registration of the user terminal 100a to the designated area, which is the third embodiment, is completed, the worker management method may perform terminal registration cancellation through the following process.

In step S301, the beacon device 300a transmits a beacon signal to the user terminal 100a.

In step S303, the user terminal (100a) checks the beacon information. The user terminal 100a detects and confirms beacon information included in the beacon signal.

In step S305, the user terminal (100a) transmits beacon information and terminal information to the management device (200).

In step S307, the management device 200 determines whether beacon information is received within a preset period. If the beacon information is received within a predetermined period, the management device 200 determines that the operator is located in the designated area and branches to step S301, and if not received, determines that the operator is not located in the designated area and performs step S309.

In step S309, the management device 200 cancels access registration of the user terminal 100a. The management device 200 updates the access registration list after excluding the user terminal 100a from the access registration list based on the terminal information.

In step S311, the management device 200 determines whether there is a user terminal currently being registered for access by using the access registration list. The management device 200 branches to step S301 if there is a user terminal currently registering access, and performs step S313 if there is no user terminal currently registering access.

In step S313, the management device 200 transmits an inactivation signal to the beacon device 300a. Here, the deactivation signal may be a signal that converts the beacon device 300a into an idle state.

In step S315, the beacon device 300a is deactivated. The beacon device 300a is converted to an idle state to minimize power consumption.

17 is a flowchart illustrating a process of managing worker's safety according to an embodiment of the present invention.

Referring to FIG. 17 , in the worker management method, safety management of the worker is performed remotely using a user terminal used by the worker.

In step S401, the user terminal 100a collects at least one of sensor information and data traffic information. Here, sensor information may include inclination information, speed information, acceleration information, biometric information, etc., and data traffic information refers to information about the amount of data used by the user terminal 100a.

In step S403, the user terminal (100a) transmits the collected information to the management device (200). For example, when the collected information is only sensor information, the user terminal 100a may transmit only the sensor information to the management device 200, and when only data traffic information is collected, the management device 200 transmits only the data traffic information. It can be transmitted to, and when sensor information and data traffic information are collected, both information can be transmitted to the management device 200.

In step S405, the management device 200 estimates the state of the operator using the received information. The management device 200 can manage the safety state of the worker only by having the user terminal 100a without a separate manipulation of the worker. If the sensor information includes gyro information, the management device 200 estimates the physical state of the worker based on the tilt and orientation included in the gyro information, and if the sensor information includes electrocardiogram information, which is biometric information, based on the electrocardiogram information. Estimate the biological state of the worker. In addition, the management device 200 estimates the usage state of the user terminal by checking the amount of data currently used by the operator based on the data traffic information. In addition, the management device 200 transmits and receives traffic information transmitted and received by the user terminal 100a periodically or continuously (background) and while performing a specific action (use of a specific application, etc.) of the user terminal 100a. The usage state of the user terminal 100a may be estimated by dividing the traffic.

In step S407, the management device 200 determines whether a safety accident has occurred by using the estimated physical state of the worker and the use state of the user terminal. Preferably, the management device 200 determines whether there is a safety accident of the worker based on the use state of the user terminal 100a when the physical condition of the worker is not measured, and the use state of the user terminal 100a is not measured. If not, the worker's safety accident is determined based on the worker's physical condition. The management device 200 performs step S409 when it is determined that a safety accident has occurred, and branches to step S401 when it is determined that no safety accident has occurred.

In step S409, the management device 200 transmits a notification message related to the safety accident to the manager terminal 400. In addition, the management device 200 may also transmit a notification message related to the safety accident to other user terminals 100b and 100c located around the user terminal 100a.

In step S411, the manager terminal 400 and other user terminals 100b and 100c output notification messages.

In step S413, the management device 200 may additionally transmit a notification message related to the safety accident to the user terminal 100a.

In step S415, the user terminal 100a outputs a notification message.

18 is a flowchart illustrating a process of managing a worker's issue according to an embodiment of the present invention.

Referring to FIG. 18 , in the worker management method, issues of the worker are managed remotely using a user terminal used by the worker.

In step S501, the user terminal 100a collects at least one of text information and image information. Here, the text information may be information including text described in chatting, social network service (SNS), and the like, and the image information may be information including photos and videos.

In step S503, the user terminal 100a transmits the collected information to the management device 200. For example, the user terminal 100a may transmit only text information to the management device 200 when only text information is collected, and transmit only image information to the management device 200 when only image information is collected. And, when text information and image information are collected, both information can be transmitted to the management device 200.

In step S505, the management device 200 detects the worker's issue using the received information. The management device 200 can manage issues of a worker only by having the user terminal 100a without a separate manipulation by the worker. When the text information is received, the management device 200 generates a text embedding vector by applying the text information to a natural language processing pre-training language model, analyzes the text embedding vector through a first analysis model that processes a long word sequence, and solves the issue. detect When the image information is received, the management device 200 generates an image embedding vector by using a feature detection filter related to at least one of image coefficients, color division, and texture division of the image information, and the generated image embedding vector is divided into a plurality of hierarchical layers. It is applied to the second analysis model for which interlearning has been completed to calculate the probability of the presence or absence of issue information, and image information including issue information is detected according to the calculated probability.

In step S507, the management device 200 determines whether an issue has occurred by using the detected issue result. The management device 200 performs step S509 when it is determined that an issue has occurred, and branches to step S501 when it is determined that no issue has occurred.

In step S509, the management device 200 transmits a notification message related to the generated issue to the manager terminal 400. In addition, the management device 200 may transmit a notification message related to an issue to other user terminals 100b and 100c located around the user terminal 100a.

In step S511, the manager terminal 400 and other user terminals 100b and 100c output notification messages.

In step S513, the management device 200 may additionally transmit a notification message related to the issue to the user terminal 100a.

In step S515, the user terminal 100a outputs a notification message.

19 is a block diagram illustrating a computing device according to an embodiment of the present invention.

Referring to FIG. 19 , the computing device TN100 may be a device (eg, a user terminal, a management device, a beacon device, a manager terminal, etc.) described in this specification.

The computing device TN100 may include at least one processor TN110, a transceiver TN120, and a memory TN130. In addition, the computing device TN100 may further include a storage device TN140, an input interface device TN150, and an output interface device TN160. Elements included in the computing device TN100 may communicate with each other by being connected by a bus TN170.

The processor TN110 may execute program commands stored in at least one of the memory TN130 and the storage device TN140. The processor TN110 may mean a central processing unit (CPU), a graphics processing unit (GPU), or a dedicated processor on which methods according to embodiments of the present invention are performed. Processor TN110 may be configured to implement procedures, functions, methods, and the like described in relation to embodiments of the present invention. The processor TN110 may control each component of the computing device TN100.

Each of the memory TN130 and the storage device TN140 may store various information related to the operation of the processor TN110. Each of the memory TN130 and the storage device TN140 may include at least one of a volatile storage medium and a non-volatile storage medium. For example, the memory TN130 may include at least one of read only memory (ROM) and random access memory (RAM).

The transmitting/receiving device TN120 may transmit or receive a wired signal or a wireless signal. The transmitting/receiving device TN120 may perform communication by being connected to a network.

Meanwhile, the embodiments of the present invention are not implemented only through the devices and/or methods described so far, and may be implemented through a program that realizes functions corresponding to the configuration of the embodiments of the present invention or a recording medium in which the program is recorded. And, such implementation can be easily implemented by those skilled in the art from the description of the above-described embodiment.

Although the embodiments of the present invention have been described in detail above, the scope of the present invention is not limited thereto, and various modifications and improvements of those skilled in the art using the basic concept of the present invention defined in the following claims are also provided. fall within the scope of the invention.

100: user terminal
110: communication department
120: GPS unit
130: sensor unit
140: control unit
150: display unit
160: storage unit
200: management device
210: device communication unit
220: device control unit
230: device storage unit
300: beacon device
400: manager terminal
500: worker management system
550: communication network

Claims (11)

User terminals used by workers;
Beacon devices installed in designated areas; and
State information of the worker is estimated using at least one of location information, beacon information, sensor information, data traffic information, video information, and text information received from the user terminal, and the worker is monitored using the estimated state information. Including; management device for managing;
If other workers are staying in the designated area first before the worker enters and exits the designated area,
The management device,
An activation signal is transmitted to the beacon device to control the beacon device to transmit a beacon signal to the user terminal, and an activation signal is transmitted to one of the other user terminals used by the other worker so that the other user terminal is the user terminal. Control to transmit an authentication signal to the terminal,
When receiving the beacon information included in the beacon signal and the authentication information included in the authentication signal from the user terminal, respectively, the beacon information and the authentication information are compared with pre-stored standard beacon information and standard authentication information, respectively, to determine whether they match. Check, and if all of the confirmed results match, it is characterized in that it is determined that the worker has entered the designated area. According to claim 1,
The management device,
When the location information received from the user terminal is included in the designated area, an activation signal is transmitted to the beacon device to control the beacon device to transmit a beacon signal to the user terminal,
Receiving beacon information included in the beacon signal from the user terminal, and determining that the worker has entered the designated area when the received beacon information matches pre-stored reference beacon information Worker management system. delete According to claim 1 or 2,
When access registration of the user terminal is completed in the designated area,
The management device,
Worker management system, characterized in that for receiving the beacon information from the user terminal every predetermined period, and determining that the worker has deviated from the designated area when the beacon information is not received even after the period. According to claim 1,
The management device,
Receiving sensor information and data traffic information from the user terminal, estimating the worker's body state using the sensor information, estimating the use state of the user terminal using the data traffic information, and estimating the estimated body state A worker management system, characterized in that for determining whether or not there is a safety accident of the worker using the state and use state. According to claim 1,
The management device,
Receiving at least one of text information and image information from the user terminal, generating an embedding vector representing a feature of at least one of the received text information and image information, and generating the embedding vector through at least one learned analysis model. A worker management system characterized by detecting issue information including information to be analyzed and shared with other users. According to claim 6,
The management device,
When the text information is received, a text embedding vector is generated by applying the text information to a natural language processing pre-training language model, and the text embedding vector is analyzed through a first analysis model that processes a long word sequence,
First, the text embedding vector is applied to a sigmoid function in the output layer of the first analysis model to determine whether or not there is an issue included in the text information, and if it is determined that the text information includes the issue, the corresponding text embedding vector A worker management system, characterized in that by applying to the softmax (softmax) function of the output layer of the first analysis model to classify the content related to the issue. According to claim 6,
The management device,
When the image information is received, an image embedding vector is generated using a feature detection filter related to at least one of image coefficients, color segmentation, and texture segmentation of the image information, and the generated image embedding vector is subjected to learning between a plurality of layers. A worker management system, characterized in that applying to the completed second analysis model to calculate the probability of the presence or absence of issue information, and detecting image information including issue information according to the calculated probability. According to claim 6,
The management device,
When the issue information is included in the text information and the image information, a commonality of the issue information included in each information is estimated, and the text information and the image information are integrated into one information using the estimated commonality. worker management system. In the worker management method performed by a worker management system including a user terminal used by a worker, a beacon device installed in a designated area, and a management device for managing workers,
Collecting, by the user terminal, at least one of location information, beacon information, sensor information, data traffic information, image information and text information;
Receiving the collected information by a management device, and estimating state information of the worker using the received information; and
Including; managing the operator using the estimated state information by the management device;
If other workers are staying in the designated area first before the worker enters the designated area,
Transmitting, by the management device, an activation signal to any one of the other user terminals used by the beacon device and the other worker;
Transmitting, by the beacon device, a beacon signal to the user terminal;
Transmitting an authentication signal to the user terminal by another worker terminal having received the activation signal; and
Further comprising, by the user terminal, transmitting beacon information included in the beacon signal and authentication information included in the authentication signal to the management device,
The estimating step is
The beacon information and the authentication information are compared with pre-stored standard beacon information and standard authentication information, respectively, to determine whether they match, and if all of the checked results match, it is determined that the operator has entered the designated area. How to manage workers who do. Information communication unit for receiving at least one of location information, beacon information, sensor information, data traffic information, image information and text information from the user terminal used by the operator; and
A device control unit for estimating state information of the worker using the received information and managing the worker using the estimated state information; including,
If other workers are staying in the designated area first before the worker enters the designated area,
The device control unit,
An activation signal is transmitted to a beacon device installed in a designated area to control the beacon device to transmit a beacon signal to the user terminal, and an activation signal is transmitted to one of the other user terminals used by the other worker to transmit the beacon signal to the other user terminal Control to transmit an authentication signal to the user terminal,
When receiving the beacon information included in the beacon signal and the authentication information included in the authentication signal from the user terminal, respectively, the beacon information and the authentication information are compared with pre-stored standard beacon information and standard authentication information, respectively, to determine whether they match. and if all of the confirmed results match, it is determined that the worker has entered and exited the designated area.

Images