KR102454671B1 - SYSTEM AND METHOD FOR MANAGING SAFETY IN WORK SITE THROUGH IoT SENSOR DATA AND VIDEO DATA ANALYSIS BASED ON ARTIFICIAL INTELLIGENCE - Google Patents

Abstract

A work site safety management system and method through AI-based IoT sensor data and image data analysis are provided. According to various embodiments of the present invention, the work site safety management system includes: an IoT sensor module and a camera module placed in a work site or attached to at least a part of the body of a worker located in the work site; and a safety management server that provides safety management solutions to the worker. The safety management server analyzes at least one pieces of sensor data collected from the IoT sensor module and image data collected from the camera module to predict the risk of the work site or the worker, and provides notification of the predicted risk to the worker as the safety management solution.

Description

SYSTEM AND METHOD FOR MANAGING SAFETY IN WORK SITE THROUGH IoT SENSOR DATA AND VIDEO DATA ANALYSIS BASED ON ARTIFICIAL INTELLIGENCE

Various embodiments of the present invention relate to a workplace safety management system and method through AI-based IoT sensor data and image data analysis.

Manufacturing refers to an industry-oriented industrial system produced in factories, and is an industry that mass-produces and provides products by processing raw materials with manpower, mechanical power, or other powers. As of 2017, the manufacturing industry accounted for 29.6% of GDP, 90% of exports, and 56% of facility investment. Compared to the average wage in the service industry of 3.21 million won, the average wage in the manufacturing industry is 3.69 million won, which is the basis of economic development and a source of quality jobs and innovative growth.

According to this trend, various technologies related to manufacturing are being developed, and manufacturing plants have also been upgraded based on these technologies, but most of them are only manufacturing and production related technologies such as manufacturing technology innovation and production efficiency improvement. The development of technologies to improve factors such as workers' safety, health, and environment is relatively backward.

In fact, despite the advancement of manufacturing technology, the number of injuries and deaths caused by falling, pinching, bumping, crushing, exposure to harmful gases and suffocation, that is, the number of injuries and deaths due to industrial accidents, continues to increase, but industrial accidents nevertheless occur Only a standardized guideline for prevention is presented, and the development of technology to establish an environment where workers can work safely is not progressing properly.

Even now, manufacturing plants in various fields such as semiconductors, chemicals, automobiles, and steel are exposed to the risk of industrial accidents, and the number of injured or dead due to industrial accidents is continuously increasing, among which the steel industry is representative.

The process of melting iron ore in a steelworks uses coke, which is used as a fuel by processing coal and is manufactured in a coke plant.

The coke oven of the coke plant produces coke by charging coal in the carbonization chamber and carbonizing it with indirect heat heated in the combustion chamber. Five jobs are being performed on average.

On the other hand, in the coke manufacturing process, various types of by-product gases such as coke produced gas (COG), blast furnace produced gas (BFG), and converter produced gas (LDG) are generated during the production of coke. These by-product gases are toxic and flammable. Although it is made of a material and can cause serious human casualties such as gas poisoning and suffocation when gas is leaked, there is a problem in that the conventional coke manufacturing process is not properly prepared for such by-product gas leakage.

In fact, in the conventional coke manufacturing process, as a measure to prevent such gas poisoning accidents, workers directly inspect gas facilities in groups of two, perform gas detection before work to check the concentration, and must wear protective equipment such as a mask. However, even such basic safety management is not properly observed.

In addition, in the conventional coke manufacturing process, the manager of the job site directly walks around the job site and instructs the workers to work in compliance with safety standards, or records the job site image taken through a camera installed at the job site (eg CCTV camera). Because it was controlled only through control and instructions by the manager, such as monitoring whether workers work in compliance with safety standards through the Therefore, there is a problem in that it is difficult to check and instruct these tasks one by one.

The problem to be solved by the present invention is for the purpose of solving the above-described conventional problems, data collected through an IoT sensor module and a camera module disposed at a work site or attached to at least a part of a worker's body located in a work site Based on artificial intelligence that predicts risks to the work site or workers by analyzing It relates to a workplace safety management system and method through IoT sensor data and image data analysis of

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

The work site safety management system through the AI-based IoT sensor data and image data analysis according to an embodiment of the present invention for solving the above-described problems is at least the body of a worker placed at the work site or located in the work site An IoT sensor module and a camera module attached to a part, and a safety management server that provides a safety management solution to the worker, wherein the safety management server includes sensor data collected from the IoT sensor module and images collected from the camera module By analyzing at least one of the data, the risk of the work site or the operator may be predicted, and a notification of the predicted risk may be provided to the operator as the safety management solution.

In various embodiments, the safety management server analyzes at least one of the collected sensor data and the collected image data to provide first attribute information about the work site and the worker corresponding to the first attribute information. The second attribute information may be extracted, and a risk to the work site or the worker may be predicted using the extracted first attribute information and the extracted second attribute information.

In various embodiments, the safety management server analyzes the collected sensor data to extract the average temperature of the work site as the first attribute information, and in response to the extracted average temperature, the collected image data Analyze to extract information about the work performed by the worker as the second attribute information, predict the risk to the worker based on the extracted average temperature and information about the extracted work, and the predicted risk Accordingly, a notification for guiding a break or work start to the worker may be provided.

In various embodiments, the safety management server, the information about the extracted job includes the type, intensity, and execution time of the job, based on the type and intensity of the job and the extracted average temperature to the worker Set a reference work time and a reference break time for the worker, provide a notification to guide the worker to a break when the worker's work performance time exceeds the reference work time, and provide a notification to guide the worker to a break As the reference rest time elapses from the time point, a notification for guiding the start of work to the worker may be provided.

In various embodiments, by analyzing the collected sensor data, as the first attribute information, the information about the generation of harmful gas at the work site - The information about the generation of the harmful gas is whether the noxious gas is generated, the type of the noxious gas and concentration - extracted and in response to the extracted harmful gas generation information, by analyzing the collected image data, information regarding the worker's wearing of protective equipment and harmful gas intake amount as the second attribute information information is extracted, and based on the extracted information on the generation of harmful gas, the extracted information about wearing protective equipment, and the amount of harmful gas intake information, the risk to the worker is predicted, and the worker according to the predicted risk Alerts can be provided to inform users to wear protective equipment or to stop work.

In various embodiments, when the work site is a coke production plant including a charging car for supplying coal, the safety management server analyzes the collected image data to obtain operation information of the charging car as the first attribute information. extraction, and in response to the extracted operation information of the charging car, the collected image data is analyzed to extract the distance between the charging car and the operator as the second attribute information, and the extracted operation information of the charging car and the It is possible to predict a risk to the operator based on the extracted distance, and provide a notification for guiding the operation attention of the charging car to the operator according to the predicted risk.

In various embodiments, the safety management server provides a notification for guiding the operation caution of the charging car to the operator when the extracted distance becomes less than the first distance as the charging car moves, but the collected image When it is determined that the operator does not perform a coping operation according to the provided notification until the extracted distance is less than a second distance shorter than the first distance by analyzing data, the operation of the charging car to the operator The worker re-provides a notification to guide attention, and the worker performs a coping action according to the re-provided notification until the extracted distance is less than a third distance shorter than the second distance by analyzing the collected image data If it is determined not to do so, the operation of the charging car may be forcibly stopped.

In various embodiments, the safety management server, when a plurality of workers are located in the work site, using the first attribute information for the work site and the separately extracted second attribute information for the plurality of workers It is possible to provide a customized notification for each of the plurality of workers.

In various embodiments, the safety management server, each of the plurality of workers according to whether the first attribute information for the work site and the individually extracted second attribute information for the plurality of workers satisfy a preset criterion However, individually predicting the risk to the plurality of workers, based on the characteristic information of the plurality of workers, including the gender, age, and physical characteristics of the workers - weights for each of the plurality of workers and by correcting the preset criteria based on the assigned weights, the criteria for each of the plurality of workers may be individually set.

In various embodiments, the safety management server analyzes at least one of the collected sensor data and the collected image data to extract first attribute information about the work site and second attribute information about the worker, Predicting a risk to the job site or the operator according to the correlation between the extracted first attribute information and the extracted second attribute information, and providing a notification to the operator according to the predicted risk, When it is determined that the state of the work site is in a dangerous state based on the attribute information, a notification is provided to the operator regardless of the correlation with the extracted second attribute information, and the operator is based on the second attribute information When it is determined that the state of is a dangerous state, a notification may be provided to the worker irrespective of a correlation with the extracted first attribute information.

In various embodiments, the safety management server analyzes a plurality of sensor data collected for a predetermined period in the past based on the first time point at which the worker starts the work, and the risk for the second time point after the first time point The probability of occurrence may be calculated, and when the calculated probability of occurrence of danger exceeds a preset value, a notification for guiding the interruption of the work after the first time point may be provided to the worker.

In various embodiments, the safety management server, when the risk of the job site or the worker is predicted, the worker's characteristic information - the worker's characteristic information includes the worker's gender, age and physical characteristics - to Based on the determination of a coping method according to the predicted risk, it is possible to provide the worker with a notification including information about the predicted risk and information about the determined coping method.

In various embodiments, the safety management server provides a notification including information on the predicted risk and a coping method according to the predicted risk to the worker when the risk of the job site or the operator is predicted, While the operator does not perform an operation according to the coping method or a predetermined time elapses after providing the notification even though the operator has provided N or more notifications by analyzing the collected sensor data and the collected image data When it is determined that the operator does not perform the operation according to the coping method, the operator may be designated as a person of interest, and a penalty may be given to the operator designated as the person of interest.

In various embodiments, the safety management server, when the risk to any one of the workers in the job site or a plurality of workers in the job site is predicted, information and the prediction about the predicted risk to any one of the workers Provide a notification including a coping method according to the risk, and even if the notification is provided to the one or more workers N times or more, the one operator does not perform the operation according to the coping method, or after providing the notification When it is determined that any one of the workers does not perform the operation according to the coping method during the elapse of time of The notification provided to any one of the workers may be provided.

Work site safety management method through AI-based IoT sensor data and image data analysis according to an embodiment of the present invention for solving the above problems A method performed through a safety management system including an IoT sensor module and a camera module attached to at least a part, and a safety management server that provides a safety management solution to the worker, the method comprising: collecting sensor data from the IoT sensor module; Collecting image data from the camera module and analyzing the collected sensor data and the collected image data to predict the risk of the work site or the worker, and provide the worker with the predicted risk as the safety management solution It may include the step of providing a notification for.

Other specific details of the invention are included in the detailed description and drawings.

According to various embodiments of the present invention, risks to the work site or workers are analyzed by analyzing data collected through an IoT sensor module and a camera module disposed at the work site or attached to at least a part of the body of a worker located in the work site. By predicting and providing a notification to a worker when a risk is predicted, there is an advantage in that it is possible to prevent the worker from being exposed to a dangerous situation such as exposure to harmful gas and being damaged.

Effects of the present invention are not limited to the effects mentioned above, and other effects not mentioned will be clearly understood by those skilled in the art from the following description.

1 is a diagram illustrating a workplace safety management system through AI-based IoT sensor data and image data analysis according to an embodiment of the present invention.
2 and 3 are views exemplarily showing a coke manufacturing plant applicable to various embodiments.
4 is a hardware configuration diagram of a workplace safety management server through AI-based IoT sensor data and image data analysis according to an embodiment of the present invention according to another embodiment of the present invention.
5 is a flowchart of a method for safety management of a workplace through AI-based IoT sensor data and image data analysis according to an embodiment of the present invention according to another embodiment of the present invention.
6 is a view exemplarily illustrating a work site to which a work site safety management system through AI-based IoT sensor data and image data analysis is applied in various embodiments.
7 is a flowchart for explaining a method of preventing the risk of exposure to high temperatures of workers as a safety management solution in various embodiments.
8 is a table for explaining high temperature exposure standards for each operation applicable in various embodiments.
9 is a flowchart for explaining a method of preventing a worker from exposure to harmful gas as a safety management solution in various embodiments.
10 is a diagram illustrating a process of providing a notification to a worker in order to prevent the worker from being exposed to harmful gas in various embodiments.
11 is a flowchart illustrating a method of preventing a collision with a charging car as a safety management solution according to various embodiments of the present disclosure;
12 is a view for explaining a process of preventing a collision with a charging car according to various embodiments of the present disclosure;

Advantages and features of the present invention and methods of achieving them will become apparent with reference to the embodiments described below in detail in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but may be implemented in various different forms, and only these embodiments allow the disclosure of the present invention to be complete, and those of ordinary skill in the art to which the present invention pertains. It is provided to fully inform those skilled in the art of the scope of the present invention, and the present invention is only defined by the scope of the claims.

The terminology used herein is for the purpose of describing the embodiments and is not intended to limit the present invention. In this specification, the singular also includes the plural, unless specifically stated otherwise in the phrase. As used herein, “comprises” and/or “comprising” does not exclude the presence or addition of one or more other components in addition to the stated components. Like reference numerals refer to like elements throughout, and "and/or" includes each and every combination of one or more of the recited elements. Although "first", "second", etc. are used to describe various elements, these elements are not limited by these terms, of course. These terms are only used to distinguish one component from another. Therefore, it goes without saying that the first component mentioned below may be the second component within the spirit of the present invention.

Unless otherwise defined, all terms (including technical and scientific terms) used herein will have the meaning commonly understood by those of ordinary skill in the art to which this invention belongs. In addition, terms defined in a commonly used dictionary are not to be interpreted ideally or excessively unless clearly specifically defined.

As used herein, the term “unit” or “module” refers to a hardware component such as software, FPGA, or ASIC, and “unit” or “module” performs certain roles. However, “part” or “module” is not meant to be limited to software or hardware. A “part” or “module” may be configured to reside on an addressable storage medium or may be configured to reproduce one or more processors. Thus, by way of example, “part” or “module” refers to components such as software components, object-oriented software components, class components and task components, processes, functions, properties, Includes procedures, subroutines, segments of program code, drivers, firmware, microcode, circuitry, data, databases, data structures, tables, arrays, and variables. Components and functionality provided within “parts” or “modules” may be combined into a smaller number of components and “parts” or “modules” or additional components and “parts” or “modules”. can be further separated.

Spatially relative terms "below", "beneath", "lower", "above", "upper", etc. It can be used to easily describe the correlation between a component and other components. A spatially relative term should be understood as a term that includes different directions of components during use or operation in addition to the directions shown in the drawings. For example, when a component shown in the drawing is turned over, a component described as “beneath” or “beneath” of another component may be placed “above” of the other component. can Accordingly, the exemplary term “below” may include both directions below and above. Components may also be oriented in other orientations, and thus spatially relative terms may be interpreted according to orientation.

In this specification, a computer means all types of hardware devices including at least one processor, and may be understood as encompassing software configurations operating in the corresponding hardware device according to embodiments. For example, a computer may be understood to include, but is not limited to, smart phones, tablet PCs, desktops, notebooks, and user clients and applications running on each device.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

Each step described in this specification is described as being performed by a computer, but the subject of each step is not limited thereto, and at least a portion of each step may be performed in different devices according to embodiments.

1 is a diagram illustrating a workplace safety management system through AI-based IoT sensor data and image data analysis according to an embodiment of the present invention.

Referring to FIG. 1 , the workplace safety management system through AI-based IoT sensor data and image data analysis according to an embodiment of the present invention includes a safety management server 100, a camera module 210, and an IoT sensor module ( 220 ), the operator terminal 300 , an external server 400 , and a network 500 .

Here, the workplace safety management system through the AI-based IoT sensor data and image data analysis shown in FIG. 1 is according to an embodiment, and its components are not limited to the embodiment shown in FIG. 1 , It may be added, changed, or deleted as necessary.

In one embodiment, the safety management server 100 analyzes various data collected at the work site to predict the risk to the work site or workers located at the work site, and provides safety management to the workers according to the risk prediction result. can provide a solution.

In various embodiments, the safety management server 100 analyzes at least one data of sensor data collected through the IoT sensor module 210 and image data collected through the camera module 220, It is possible to predict the risks to the workers and provide a safety management solution to the workers according to the risk prediction results.

Here, the work site may be a coke manufacturing plant that manufactures coke, and the worker may mean a person who performs a coke manufacturing operation in the coke manufacturing plant. For example, as shown in FIGS. 2 and 3 in a coke manufacturing plant, which is a work site applicable to various embodiments of the present invention, the coke oven 10 of the coke plant has an upper surface in which coal is charged. ) is installed, and a plurality of doors 22 having fire bricks on the inner surface are provided on the outer surface, while the charging car 30 running along the rail 31 in the upper part of the oven 10 has a charging port. A plurality of feeder cones 32 for charging coal through (21) may be provided.

And, the coal in the colbin 40 is supplied into the hopper 50 of the charging car 30 when the gate 42 is opened, and the coal in the hopper 50 is a hollow feeder pipe 54 when the bunker gate 52 is opened. After being supplied to the feeder cone 32 with the lower end in close contact with each of the charging ports 21 through the , it is supplied into the oven 10 to perform the coal drying operation.

Accordingly, the carbonized coke in the oven 10 is transferred from the extruder side to the opposite side transfer car (not shown) side by an extruder (not shown) with both doors 22 open at the end of the carbonization operation of the coal charged in the oven 10 . It can be extruded and discharged. However, the present invention is not limited thereto.

In addition, here, the safety management solution may mean to guide the worker with information on the risk expected to occur at the job site or the worker, as well as a coping method for escaping the risk, but is not limited thereto.

In various embodiments, the safety management server 100 may be connected to the operator terminal 300 through the network 500, and when the occurrence of a risk to the work site or the operator is predicted, to the operator terminal 300 of the operator. Notifications can be provided.

Here, the operator terminal 300 is a wireless communication device that ensures portability and mobility, navigation, PCS (Personal Communication System), GSM (Global System for Mobile communications), PDC (Personal Digital Cellular), PHS (Personal Handyphone System) ), PDA (Personal Digital Assistant), IMT (International Mobile Telecommunication)-2000, CDMA (Code Division Multiple Access)-2000, W-CDMA (W-Code Division Multiple Access), Wibro (Wireless Broadband Internet) terminal, smartphone (Smartphone), smart pad (Smartpad), tablet PC (Tablet PC), such as all types of handheld (Handheld) based wireless communication device may include, but is not limited thereto.

Also, here, the network 500 may refer to a connection structure capable of exchanging information with each other, such as a plurality of terminals and servers. For example, the network 500 includes a local area network (LAN), a wide area network (WAN), the Internet (WWW), a wired/wireless data communication network, a telephone network, a wired/wireless television communication network, and the like. can do.

In addition, here, the wireless data communication network is 3G, 4G, 5G, 3rd Generation Partnership Project (3GPP), 5th Generation Partnership Project (5GPP), Long Term Evolution (LTE), World Interoperability for Microwave Access (WIMAX), Wi-Fi (Wi- Fi), Internet, LAN (Local Area Network), Wireless LAN (Wireless Local Area Network), WAN (Wide Area Network), PAN (Personal Area Network), RF (Radio Frequency), Bluetooth network, It may include, but is not limited to, a Near-Field Communication (NFC) network, a satellite broadcasting network, an analog broadcasting network, a Digital Multimedia Broadcasting (DMB) network, and the like.

In an embodiment, the IoT sensor module 210 may be disposed at a work site or may be attached to at least a portion of a body of a worker located in the work site, and may measure sensor data for the work site or the worker.

In various embodiments, the IoT sensor module 210 may be disposed at a work site, and may include a sensor (eg, a temperature measuring sensor, a noxious gas detecting sensor, etc.) for measuring information about the work environment. In addition, the IoT sensor module 210 is implemented in the form of a wearable device (eg, smart watch) that is attached to at least a part of protective clothing and protective equipment worn by the worker, or that the worker can directly wear, information about the worker It may include, but is not limited to, a sensor for measuring the temperature (eg, a temperature sensor inside or outside the protective suit, a sensor for detecting an inhalation amount of harmful gas coupled with a mask, a sensor for measuring biometric information (eg, heart rate)).

In one embodiment, the camera module 220 may be disposed at a work site (eg, a structure disposed/installed in the work site, the charging car 30, etc.) or attached to at least a part of the body of a worker located in the work site. And, it is possible to generate image data about the job site or the operator.

In various embodiments, the camera module 220 may include a camera (eg, CCTV) that is disposed at the work site and generates image data by photographing the work site and the worker. In addition, the camera module 220 is attached to at least a portion of the protective clothing and protective equipment worn by the worker, or is implemented in the form of a wearable device that the worker can directly wear, and a camera (eg: body cam), but is not limited thereto.

In one embodiment, the external server 400 may be connected to the safety management server 100 through the network 500, and the safety management server 100 is a job site through the AI-based IoT sensor data and image data analysis. Various information and data (eg, IoT sensor data and image data) necessary to provide a safety management solution can be provided.

In various embodiments, the external server 400 is connected to the IoT sensor module 210 and transmits the sensor data collected from the IoT sensor module 210 to the safety management server 100. The IoT sensor data transmission server 410 and It may include, but is not limited to, an image transmission media server 420 that is connected to the camera module 220 and transmits the image data collected from the camera module 220 to the safety management server 100 . Hereinafter, with reference to FIG. 4 , a hardware configuration of the safety management server 100 that performs a work site safety management method through AI-based IoT sensor data and image data analysis will be described.

4 is a hardware configuration diagram of a workplace safety management server through AI-based IoT sensor data and image data analysis according to an embodiment of the present invention according to another embodiment of the present invention.

Referring to FIG. 4 , a workplace safety management server 100 (hereinafter, “safety management server 100”) through AI-based IoT sensor data and image data analysis according to another embodiment of the present invention is one or more The processor 110, the memory 120 for loading the computer program 151 executed by the processor 110, the bus 130, the communication interface 140, and the storage for storing the computer program 151 ( 150) may be included. Here, only the components related to the embodiment of the present invention are shown in FIG. 4 . Accordingly, those skilled in the art to which the present invention pertains can see that other general-purpose components other than the components shown in FIG. 4 may be further included.

The processor 110 controls the overall operation of each component of the safety management server 100 . The processor 110 includes a central processing unit (CPU), a micro processor unit (MPU), a micro controller unit (MCU), a graphic processing unit (GPU), or any type of processor well known in the art. can be

In addition, the processor 110 may perform an operation for at least one application or program for executing the method according to the embodiments of the present invention, and the safety management server 100 may include one or more processors. .

In various embodiments, the processor 110 temporarily and/or permanently stores a signal (or data) processed inside the processor 110 . Random Access Memory (RAM) and ROM (Read -Only Memory, not shown) may be further included. In addition, the processor 110 may be implemented in the form of a system on chip (SoC) including at least one of a graphic processing unit, a RAM, and a ROM.

The memory 120 stores various data, commands and/or information. The memory 120 may load the computer program 151 from the storage 150 to execute methods/operations according to various embodiments of the present disclosure. When the computer program 151 is loaded into the memory 120 , the processor 110 may perform the method/operation by executing one or more instructions constituting the computer program 151 . The memory 120 may be implemented as a volatile memory such as RAM, but the technical scope of the present disclosure is not limited thereto.

The bus 130 provides a communication function between the components of the safety management server 100 . The bus 130 may be implemented as various types of buses, such as an address bus, a data bus, and a control bus.

The communication interface 140 supports wired and wireless Internet communication of the safety management server 100 . Also, the communication interface 140 may support various communication methods other than Internet communication. To this end, the communication interface 140 may be configured to include a communication module well known in the technical field of the present invention. In some embodiments, the communication interface 140 may be omitted.

The storage 150 may non-temporarily store the computer program 151 . When performing the process of providing a job site safety management solution through AI-based IoT sensor data and image data analysis through the safety management server 100, the storage 150 performs AI-based IoT sensor data and image data analysis. It is possible to store various information necessary to provide a process of providing a work site safety management solution through

The storage 150 is a non-volatile memory such as a read only memory (ROM), an erasable programmable ROM (EPROM), an electrically erasable programmable ROM (EEPROM), a flash memory, a hard disk, a removable disk, or well in the art to which the present invention pertains. It may be configured to include any known computer-readable recording medium.

The computer program 151 may include one or more instructions that, when loaded into the memory 120 , cause the processor 110 to perform methods/operations according to various embodiments of the present invention. That is, the processor 110 may perform the method/operation according to various embodiments of the present invention by executing the one or more instructions.

In one embodiment, the computer program 151 collects sensor data from the IoT sensor module, collects image data from the camera module, and analyzes the collected sensor data and the collected image data to risk the work site or workers. Including one or more instructions to perform a worksite safety management method through artificial intelligence-based IoT sensor data and image data analysis, including the step of predicting can do.

The steps of the method or algorithm described in relation to the embodiment of the present invention may be implemented directly in hardware, implemented as a software module executed by hardware, or implemented by a combination thereof. A software module may include random access memory (RAM), read only memory (ROM), erasable programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), flash memory, hard disk, removable disk, CD-ROM, or It may reside in any type of computer-readable recording medium well known in the art to which the present invention pertains.

The components of the present invention may be implemented as a program (or application) to be executed in combination with a computer, which is hardware, and stored in a medium. Components of the present invention may be implemented as software programming or software components, and similarly, embodiments may include various algorithms implemented as data structures, processes, routines, or combinations of other programming constructs, including C, C++ , Java, assembler, etc. may be implemented in a programming or scripting language. Functional aspects may be implemented in an algorithm running on one or more processors. Hereinafter, with reference to FIGS. 5 and 6 , a work site safety management method through artificial intelligence-based IoT sensor data and image data analysis performed by the safety management server 100 will be described.

5 is a flowchart of a worksite safety management method through AI-based IoT sensor data and image data analysis according to an embodiment of the present invention according to another embodiment of the present invention, and FIG. It is a diagram showing an example of a work site to which a work site safety management system through intelligence-based IoT sensor data and image data analysis is applied.

Referring to FIGS. 5 and 6 , in step S110 , the safety management server 100 may collect sensor data and image data for a work site and a worker.

In various embodiments, the safety management server 100 includes various sensor data measured through the IoT sensor module 210 (eg, the temperature of the work site, harmful gases, the worker's body temperature, the amount of inhaled harmful gases, and biometric information (eg, heart rate). ), etc.) can be collected.

In addition, the safety management server 100 may collect various image data generated through the camera module 220 (eg, image data of a work site and a worker). However, the present invention is not limited thereto.

In various embodiments, the safety management server 100 collects sensor data and image data every preset period, but sensor data and image data based on the first attribute information and the second attribute information extracted through step S120 to be described later. collection cycle can be adjusted. For example, the safety management server 100 collects sensor data in a first cycle as work is started at the work site, but when it is determined that harmful gas is generated by analyzing the sensor data through step S120 to be described later, The sensor data collection period may be set to a second period shorter than the first period. However, the present invention is not limited thereto.

At this time, the worker terminal 300 of the worker may be directly connected to the IoT sensor modules 210 located in the work site based on short-range wireless communication as shown in FIG. 6 , and not through the safety management server 100 . By directly receiving the sensor data collected from the IoT sensor modules 210 without the use of the IoT sensor module 210, it may be implemented so that the operator can recognize information about the work site in real time, but is not limited thereto.

In step S120 , the safety management server 100 may analyze the sensor data and image data collected through step S110 to extract attribute information about the work site and the worker.

In various embodiments, the safety management server 100 may extract the first attribute information about the work site and the second attribute information about the worker by analyzing the sensor data and the image data using the pre-learned artificial intelligence model. have.

Here, the pre-trained artificial intelligence model may be a model learned using learning data generated by labeling attribute information on the work site and attribute information on the work site on sensor data and image data. At this time, the pre-learned AI model needs to extract attribute information from different types of input data (eg, sensor data in numerical form and image data in image form). It may include, but is not limited to, a first model for extracting information and a second model for extracting attribute information from image data.

An artificial intelligence model (e.g., a neural network) consists of one or more network functions, which may consist of a set of interconnected computational units, which can generally be referred to as ‘nodes’. These 'nodes' may also be referred to as 'neurons'. The one or more network functions are configured by including at least one or more nodes. Nodes (or neurons) constituting one or more network functions may be interconnected by one or more ‘links’.

In the AI model, one or more nodes connected through a link may relatively form a relationship between an input node and an output node. The concepts of an input node and an output node are relative, and any node in an output node relationship with respect to one node may be in an input node relationship in a relationship with another node, and vice versa. As described above, an input node to output node relationship may be created around a link. One or more output nodes may be connected to one input node through a link, and vice versa.

In the relationship between the input node and the output node connected through one link, the value of the output node may be determined based on data input to the input node. Here, a node interconnecting the input node and the output node may have a weight. The weight may be variable, and may be changed by a user or an algorithm in order for the AI model to perform a desired function. For example, when one or more input nodes are interconnected to one output node by respective links, the output node sets values input to input nodes connected to the output node and links corresponding to the respective input nodes. An output node value may be determined based on the weight.

As described above, in the AI model, one or more nodes are interconnected through one or more links to form an input node and an output node relationship within the AI model. The characteristics of the AI model may be determined according to the number of nodes and links in the AI model, the correlation between nodes and links, and the value of a weight assigned to each of the links. For example, when the same number of nodes and links exist and there are two AI models having different weight values between the links, the two AI models may be recognized as different from each other.

Some of the nodes constituting the artificial intelligence model may configure one layer based on distances from the initial input node. For example, a set of nodes having a distance n from the initial input node may constitute n layers. The distance from the initial input node may be defined by the minimum number of links that must be traversed to reach the corresponding node from the initial input node. However, the definition of such a layer is arbitrary for description, and the order of the layer in the AI model may be defined in a different way than the above. For example, a layer of nodes may be defined by a distance from the final output node.

The initial input node may mean one or more nodes to which data is directly input without going through a link in a relationship with other nodes among nodes in the AI model. Alternatively, in the relationship between nodes based on the link in the artificial intelligence model network, it may mean nodes that do not have other input nodes connected by a link. Similarly, the final output node may refer to one or more nodes that do not have an output node in relation to other nodes among nodes in the AI model. In addition, the hidden node may refer to nodes constituting an artificial intelligence model other than the first input node and the last output node. The artificial intelligence model according to an embodiment of the present invention may have more nodes in the input layer than nodes in the hidden layer close to the output layer, and is an artificial intelligence model in which the number of nodes decreases as the input layer progresses to the hidden layer. can

An AI model may include one or more hidden layers. The hidden node of the hidden layer may use the output of the previous layer and the output of the neighboring hidden nodes as inputs. The number of hidden nodes for each hidden layer may be the same or different. The number of nodes of the input layer may be determined based on the number of data fields of the input data and may be the same as or different from the number of hidden nodes. Input data input to the input layer may be calculated by a hidden node of the hidden layer and may be output by a fully connected layer (FCL) that is an output layer.

In various embodiments, the artificial intelligence model may be a deep learning model.

A deep learning model (eg, deep neural network (DNN)) may refer to an artificial intelligence model including a plurality of hidden layers in addition to an input layer and an output layer. Be able to identify latent structures, i.e., the latent structures of photos, texts, videos, voices, and music (e.g., what objects are in the photos, what the content and emotions of the text are, and the content and emotions, etc.).

Deep neural networks include convolutional neural networks (CNNs), recurrent neural networks (RNNs), auto encoders, generative adversarial networks (GANs), and restricted Boltzmann machines (RBMs). boltzmann machine), a deep belief network (DBN), a Q network, a U network, a Siamese network, and the like, but is not limited thereto.

In various embodiments, the network function may include an auto-encoder. Here, the auto-encoder may be a type of artificial neural network for outputting output data similar to input data.

The auto encoder may include at least one hidden layer, and an odd number of hidden layers may be disposed between the input/output layers. The number of nodes in each layer may be reduced from the number of nodes in the input layer to an intermediate layer called the bottleneck layer (encoding), and then expanded symmetrically with reduction from the bottleneck layer to the output layer (symmetrical to the input layer). The nodes of the dimensionality reduction layer and the dimension restoration layer may or may not be symmetrical. In addition, the auto-encoder can perform non-linear dimensionality reduction. The number of input layers and output layers may correspond to the number of sensors remaining after preprocessing of input data. In the auto-encoder structure, the number of nodes of the hidden layer included in the encoder may have a structure that decreases as the distance from the input layer increases. If the number of nodes in the bottleneck layer (the layer with the fewest nodes between the encoder and decoder) is too small, a sufficient amount of information may not be conveyed, so a certain number or more (e.g., more than half of the input layer, etc.) ) may be maintained.

The neural network may be trained by at least one of supervised learning, unsupervised learning, and semi-supervised learning. The training of the neural network is to minimize the error in the output. More specifically, the training of the neural network repeatedly inputs the training data into the neural network, calculates the output of the neural network for the training data and the error of the target, and outputs the error of the neural network in a direction to reduce the error. It is a process of updating the weight of each node of the neural network by backpropagating from the layer to the input layer.

First, in the case of teacher learning, learning data in which correct answers are labeled in each learning data is used (ie, labeled learning data), and in the case of comparative learning, the correct answers may not be labeled in each learning data. That is, for example, the learning data in the case of teacher learning regarding data classification may be data in which categories are labeled in each of the learning data. Labeled training data is input to the neural network, and an error can be calculated by comparing the output (category) of the neural network with the label of the training data.

Next, in the case of comparison learning on data classification, an error may be calculated by comparing the input training data with the neural network output. The calculated error is back propagated in the reverse direction (ie, from the output layer to the input layer) in the neural network, and the connection weight of each node of each layer of the neural network may be updated according to the back propagation. A change amount of the connection weight of each node to be updated may be determined according to a learning rate. The computation of the neural network on the input data and the backpropagation of errors can constitute a learning cycle (epoch). The learning rate may be applied differently depending on the number of repetitions of the learning cycle of the neural network. For example, in the early stage of learning of a neural network, a high learning rate can be used to enable the neural network to quickly acquire a certain level of performance, thereby increasing efficiency, and using a low learning rate at the end of learning can increase accuracy.

In training of a neural network, in general, the training data may be a subset of real data (that is, data to be processed using the trained neural network), and thus the error on the training data is reduced, but the error on the real data is reduced. There may be increasing learning cycles. Overfitting is a phenomenon in which errors on actual data increase by over-learning on training data as described above. For example, a phenomenon in which a neural network that has learned a cat by seeing a yellow cat does not recognize that it is a cat when it sees a cat other than yellow may be a type of overfitting. Overfitting can act as a cause of increasing errors in machine learning algorithms. In order to prevent such overfitting, various optimization methods can be used. In order to prevent overfitting, methods such as increasing training data, regularization, or dropout in which a part of nodes of the network are omitted in the process of learning, may be applied.

In various embodiments, the safety management server 100 may train the artificial intelligence model using a plurality of first learning data (eg, a plurality of images each annotated with abnormal cells, normal cells, and background regions).

More specifically, the safety management server 100 inputs each of the learning input data sets to one or more network functions, and a learning output data set corresponding to each of the output data calculated by the one or more network functions and the label of each of the learning input data sets. An error can be derived by comparing each of them. That is, in the training of the artificial intelligence model, learning input data may be input to an input layer of one or more network functions, and the training output data may be compared with outputs of one or more network functions. The safety management server 100 may train the neural network based on an error between an operation result of one or more network functions on the learning input data and an error of the learning output data (label).

In addition, the safety management server 100 may adjust the weight of one or more network functions in a backpropagation method based on the error. That is, the safety management server 100 may adjust the weight so that the output of the one or more network functions approaches the learning output data based on the error between the operation result of the one or more network functions on the learning input data and the learning output data.

When the learning of one or more network functions is performed over a predetermined epoch, the safety management server 100 may determine whether to stop the learning by using the verification data. The predetermined epoch may be a part of the overall learning objective epoch. The validation data may consist of at least a portion of the labeled training data set. That is, the safety management server 100 performs learning of the neural network through the learning data set, and after the learning of the artificial intelligence model is repeated more than a predetermined epoch, the learning effect of the artificial intelligence model is a predetermined level using the verification data. It can be determined whether it is abnormal or not. For example, when the safety management server 100 performs learning with a target repetition learning number of 10 times using 100 pieces of learning data, after performing repeated learning 10 times, which is a predetermined epoch, 10 verification data is used. Thus, by performing repeated learning three times, if the change in the output of the AI model during the three times of repeated learning is below a predetermined level, it is determined that further learning is meaningless and learning can be terminated. That is, the verification data may be used to determine the completion of learning based on whether the effect of learning for each epoch is greater than or equal to a certain level in the iterative learning of the artificial intelligence model. The above-described number of training data, verification data, and number of repetitions are merely examples and are not limited thereto.

The safety management server 100 may generate a sleep evaluation model by testing the performance of one or more network functions using the test data set and determining whether to activate one or more network functions. The test data may be used to verify the performance of the artificial intelligence model, and may be composed of at least a part of the training data set. For example, 80% of the training dataset can be utilized for training the AI model (i.e., learning to adjust the weights to output results similar to labels), and 20% can be used to improve the performance of the AI model. It can be used as test data for verification.

The safety management server 100 may determine whether to activate the artificial intelligence model according to whether it is above a predetermined performance by inputting a test data set to the learned artificial intelligence model and measuring the error.

The safety management server 100 verifies the performance of the trained artificial intelligence model using test data on the learned artificial intelligence model, and when the performance of the trained artificial intelligence model is higher than a predetermined standard, the AI model is transferred to another application. You can enable it for use. In addition, the safety management server 100 may inactivate and discard the artificial intelligence model when the performance of the learned artificial intelligence model is less than or equal to a predetermined criterion. For example, the safety management server 100 may determine the performance of the generated artificial intelligence model based on factors such as accuracy, precision, and recall. The above-described performance evaluation criteria are merely examples, and the present invention is not limited thereto.

In various embodiments, the safety management server 100 may extract attribute information from sensor data and image data based on a preset correlation relationship. At least one of the sensor data and the image data may be analyzed to extract first attribute information about the job site, and based on a preset correlation, second attribute information about the worker corresponding to the first attribute information may be extracted. can

Here, the preset correlation may mean that a combination of attribute information to be considered together in order to predict a specific risk is defined in advance.

As an example, when the safety management server 100 extracts the average temperature of the work site as first attribute information as the sensor data is analyzed, in order to predict the "high temperature exposure risk" of the worker according to a preset correlation, the work site As the second attribute information corresponding to the average temperature of , information about the work performed by the worker (eg, the type of work, intensity, and execution time, etc.) may be extracted.

As another example, when the safety management server 100 extracts information on harmful gas generation (eg, whether harmful gas is generated, type and concentration of noxious gas, etc.) as the first attribute information as the sensor data is analyzed, the preset In order to predict the "risk of exposure to harmful gas" of workers according to the relationship, information about the worker's wearing of protective equipment (e.g., whether or not protective equipment is worn and the protection worn type of equipment, etc.) and harmful gas intake amount information (eg, based on the harmful gas intake amount detection sensor provided in the mask) can be extracted.

As another example, when the safety management server 100 extracts operation information of the charging car 30 disposed in the work site as the first attribute information as the image data is analyzed, the operator's "charging car In order to predict the “risk of collision with 30 ,” the distance between the charging car 30 and the operator may be extracted as second attribute information corresponding to the operation information of the charging car 30 .

Here, the safety management server 100 analyzes sensor data and image data to first extract first attribute information about the work site, and then extracts second attribute information corresponding to the first attribute information according to a preset correlation relationship. However, the present invention is not limited thereto, and in some cases, the second attribute information is first extracted and then the first attribute information corresponding to the second attribute information is extracted, or after extracting arbitrary attribute information, the extraction is performed. It may be implemented in the form of extracting other attribute information corresponding to the specified attribute information.

In various embodiments, the safety management server 100 may extract all extractable attribute information (eg, first attribute information and second attribute information) by analyzing sensor data and image data, and Accordingly, the extracted attribute information can be classified.

More specifically, the safety management server 100 analyzes the sensor data and the image data, respectively, and analyzes the average temperature of the work site, information on harmful gas generation, operation information of the charging car 30, and information on the work performed by the operator. , it is possible to extract information on the worker's wearing protective equipment, harmful gas intake information, and the distance between the charging car 30 and the worker.

Thereafter, the safety management server 100 classifies "information about the average temperature of the work site and the work performed by the worker" into one category in order to predict the risk of high temperature exposure according to a preset relationship, and the risk of exposure to harmful gases In order to predict , "information on the occurrence of harmful gases, information on the wearing of protective equipment and information on the amount of inhalation of harmful gases" are classified into one category, and in order to predict the risk of collision with the charging car (30), the "charging car ( 30) operation information and the distance between the charging car 30 and the operator" may be classified into one category. In this case, when one piece of attribute information is used for predicting multiple risks, one piece of attribute information may be simultaneously included in a plurality of categories, but is not limited thereto.

In step S130, the safety management server 100 uses the first attribute information and the second attribute information extracted through the step S120 to predict the risk of the job site or the worker, and provides a safety management solution to the worker according to the risk prediction result. can provide

Here, the safety management solution provides information and response to the predicted risk so that the worker can recognize what the predicted risk is at the job site or to himself and how to deal with it as the occurrence of a risk to the job site or worker is predicted. It may be to provide a notification including a method, but is not limited thereto.

In various embodiments, the safety management server 100 includes first attribute information and second attribute information extracted according to a preset relationship, or first attribute information and second attribute information matched with each other according to a preset relationship. It is possible to predict a risk to the operator depending on whether the set criteria are satisfied, and when a risk is predicted to occur, a notification may be provided to the operator. Hereinafter, with reference to FIGS. 7 to 12, it will be described in more detail.

7 is a flowchart for explaining a method of preventing the risk of exposure to high temperatures of workers as a safety management solution in various embodiments.

Referring to Figure 7, the safety management server 100 is the average temperature of the work site extracted as the first attribute information and the information about the work extracted as the second attribute information (eg, the type, intensity, and execution time of the task) A safety management solution can be provided to workers based on the relationship.

In step S210, the safety management server 100 may extract information about the average temperature and work of the work site.

More specifically, the safety management server 100 may measure the temperature for the work site for a predetermined period through the IoT sensor module 210 (eg, a temperature sensor) provided in at least a part of the work site or the body of the worker, , by calculating the average temperature for the work site for a predetermined period using the measured temperature, the average temperature of the work site may be extracted as the first attribute information.

In addition, the safety management server 100 can extract information about the work site and the operation pattern of the worker by analyzing the image data captured through the camera module 220 provided on the work site or at least a part of the worker's body, , information on the job may be extracted as the second attribute information based on the extracted information on the job site and the operation pattern of the worker. Here, the information about the task may include the type, intensity, and execution time of the task. However, the present invention is not limited thereto, and the safety management server 100 determines the type and intensity of a task performed at the current time based on the previously registered task schedule information, and the operator's operation pattern collected through image data analysis. The execution time can be extracted according to

In step S220, the safety management server 100 may set a standard for high temperature exposure in order to prevent the risk of exposure to high temperature of the worker.

In various embodiments, the safety management server 100 may set the reference work time and the reference rest time for the worker based on the type and intensity of the work and the average temperature.

Here, the reference work time and the reference rest time may be determined according to the high temperature exposure standard shown in FIG. 8 , but is not limited thereto.

In addition, here, light work refers to work that requires up to 200 kcal of calories, and refers to work that requires light use of hands or arms to control the machine while sitting or standing, and moderate work requires 200 to 350 kcal per hour. It refers to the work that is done, and it means walking while lifting or pushing an object, it refers to a work that requires 350 to 500 kcal per hour, and may refer to work such as pickaxe or shovel, but is not limited thereto.

In step S230, the safety management server 100 may predict the risk of the worker according to whether the criteria set through step S220 are satisfied, and may provide a notification to guide the worker to rest or start work according to the predicted risk. . For example, the safety management server 100 sets the standard work time and the standard break time to 45 minutes/15 minutes per hour, respectively, when the average temperature of the worker is 28 degrees and the type and intensity of the worker's work is moderate work. can Thereafter, the safety management server 100 may provide a notification to guide the worker to a break when the worker's work execution time exceeds 45 minutes, and 15 minutes elapse from the time when the worker provides a break notification to the worker A notification may be provided to guide the initiation of work.

9 is a flowchart for explaining a method of preventing the risk of exposure to harmful gases by workers as a safety management solution in various embodiments, and FIG. It is a diagram showing the process of providing.

Referring to FIGS. 9 and 10 , in various embodiments, the safety management server 100 includes information on the generation of harmful gas extracted as first attribute information (eg, whether noxious gas is generated, type and concentration of noxious gas), and 2 It is possible to provide a safety management solution to the worker based on the correlation between the information on wearing protective equipment and the information on the amount of inhalation of harmful gas extracted as attribute information.

In step S310, the safety management server 100 may extract information on the generation of harmful gas at the work site, information on the worker's wearing of protective equipment, and information on the amount of harmful gas intake.

More specifically, the safety management server 100 uses the sensor data collected through the IoT sensor module 210 (eg, harmful gas detection sensor) provided on at least a part of the body of the worker or the work site, at the work site. Whether or not harmful gas is generated, the type and concentration of harmful gas can be extracted.

In addition, the safety management server 100 analyzes the image data photographed through the camera module 220 provided on the work site or at least a part of the worker's body, thereby protecting the worker from exposure to harmful gases (eg, a mask). ) can be extracted.

In addition, the safety management server 100 uses the sensor data collected through the IoT sensor module 210 (eg, a harmful gas intake detection sensor provided in a mask) provided on at least a part of the worker's body, causing harmful effects to the worker. Gas intake amount information can be extracted.

In step S320, the safety management server 100 may set standards for exposure to harmful gases in order to prevent the risk of exposure to harmful gases of workers.

Here, the standards for exposure to harmful gases are exposure standards of the Ministry of Labor, and may be determined based on Ministry of Labor Notification No. 2002-8, exposure standards for chemical substances and physical factors, but is not limited thereto.

In step S330, the safety management server 100 can predict the risk of the worker according to whether the criteria set through the step S320 are satisfied, and a notification to guide the worker wearing protective equipment, start work, or stop work according to the predicted risk can provide For example, the safety management server 100 guides the worker to wear a protective device when the generation of harmful gas is detected at the work site, but it is determined that the worker does not wear protective equipment (eg, a mask) for blocking harmful gas. notification can be provided. In addition, the safety management server 100 may determine whether the inhalation amount of the harmful gas of the worker satisfies the standard, and may provide a notification for guiding the start of the work or the stop of the work according to the satisfaction.

11 is a flowchart illustrating a method of preventing a collision with a charging car as a safety management solution in various embodiments, and FIG. 12 is a diagram illustrating a process of preventing a collision with a charging car in various embodiments.

11 and 12, in various embodiments, the safety management server 100 is the operation information of the charging car 30 extracted as the first attribute information and the charging car 30 and the worker extracted as the second attribute information. It is possible to provide a safety management solution to the operator based on the relationship between the distances.

In step S410 , the safety management server 100 may extract operation information of the charging car 30 and the distance d between the charging car 30 and the operator.

More specifically, the safety management server 100 may detect a change in the position of the charging car 30 by analyzing the image data taken through the camera module 220 installed on the work site or at least a part of the worker's body. , through this, operation information including whether or not the charging car 30 operates and an operation speed may be extracted. However, the present invention is not limited thereto, and the safety management server 100 may be connected to a computing device that controls the operation of the charging car 30 to receive operation information of the charging car 30 from the computing device.

In addition, the safety management server 100 analyzes the image data taken through the camera module 220 installed on at least a part of the work site or the worker's body to calculate the distance (d) between the charging car 30 and the worker. can

In various embodiments, the safety management server 100 may calculate the distance (d) between the charging car 30 and the worker based on the structure in which information regarding the size and length is registered in advance among the structures disposed at the work site. have. For example, the safety management server 100 may calculate the distance between the charging car 30 and the operator based on the distance (pre-set value) between the bulkheads provided in the coke manufacturing plant, but is not limited thereto, Any method of calculating the distance between two or more objects is applicable.

In step S420 , the safety management server 100 may set a standard for collision prevention in order to prevent a collision of the operator's charging car 30 .

In step S430, the safety management server 100 can predict the risk of the operator according to whether the criteria set through the step S420 are satisfied, and a notification to guide the operation attention of the charging car 30 to the operator according to the predicted risk can provide

In various embodiments, the safety management server 100 provides an operation of the charging car 30 to the operator when the distance (d) between the charging car 30 and the operator is less than the first distance as the charging car 30 moves. Provide a notification to guide attention, but the operator does not perform a coping action according to the notification until the distance (d) between the charging car 30 and the operator is less than the second distance shorter than the first distance by analyzing the image data If it is determined that not, a notification guiding the operation caution of the charging car 30 to the operator may be provided again.

In addition, the safety management server 100 analyzes the image data and the operator does not perform a coping action according to the notification until the distance (d) between the charging car 30 and the operator becomes less than the third distance shorter than the second distance. When it is determined that not, by forcibly stopping the operation of the charging car 30 , it is possible to prevent an accident due to the moving operation of the charging car 30 .

Again, referring to FIGS. 5 and 6 , the safety management server 100 determines the correlation between the first attribute information for the work site and the second attribute information for the worker extracted by analyzing at least one of sensor data and image data. Predict the risk to the job site or worker according to the forecast, and provide a notification to the worker according to the predicted risk, but if it is determined that the status of the job site or worker is in a dangerous state only with the first attribute information or the first attribute information, Notifications can be provided to workers regardless of relationship.

As an example, the safety management server 100 extracts information on the generation of harmful gases by analyzing the sensor data, but as the concentration of the harmful gas detected at the work site exceeds a preset reference value, it is a dangerous situation to proceed with the work. If it is determined that the worker is wearing a mask or not, regardless of the worker's inhalation amount of harmful gas, it is possible to provide a notification instructing the worker to stop the work and leave the job site.

As another example, when it is determined that the safety management server 100 extracts information about the worker's wearing of protective equipment by analyzing the image data, but the worker does not wear the protective equipment that is essential to perform the work In other words, it is possible to provide a notification instructing workers to wear protective equipment regardless of the attribute information of the job site.

In various embodiments, when a plurality of workers are located in one work site, the safety management server 100 may individually extract second attribute information for each of the plurality of workers, A customized notification may be provided to each of the plurality of workers by using the first attribute information and each of the plurality of individually extracted second attribute information. For example, the safety management server 100 is a risk to each of a plurality of workers according to whether the first attribute information for the work site and the second attribute information individually extracted for the plurality of workers satisfy a preset criterion. can be individually predicted, and customized notifications for each of a plurality of workers can be provided according to the individually predicted risk.

At this time, the preset standard is a standard for judging the risk to the work site or the worker, and may be individually set for each of a plurality of workers according to the worker's characteristic information (eg, the worker's gender, age, and physical characteristics, etc.) , but not limited thereto. For example, when the charging car 30 needs to leave the job site due to a moving operation or detection of harmful gas, typically a younger worker can leave the job site relatively faster than an older worker. In other words, the standard is set in consideration of the age of the worker, that is, the harmful gas intake standard for the older worker is set lower than that of the younger worker, and the distance standard for the older worker (e.g., a charging car (30) distance) can be set higher than the distance standard for younger workers.

In various embodiments, the computing device 100 sets a risk determination criterion for determining the risk of the operator, and determines the risk to the operator according to the set risk determination criterion, sets a weight based on the characteristic information of the operator, and , by correcting the risk determination criteria using the set weight, it is possible to determine the risk to the operator in consideration of the characteristic information of the operator. For example, when the criterion for determining the risk of collision with the charging car 30 is 100 m, the computing device 100 gives an older worker a higher weight than a younger worker, and through this, by correcting the standard , the distance standard for the older worker (eg, the distance standard with the charging car 30) may be set higher than the distance standard for the younger worker.

In various embodiments, the safety management server 100 predicts a risk to the job site or worker based on the first attribute information for the job site and the second attribute information for the worker, and predicts the risk to the risk when the occurrence of the risk is predicted. A notification including information on information and a method for coping with a risk is provided to the worker, but based on the worker's characteristic information (eg, gender, age, and physical characteristics, etc.), it is possible to determine a coping method according to the risk. For example, the safety management server 100 may provide a notification to guide older workers to take a longer break than younger workers as a method of coping with the risk of exposure to high temperatures. In addition, the safety management server 100 provides a specific user with a coping method according to the risk, but may provide a notification including a more detailed coping method to a worker with less experience than a worker with a lot of work experience, but is not limited thereto does not

In various embodiments, the safety management server 100 analyzes (eg, big data analysis) a plurality of sensor data collected for a predetermined period in the past based on a first time point at which the worker starts work, The risk occurrence probability for the second time point may be calculated, and when the calculated risk occurrence probability exceeds a preset value, a notification for guiding the operator to stop the work after the first time point may be provided.

More specifically, the safety management server 100 analyzes a plurality of sensor data (eg, sensor data collected on the previous day) collected during a predetermined period in the past using a pre-learned risk occurrence probability calculation model to generate the next day's risk Probability can be calculated, and when the probability that the job site or worker will be exposed to a dangerous situation is high as the calculated probability of occurrence of danger exceeds a preset value, a notification to inform the worker of cessation of work after the first time point can provide

At this time, the pre-learned risk occurrence probability calculation model uses the pattern extracted by analyzing the sensor data of the day of the accident and the patterns extracted by analyzing the sensor data collected for a predetermined period in the past based on the day of the accident as the learning data. It may be a model trained by For example, the pre-learned risk occurrence probability calculation model may include a plurality of models having different structures while being learned using the same learning data, and may include a plurality of sensor data collected during a predetermined period in the past or a predetermined amount of data in the past. By analyzing the plurality of sensor data collected during the period of , the extracted pattern is input to each of the plurality of models, and the extracted result value is soft-voted to calculate the risk occurrence probability of the next day.

In various embodiments, the safety management server 100 provides a notification including information on the predicted risk and a method of coping according to the predicted risk to the operator when the risk of the job site or the operator is predicted, but sensor data and If it is determined that the worker does not perform the action according to the coping method by analyzing the collected image data, a notification including information on the predicted risk and the coping method according to the predicted risk can be provided again to the worker have.

At this time, the safety management server 100 does not perform an operation according to the coping method even though the notification of the same content is repeatedly provided to a specific worker N times or more, or a predetermined time elapses after the notification is provided. If it is determined that the operator does not perform an operation according to the coping method during the period, a penalty may be given by designating the worker as a person of interest.

Here, the penalty is to allocate a number of missions (for example, mandatory completion of various training related to safety management, etc.) so that the worker can perform safer work, and to allocate work/salary, etc. according to the result of the mission. It may include, but is not limited to, giving a disadvantage or providing information to the manager of the work site by converting it into data so that it can be reflected in the personnel evaluation.

On the other hand, the safety management server 100 may set a model person based on the operation histories of each of a plurality of workers for a predetermined period (eg, whether an operation is performed according to the coping method, etc.), and may benefit from the set model person. ) can be provided.

In various embodiments, the safety management server 100 selectively provides a notification to only one operator when the risk to any one of the workers in the workplace or a plurality of workers in the workplace is predicted, but to any one operator If it is determined that any one operator does not perform an operation according to the coping method while providing N or more notifications or a predetermined time has elapsed after providing the notification, the image data is analyzed to determine whether any one of the plurality of workers It is possible to provide notifications provided to any one operator to other workers adjacent to the operator.

In various embodiments, the safety management server 100 provides information on each of two or more risks and a method for coping with each of the two or more risks, when two or more different types of risks are predicted for one job site or one worker. It is possible to provide a notification that includes, but to minimize the number of guided countermeasures. For example, if three types of risks are predicted for one worker, the safety management server 100 determines one coping method that can overcome all three types of predicted risks, and sets the determined coping method to the worker. can guide you.

At this time, the safety management server 100 provides a notification including two or more coping methods so as to solve all the risks when it is determined that only one coping method cannot solve all the risks, but the most of the three types of risks. It is possible to give priority to countermeasures that respond to significant risks (eg, the most damaging risks).

Here, the workplace safety management method through the AI-based IoT sensor data and image data analysis according to FIGS. 5 to 12 is the sensor data collected through the IoT sensor module 210 and the sensor data collected through the camera module 220. The image data is provided to the safety management server 100, and the safety management server 100 predicts a risk to the worker and provides a notification for this, but is not limited thereto, and the safety management server 100 Without going through, it is possible to perform a risk prediction and notification providing operation by the operator through the operator terminal 200 of the operator.

More specifically, the worker terminal 200 may include a processor capable of performing an operation for an application or program provided from the safety management server 100, and an IoT sensor module 210 attached to at least a part of the worker's body. ) and the sensor data and image data collected through the camera module 220 can be analyzed on their own without providing the safety management server 100 to predict the risk, and when the risk is predicted, the operator terminal 200 itself You can control to output notifications. Here, the risk prediction and notification providing operation performed through the processor of the worker terminal 200 is the safety management server 100 in the workplace safety management method through the AI-based IoT sensor data and image data analysis according to FIGS. 5 and 6 . ) may be the same as the operation performed by

In various embodiments, the camera module 220 is attached to at least a portion of the protective clothing and protective equipment worn by the first worker, or is implemented in the form of a wearable device that the first worker can directly wear. When implemented in the form of taking pictures of other second workers, the operator terminal 200 includes the sensor data collected from the IoT sensor module 210 and the image data collected from the camera sensor 220 (eg, at the job site and the job site). By analyzing the image data generated by photographing the second workers located), it is possible to predict the occurrence of a risk to the second workers, and when the occurrence of a risk to a specific second worker is predicted, the worker terminal 200 of the specific second worker ) and may provide a notification to a specific second worker terminal 200. At this time, the result of risk determination and notification provided by the worker terminal 200 itself is displayed in real time or at a preset period by the safety management server 100 ), so that the history of risk determination and notification to the job site or worker is stored and managed through the safety management server 100 .

The above-described artificial intelligence-based IoT sensor data and work site safety management method through image data analysis has been described with reference to the flowchart shown in the drawings. For simple explanation, the workplace safety management method through AI-based IoT sensor data and image data analysis has been described with a series of blocks, but the present invention is not limited to the order of the blocks, and some blocks are It may be performed in an order different from that shown and performed, or may be performed simultaneously. In addition, new blocks not described in the present specification and drawings may be added, or some blocks may be deleted or changed.

As mentioned above, although embodiments of the present invention have been described with reference to the accompanying drawings, those skilled in the art to which the present invention pertains can realize that the present invention can be embodied in other specific forms without changing its technical spirit or essential features. you will be able to understand Therefore, it should be understood that the embodiments described above are illustrative in all respects and not restrictive.

100: safety management server
210: IoT sensor module
220: camera module
300: operator terminal
400 : external server
500: network

Claims (15)

an IoT sensor module and a camera module disposed at the job site or attached to at least a part of the body of a worker located in the job site; and
and a safety management server that provides a safety management solution to the worker.
The safety management server,
At least one of the sensor data collected from the IoT sensor module and the image data collected from the camera module is analyzed to extract first attribute information about the work site, and use the extracted first attribute information to enter the work site. predicts the risk of the worker, and analyzes at least one of the collected sensor data and the collected image data according to a preset correlation to extract second attribute information about the worker corresponding to the extracted first attribute information, , predicts a risk to the operator according to whether the extracted first attribute information and the extracted second attribute information satisfy a preset criterion, and notifies the operator of the predicted risk as the safety management solution to provide,
When a plurality of workers are located in the work site, the plurality of workers according to whether the first attribute information for the work site and the individually extracted second attribute information for the plurality of workers satisfy the preset criteria For each of the plurality of workers based on the individual predicting the risk for each worker, the characteristic information of the plurality of workers, wherein the characteristic information of the plurality of workers includes the gender, age, and physical characteristics of the workers. By assigning weights and correcting the preset criteria based on the assigned weights, the criteria for each of the plurality of workers are individually set,
When the risk of the job site or the worker is predicted, the worker's characteristic information - the worker's characteristic information includes the worker's gender, age, and physical characteristics - based on the predicted risk coping method to determine, and to provide the worker with a notification including information on the predicted risk and information on the determined coping method,
When two or more different types of risks are predicted for one job site or one operator, one method of coping with the predicted two or more risks is determined, and information about the two or more predicted risks to the operator and Provide a notification including information on the determined one coping method, if there is no one coping method for resolving the predicted two or more risks, determining two or more coping methods for the predicted two or more risks, , to provide preferentially from the coping method determined in response to the most damaging risk among the predicted two or more risks,
Worksite safety management system through AI-based IoT sensor data and image data analysis. delete According to claim 1,
The safety management server,
By analyzing the collected sensor data, the average temperature of the work site is extracted as the first attribute information, and in response to the extracted average temperature, the collected image data is analyzed to allow the operator to extracts information about the work being performed;
Predicting the risk to the worker based on the extracted average temperature and information about the extracted work, and providing a notification to guide the worker to a break or start work according to the predicted risk,
Worksite safety management system through AI-based IoT sensor data and image data analysis. 4. The method of claim 3,
The safety management server,
The information about the extracted job includes the type, intensity, and execution time of the job, and sets a reference working time and a reference rest time for the worker based on the type and intensity of the job and the extracted average temperature, , When the work execution time of the worker exceeds the reference work time, a notification is provided to guide the worker to a break, and as the reference break time elapses from the time the worker is provided with a notification guiding a break to the worker, the providing a notification to the operator to guide the start of the job;
Worksite safety management system through AI-based IoT sensor data and image data analysis. According to claim 1,
The safety management server,
Information on the generation of harmful gas at the work site as the first attribute information by analyzing the collected sensor data - The information on the generation of harmful gas includes whether the noxious gas is generated, the type and concentration of the noxious gas - extracts, and in response to the extracted harmful gas generation information, analyzes the collected image data to extract information about the worker's wearing protective equipment and harmful gas intake information as the second attribute information,
Based on the extracted harmful gas generation information, the extracted protective equipment wearing information, and harmful gas intake information, the risk to the worker is predicted, and the worker wears or wears protective equipment according to the predicted risk. Provides a notification to inform you of the interruption of work;
Worksite safety management system through AI-based IoT sensor data and image data analysis. According to claim 1,
The safety management server,
When the work site is a coke production plant including a charging car for supplying coal, the operation information of the charging car is extracted as the first attribute information by analyzing the collected image data, and the extracted operation information of the charging car In response, analyzing the collected image data to extract the distance between the charging car and the operator as the second attribute information,
Predicting the danger to the operator based on the extracted operation information and the extracted distance of the charging car, and providing a notification to guide the operation attention of the charging car to the operator according to the predicted risk,
Worksite safety management system through AI-based IoT sensor data and image data analysis. 7. The method of claim 6,
The safety management server,
When the extracted distance becomes less than the first distance as the charging car moves, a notification is provided to the operator to guide attention to the operation of the charging car, and the extracted distance is determined by analyzing the collected image data. When it is determined that the operator does not perform a coping operation according to the provided notification until it is less than a second distance shorter than 1 distance, a notification is provided again to guide the operation attention of the charging car to the operator,
When it is determined that the operator does not perform a coping operation according to the re-provided notification until the extracted distance is less than a third distance shorter than the second distance by analyzing the collected image data, the charging car to forcibly stop the operation of
Worksite safety management system through AI-based IoT sensor data and image data analysis. delete delete According to claim 1,
The safety management server,
At least one of the collected sensor data and the collected image data is analyzed to extract first attribute information about the work site and second attribute information about the worker, and the extracted first attribute information and the extracted Predicting the risk to the job site or the operator according to the correlation between the second attribute information, and providing a notification to the operator according to the predicted risk,
When it is determined that the state of the work site is in a dangerous state based on the first attribute information, a notification is provided to the worker regardless of the correlation with the extracted second attribute information,
When it is determined that the state of the worker is in a dangerous state based on the second attribute information, providing a notification to the operator regardless of the correlation with the extracted first attribute information,
Worksite safety management system through AI-based IoT sensor data and image data analysis. According to claim 1,
The safety management server,
Analyze a plurality of sensor data collected for a predetermined period in the past based on a first time point at which the worker starts work to calculate a risk occurrence probability for a second time point after the first time point, and the calculated risk occurrence When the probability exceeds a preset value, providing a notification to guide the worker to stop the work after the first time point,
Worksite safety management system through AI-based IoT sensor data and image data analysis. delete According to claim 1,
The safety management server,
When the risk of the job site or the operator is predicted, a notification including information on the predicted risk and a coping method according to the predicted risk is provided to the operator, the collected sensor data and the collected image data Even though the worker has provided N times or more notifications by analyzing If it is determined that not, designating the worker as a person of interest, and giving a penalty to the worker designated as the person of interest,
Worksite safety management system through AI-based IoT sensor data and image data analysis. According to claim 1,
The safety management server,
When the risk to any one of the workers in the workplace or a plurality of workers in the workplace is predicted, a notification including information on the predicted risk and a coping method according to the predicted risk to the one operator provided, but
Even when the notification is provided to the one operator N or more times, the one operator does not perform the operation according to the coping method, or while a predetermined time elapses after providing the notification, the one operator responds to the coping When it is determined that the operation according to the method is not performed, the collected image data is analyzed to provide a notification provided to the one operator to another operator adjacent to the one operator among the plurality of workers,
Worksite safety management system through AI-based IoT sensor data and image data analysis. A method performed through a safety management system comprising an IoT sensor module and a camera module disposed at a work site or attached to at least a part of a worker's body located in the work site, and a safety management server that provides a safety management solution to the worker In
collecting sensor data from the IoT sensor module;
collecting image data from the camera module; and
The collected sensor data and the collected image data are analyzed to extract first attribute information on the work site, predict a risk to the work site using the extracted first attribute information, and a preset correlation relationship analyzes at least one of the collected sensor data and the collected image data according to Predicting the risk to the worker according to whether the extracted second attribute information satisfies a preset criterion, and providing a notification of the predicted risk to the worker as the safety management solution,
The step of providing a notification of the predicted risk comprises:
When a plurality of workers are located in the work site, the plurality of workers according to whether the first attribute information for the work site and the individually extracted second attribute information for the plurality of workers satisfy the preset criteria For each of the plurality of workers based on the individual predicting the risk for each worker, the characteristic information of the plurality of workers, wherein the characteristic information of the plurality of workers includes the gender, age, and physical characteristics of the workers. setting criteria for each of the plurality of workers individually by assigning weights and correcting the preset criteria based on the assigned weights;
When the risk of the job site or the worker is predicted, the worker's characteristic information - the worker's characteristic information includes the worker's gender, age, and physical characteristics - based on the predicted risk coping method determining and providing a notification including information about the predicted risk and information on the determined coping method to the operator; and
When two or more different types of risks are predicted for one job site or one operator, one method of coping with the predicted two or more risks is determined, and information about the two or more predicted risks to the operator and Provide a notification including information on the determined one coping method, if there is no one coping method for resolving the predicted two or more risks, determining two or more coping methods for the predicted two or more risks, , Comprising the step of preferentially providing a coping method determined in response to the most damaging risk among the predicted two or more risks,
Worksite safety management method through AI-based IoT sensor data and image data analysis.

Images