System and method for construction site safety and healty management
The construction site safety management system uses AI models and engines to automatically recognize and notify risks based on work type, addressing the challenge of accurately reporting task risks and equipment needs, ensuring efficient risk management.
Patent Information
- Application Number
- KR1020230089359
- Authority / Receiving Office
- KR · KR
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2023-07-10
- Publication Date
- 2026-11-06
- Estimated Expiration
- 2043-07-10
AI Technical Summary
Workers at construction sites face challenges in accurately and efficiently reporting risk levels of their tasks due to varying work categories and equipment requirements, leading to difficulties for managers in quickly identifying overall site risks.
A construction site-specific safety and health management system that utilizes AI models and engines to recognize risks based on work type information, including standard operating procedures and wearable equipment, automatically generating a work framework to identify and notify risk factors.
Enables rapid and accurate determination of construction site risks without manual input, allowing for quick identification and notification of hazardous situations.
Smart Images

Figure R1020230089359_ABST
Abstract
Description
Technology Field
[0001] The present invention relates to a construction site-specific safety and health management system and a method of operation thereof. Background Technology
[0002] Workers deployed to industrial sites, particularly construction sites, are exposed to various hazardous environments. Therefore, for effective safety and health management in industrial settings, it is crucial to identify the risk level of work performed at the site based on the type of work.
[0003] In particular, at large-scale industrial sites (e.g., the construction industry or construction maintenance sites), various tasks proceed simultaneously depending on the type of work. Therefore, it is crucial to identify hazardous situations and risk levels associated with each type of work and to implement safety and health management for the entire industrial site.
[0004] Traditionally, workers at industrial sites individually recorded information regarding the risk levels of the tasks they were performing or scheduled to perform in unspecified formats and submitted them to managers, who then aggregated this information to determine the overall risk level of the entire site. In this case, there is a problem in that it takes a significant amount of time for workers to provide managers with information regarding the risk levels of their specific tasks.
[0005] Furthermore, the level of risk varies across the dozens of different work categories, and the equipment required for workers in each category also differs. However, workers may not be able to clearly input which work category their task belongs to, nor may they record all information regarding the equipment they wear.
[0006] Therefore, there is a problem in that it is difficult for managers to quickly and accurately identify the risk level of the numerous tasks performed in industrial sites. The problem to be solved
[0007] Accordingly, the present invention provides a construction site-customized safety and health management system and a method of operation thereof, which recognizes risks in a corresponding work type based on work type information input by a safety manager and based on engines or artificial intelligence models corresponding to the work type. means of solving the problem
[0008] A method of operation of a construction site-customized safety management system operated by at least one processor for recognizing risk factors at a construction site, which is one feature of the present invention for achieving the technical problem of the present invention, wherein
[0009] The method includes the steps of mapping multiple engines to each type of work at the construction site and setting environment information so that the engines mapped in each type of work are executed; receiving information on a work type in progress from an administrator terminal and executing the multiple engines mapped to the work type based on environment information corresponding to the work type information; and using the executed multiple engines to identify risk factors at the construction site.
[0010] The above environmental information may additionally map standard work procedures corresponding to each type of work and information on equipment that workers must wear according to each type of work.
[0011] The above plurality of engines may include a plurality of engines among a standard operation procedure analysis engine, a customized complex risk recognition engine, a dialogue engine, and an equipment interoperability engine.
[0012] Prior to the step of setting the above environment information, the method may further include a step of training AI models constituting the mapped plurality of AI engines using work type learning data collected for each work type.
[0013] The step of training the above AI models allows the AI model of the customized complex risk recognition engine to be trained so that risk factors are output when the construction video included in the construction type training data and the biometric information of the worker carrying out the construction are input into the AI model of the customized complex risk recognition engine executed in the above construction type.
[0014] The step of training the above AI models allows the AI model of the conversational engine to be trained to output a work sequence according to the standard work procedure or a risk factor of the worker or construction site as a notification when a standard work procedure included in the above work type training data is input into the AI model of the conversational engine executed in the above work type.
[0015] As a method of operation of a construction site-customized safety management system operated by at least one processor for recognizing risk factors at a construction site, which is another feature of the present invention for achieving the technical problem of the present invention above,
[0016] The method includes the steps of receiving work type information of a construction site from an administrator terminal, generating a work framework including a plurality of pre-trained engines based on work type-specific environmental information corresponding to the work type information, and inputting information collected at the construction site into the work framework to identify risk factors at the construction site.
[0017] In the step of generating the above work framework, an equipment linkage engine, a standard work procedure analysis engine, a customized complex risk recognition engine, and a dialogue engine corresponding to the above work type information may be executed.
[0018] The steps executed above may include the step of the equipment linkage engine collecting video of the construction site and biometric information of workers deployed to the construction site; the step of the standard work procedure analysis engine analyzing the standard work procedure corresponding to the work type information to confirm the work information and the risk defined for the work type; and the step of the customized complex risk recognition engine determining the risk level of the construction site based on the video and biometric information collected by the equipment linkage engine and the work information and risk level analyzed by the standard work type analysis engine.
[0019] The above-described steps may further include the step of notifying the work information or risk level to at least one of the worker terminal or manager terminal at the construction site.
[0020] As another feature of the present invention for achieving the technical problem of the present invention, a construction site customized safety management system that determines the risk level according to the type of work at the construction site,
[0021] The system includes an interface for receiving video of the construction site and biometric information of a worker deployed to the construction site, a memory for storing environmental information including information on a plurality of engines mapped to be executed for each type of work occurring at the construction site, and a processor. The processor sets the environmental information by mapping a plurality of engines to be executed for each type of work at the construction site, and upon receiving information on a work type in progress from a manager terminal, identifies risk factors at the construction site using a plurality of engines executed based on the environmental information corresponding to the work type information.
[0022] The above memory may additionally have standard work procedures corresponding to each type of work and information on wearable equipment that a worker must wear according to each type of work mapped to the above environment information.
[0023] The above processor can train AI models constituting the mapped plurality of engines using work type learning data collected for each work type.
[0024] The processor can train the AI model of the customized complex risk recognition engine to output risk factors when the construction video included in the construction type training data and the biometric information of the worker carrying out the construction are input into the AI model of the customized complex risk recognition engine executed in the construction type.
[0025] The above processor can train the AI model of the conversational engine to output a work sequence according to the standard work procedure or a risk factor of the worker or construction site as a notification when the standard work procedure included in the work type learning data is input into the AI model of the conversational engine executed in the work type.
[0026] The above processor generates an equipment linkage engine, a standard work procedure analysis engine, a customized complex risk recognition engine, and a dialogue engine executed in response to the above work type information as a work framework, and can determine the risk level of the above construction site based on the above collected video and biometric information and the above defined risk. Effects of the invention
[0027] According to the present invention, even without an industrial site worker inputting information regarding the risk level of the work they are currently performing or are scheduled to perform, they can quickly and accurately determine the risk situation of the work currently being carried out at the construction site by using work-type engines.
[0028] In addition, even without manually running engines for a very large number of work types, multiple engines corresponding to the work type can be automatically executed and configured into a single framework when work type information is entered based on pre-configured environment information. Brief explanation of the drawing
[0029] FIG. 1 is an example of an environment in which a construction site-customized safety and health management system according to an embodiment of the present invention is applied. FIG. 2 is an exemplary diagram of a safety and health management system according to an embodiment of the present invention. FIG. 3 is a structural diagram of a customized complex risk recognition engine control unit according to an embodiment of the present invention. FIG. 4 is an operation flowchart of a safety and health management system according to an embodiment of the present invention. FIG. 5 is a flowchart illustrating a method for determining risks at a construction site in a safety and health management system according to an embodiment of the present invention. FIG. 6 is a detailed operation flowchart of a safety and health management system according to an embodiment of the present invention. FIGS. 7 to 10 are exemplary diagrams showing the operation of each component of a safety and health management system according to an embodiment of the present invention. FIG. 11 is an example diagram of the operation of a conversation engine according to an embodiment of the present invention. FIG. 12 is a structural diagram of a computing device according to an embodiment of the present invention. Specific details for implementing the invention
[0030] The present invention will be described in detail below by explaining embodiments of the invention with reference to the attached drawings. In describing the present invention, if it is determined that a detailed description of related known configurations or functions may obscure the essence of the invention, such detailed description will be omitted. Identical reference numerals in each drawing may indicate identical components.
[0031] The terms used herein are used merely to describe specific embodiments and are not intended to limit the invention. Singular expressions include plural expressions unless the context clearly indicates otherwise. In this specification, terms such as “comprising” or “having” are intended to indicate the existence of the features, numbers, steps, actions, components, parts, or combinations thereof described herein, and should be understood as not precluding the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.
[0032] Throughout the specification, when a part is described as being "connected" to another part, this includes not only cases where they are "directly connected" but also cases where they are "electrically or mechanically connected" with other components in between.
[0033] Unless otherwise defined, the terms used herein, including technical or scientific terms, have the same meaning as generally understood by a person of ordinary skill in the art to which the present invention pertains. Terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology, and should not be interpreted in an ideal or overly formal sense unless explicitly defined in this specification.
[0034] Hereinafter, a construction site-specific safety and health management system and a method of operation thereof according to an embodiment of the present invention will be described in detail with reference to the drawings.
[0035] FIG. 1 is an example of an environment in which a construction site-customized safety and health management system according to an embodiment of the present invention is applied.
[0036] As illustrated in FIG. 1, a construction site-specific safety and health management system (hereinafter referred to as the "safety management system" for convenience of explanation) (100) is linked with the site manager's manager terminal (200). Additionally, the safety management system (100) can be linked with the site worker's worker terminal (300) or wearable equipment (400) such as a safety helmet, insulating gloves, and safety vest worn by the worker.
[0037] In addition, the safety management system (100) can also be linked with information collection means (500), such as a CCTV that collects video of the construction site or a camera installed on a robot deployed to the construction site.
[0038] In the safety management system (100), environmental information for each type of work that occurs during the entire construction phase is set in advance.
[0039] Here, the environmental information by work type includes the Standard Operating Procedure (SOP) for each work type. Additionally, the environmental information by work type defines at least one engine to be executed according to the work type. Furthermore, the environmental information by work type maps information on the wearable equipment (400) that a worker must wear according to the work type.
[0040] When environmental information for each type of work is set, the safety management system (100) trains a customized composite AI model that constitutes at least one engine to be executed for each type of work based on information such as construction video collected at each stage of work, bio information, standard work procedures for each type of work, and risk levels defined in the standard work procedures.
[0041] At this time, the safety management system (100) trains an AI model using construction videos, biometric information, standard work procedures, risk levels, etc., as training data for each type of work. Since there are various ways for the safety management system (100) to train an AI model using training data for each type of work, the embodiment of the present invention is not limited to any one method.
[0042] The safety management system (100) trains an AI model mounted on an engine using training data, and then receives construction-related information including work type information from a manager terminal (200) or a terminal (not shown in the drawing) held by a person involved in the construction. Based on the work type information included in the received construction information, the safety management system (100) verifies the standard work procedure for the corresponding process.
[0043] And the safety management system (100) creates a work framework by executing learned engines based on pre-set work-specific environmental information for the work type. Here, the work framework consists of multiple engines mapped to the work type to process information generated in the work type and to determine whether there are risk factors for the construction site or workers.
[0044] The safety management system (100) inputs at least one piece of information, including a construction site image transmitted from an information collection means (500), a worker's biometric information transmitted from a wearable device (400), and a verified standard work procedure and risk level, into a generated work framework to recognize whether there are risk factors at the construction site.
[0045] At this time, the safety management system (100) can provide voice or text to a worker terminal or a manager terminal, guiding information related to construction procedures or construction types using a pre-trained conversational engine for each construction type, depending on the type of work being carried out at the construction site. Since the terminology used for each type of work and the construction procedures for each type of work differ depending on the type of work, information can be easily provided by using a conversational engine for each type of work that has been pre-trained for each type of work.
[0046] In addition, if the safety management system (100) determines that there is a risk factor at the construction site or for a worker, it provides risk information as a notification to a terminal held by the worker or a terminal held by the manager. In the embodiment of the present invention, the conversation engine for each type of work is described as providing a notification in either voice or text, but notifications can be provided in various forms.
[0047] The following describes an example of the configuration of a safety management system (100) with reference to FIG. 2.
[0048] FIG. 2 is an exemplary diagram of a safety management system according to an embodiment of the present invention.
[0049] As illustrated in FIG. 2, the safety management system (100) includes a plurality of engines (110 to 140). In the embodiment of the present invention, the engines (110 to 140) are described as including a standard work procedure analysis engine (110), a customized risk recognition engine (120), a conversation engine (130), and an equipment interlocking engine (140), but are not necessarily limited to this.
[0050] Additionally, the safety management system (100) may further include a work framework generation engine (160) and a customized complex risk recognition engine control unit (170). In the embodiment of the present invention, for convenience of explanation, it is referred to as a customized complex risk recognition engine control unit (170), but the customized complex risk recognition engine control unit (170) may also be a type of engine.
[0051] Here, the conversation engine (130) can be used as an artificial intelligence engine. Also, the equipment linkage engine (140) can use an interface and may be an IoT device.
[0052] Additionally, the safety management system (100) also includes a work-type environmental information database (150). The work-type environmental information database (150) contains information mapped to each work type, such as standard work procedures for each work type, information on at least one engine to be executed according to the work type, and information on wearable equipment (400) that a worker must wear according to the work type.
[0053] The standard work procedure analysis engine (110), customized risk recognition engine (120), conversation engine (130), and equipment linkage engine (140) corresponding to the engines each have engines added for each type of work at the construction site. In addition, each engine includes AI models trained with work type learning data.
[0054] For example, the standard work procedure analysis engine (110) is shown to include a transmission work type SOP analysis engine, a substation work type SOP analysis engine, a distribution work type SOP analysis engine, and a material center SOP analysis engine. In addition, in the embodiment of the present invention, it is represented as a single engine such as the transmission work type SOP analysis engine, but 27 transmission process SOP analysis engines corresponding to transmission work may each be included.
[0055] Likewise, in the embodiments of the present invention, only the safety helmet interlocking engine, the safety vest interlocking engine, and the insulating glove interlocking engine are shown as examples of the equipment interlocking engine (140). An interlocking engine that interlocks with all wearable equipment (400) worn by a worker may be additionally included.
[0056] When the safety management system (100) receives information on the type of work at the construction site from the manager terminal (200), it checks the environmental information stored in the environmental information database (150) for each type of work. Then, based on the checked environmental information, the work framework generation engine (160) generates a work framework by including a plurality of engines that are mapped to and executed for the corresponding type of work. The work framework may include a message queue-based AI framework (not shown in the drawing) capable of simultaneously recognizing risks in a complex manner based on images, biometric information, and risk levels.
[0057] That is, in the embodiment of the present invention, the customized risk recognition engine (120) is described as an example in which at least one customized risk recognition engine is included, such as a transmission process customized risk recognition engine, a substation process customized risk recognition engine, a distribution process customized risk recognition engine, and a material center customized risk recognition engine. Each customized risk recognition engine is composed of a plurality of engines, such as an engine that recognizes a worker's face in an image, an engine that recognizes biometric information from wearable equipment worn by the worker, and an engine that calculates the risk level.
[0058] The customized complex risk recognition engine control unit (170), based on work type information, instructs the activation of the engine requiring information collection according to the work type among the multiple deactivated engines included in each customized risk recognition engine. Then, it receives the information collected by each activated engine and transmits it to the work framework generation engine (160). At this time, the customized complex risk recognition engine control unit (170) inserts the information collected by each engine into each message queue-based AI framework and transmits it to the work framework generation engine (160).
[0059] In an embodiment of the present invention, the customized composite risk recognition engine control unit (170) is described as being implemented separately from the customized risk recognition engine (120), but the customized composite risk recognition engine control unit (170) may also be implemented to be included in the customized risk recognition engine (120).
[0060] Here, the structure of the customized complex risk recognition engine control unit (170) is described with reference to FIG. 3.
[0061] FIG. 3 is a structural diagram of a customized complex risk recognition engine control unit according to an embodiment of the present invention.
[0062] As illustrated in FIG. 3, the customized complex risk recognition engine control unit (170) is composed of an AI interface (171) and a plurality of processing modules (172) that collect and process information.
[0063] In the embodiments of the present invention, examples are provided that include a biometric information collection module for collecting biometric information in a plurality of processing modules (172), image receiving modules for receiving images through various channels, a biometric danger situation recognition AI, a human face recognition AI, a safety equipment recognition AI, a danger situation recognition AI, etc., but the invention is not necessarily limited to this.
[0064] Additionally, the customized complex risk recognition engine control unit (170) may further include a message queue generation module (173) corresponding to the AI framework.
[0065] The AI interface (171) can retrieve or display information to manage safety and health based on work type information through methods such as GET and POST. The expression of the AI interface (171) is a known technology, and the embodiment of the present invention is not limited to any one of the technologies.
[0066] The processing module (172) collects biometric information according to the representation of the AI interface (171) or collects construction videos through various channels. Additionally, the processing module (172) can recognize biometric danger situations from the biometric information. Additionally, the processing module (172) can recognize the worker's face from the construction videos, recognize safety equipment worn by the worker, or recognize whether the worker is in a dangerous situation.
[0067] The message queue generation module (173) generates an AI framework including the biological danger status, worker face recognition result, safety equipment recognition result, and danger status result processed by the processing module (172). Since the method by which the message queue generation module (173) generates an AI framework including various information can be executed through various methods, the embodiment of the present invention is not limited to any one method.
[0068] And the generated AI framework can be provided to an app installed on a message notification module, a supervisor terminal, or a worker terminal. Alternatively, the message queue generation module (173) can provide the generated AI framework to a work framework generation engine (160).
[0069] In such an environment, the operation of the safety management system (100) identifying risk factors at the construction site is explained with reference to FIGS. 4 to 6.
[0070] FIG. 4 is an operation flowchart of a safety management system according to an embodiment of the present invention.
[0071] As illustrated in FIG. 4, the safety management system (100) sets environmental information for each type of work that may occur at the construction site from the outside (S100). Here, setting environmental information for each type of work means mapping standard work procedures for each type of work, at least one engine executed for each type of work, information on wearable equipment (400) that a worker must wear in each type of work, and information on risks that may occur for each type of work, etc., to each type of work and registering them in the safety management system (100).
[0072] At this time, since there are various methods for setting environmental information for each type of work in the safety management system (100), the embodiment of the present invention is not limited to any one method.
[0073] The safety management system (100) receives construction information from a manager terminal (200) at an actual construction site (S200). Here, the construction information may include not only basic information about the construction, such as the location of the construction site and the construction schedule, but also the type of process (e.g., transmission, substation, distribution, material center, etc.) and the stage of the construction (e.g., work type name or work type code, etc.).
[0074] Here, information regarding the construction phase can be registered in the construction phase environmental information database (150) as shown in Table 1 below.
[0075] division Work type Work type code Monitoring revision page Power distribution (86 types) Processing Article 39 General processing (dry poles, long poles, overhead wires) D01 Target '22.07 128 Transformer installation D02 Target '22.07 130 Switch installation D03 Target '22.07 131 Intelligent power distribution work (excluding switchgear work) D04 Target '22.07 132 Multifunctional wire installation method for power distribution work D05 Target '22.07 133 Periodic check D06 Target '22.07 134 Low-voltage maintenance work, service line construction D07 Target '22.07 135 Grounding, leakage current, voltage / current measurement D08 Target '22.07 136 … … … … … …
[0076] As shown in Table 1 above, information regarding the work type stage may include the work type name, work type code, and information on whether the work is subject to monitoring. Additionally, it may include revision information for the work type and information on the page containing the information for the work type.
[0077] The safety management system (100) checks the environmental information stored and mapped to the corresponding work type based on the received work type information (S300). The environmental information includes a standard work procedure that includes multiple engines executed to process information generated in the corresponding work type or to determine the risk level of the construction site or workers, the name of the work type, the work type code, and information on the risk level for each work type.
[0078] Then, the safety management system (100) executes a plurality of engines corresponding to the work type based on the environmental information confirmed in step S300, and creates a work framework including the executed plurality of engines (S400). Since the method by which the safety management system (100) creates a work framework with a plurality of engines can be executed using various techniques, the embodiment of the present invention is not limited to any one method.
[0079] The safety management system (100) collects construction footage or biometric information of workers at the construction site based on the generated work framework. Then, the safety management system (100) analyzes the collected information to recognize whether there are risk factors at the construction site or for the workers (S500).
[0080] At this time, although not illustrated in the embodiments of the present invention, a learning procedure for training the AI models inserted into the engine may be additionally executed prior to step S100. In the embodiments of the present invention, a deep learning ensemble model is used as an example for the AI model, but it is not necessarily limited to this.
[0081] That is, the safety management system (100) can train an AI model included in an engine executed for each type of work based on at least one of construction videos collected for each type of work, biometric information of workers at the construction site, and information on the type of work. In other words, when the safety management system (100) inputs construction videos, biometric information, and standard work procedures into a work framework created for each type of work, it trains an AI model included in each of the multiple engines that form a composite AI framework within the work framework so that a risk level or risk factor is output.
[0082] Additionally, the safety management system (100) can train an AI model to run a conversational engine that notifies a worker or manager of a danger in voice or text using terms used for each type of work in order to notify the worker that there is a dangerous situation. Additionally, the safety management system (100) can train an AI model to run a standard work procedure analysis engine that verifies a standard work procedure corresponding to the type of work information.
[0083] In an embodiment of the present invention, a standard work procedure analysis engine (110), a customized complex risk recognition engine control unit (170), and a pre-trained AI model within a conversational engine (130) may be included. Since the method by which the safety management system (100) trains the AI model according to each work type can be implemented using various techniques, the embodiment of the present invention is not limited to any one method.
[0084] In the aforementioned Fig. 4, a method for recognizing the presence or absence of a risk factor at step S500 is explained with reference to Fig. 5.
[0085] FIG. 5 is a flowchart illustrating a method for determining the risk of a construction site in a safety management system according to an embodiment of the present invention.
[0086] As illustrated in FIG. 5, the safety management system (100) simultaneously receives construction video and biometric information from the construction video and the wearable equipment (400) worn by the worker, respectively, through the information collection means (500) of the construction site linked via the executed engines (S610, S620).
[0087] The safety management system (100) measures the risk level by analyzing construction images and biometric information based on a risk assessment to determine if there are risk factors at the construction site, using a customized risk recognition engine among the pre-trained engines (S630). Then, the safety management system (100) checks whether there are risk factors at the construction site based on the analysis results (S640).
[0088] For example, if the worker's body temperature among the collected biometric information is higher than the pre-set normal range of body temperature, the safety management system (100) recognizes that the worker is in a dangerous situation. As such, since the method by which the customized risk recognition engine determines whether there are risk factors based on construction video and biometric information can be determined based on various criteria, the embodiment of the present invention is not limited to any one method.
[0089] When the safety management system (100) determines at step S640 that there is a risk factor at the construction site, the conversation engine notifies the safety manager or worker that there is a dangerous situation in at least one of voice or text (S650).
[0090] Next, an embodiment in which each component of the safety management system (100) operates according to the type of work is described with reference to FIG. 6.
[0091] FIG. 6 is a detailed operation flowchart of a safety management system according to an embodiment of the present invention.
[0092] As illustrated in FIG. 6, an administrator generates environmental information using an administrator terminal (200) (S700). That is, environmental information is generated that maps information about the process and which engines operate in the corresponding process.
[0093] Environmental information generated by the administrator through the administrator terminal (200) is transmitted to the safety management system (100) (S701), and the environmental information is set so that a plurality of engines (110~160) constituting the safety management system (100) are executed according to the type of work (S702). The method of setting the engines (110~160) based on the environmental information by the safety management system (100) is already known, and the embodiment of the present invention is not limited to any one method.
[0094] After environmental information is set in the safety management system (100) in this manner, when an administrator inputs work type information using the administrator terminal (200) (S703), the safety management system (100) executes multiple engines (110~140) that match the work type included in the work type information (S704). At this time, the engines executed are described as including an SOP analysis engine (110), a customized complex risk recognition engine control unit (170), a dialogue engine (130), and an interlocking engine (140).
[0095] When the executed engines (110 to 140) each transmit engine information to the work framework generating engine (160) (S705), the work framework generating engine (160) generates a work framework according to one type of work from the multiple executed engines (110 to 140) (S706). Since there are various ways for the work framework generating engine (160) to generate a work framework from the multiple engines (110 to 140, 170) into one work framework, the embodiment of the present invention is not limited to any one method.
[0096] After the work framework generation engine (160) generates the work framework, the work framework generation engine (160) receives biometric information from the worker wearable equipment (400) linked through the linkage engine (140) (S707). At this time, the work framework generation engine (160) may also receive images of the construction site from the worker wearable equipment (400) or through the information collection means (500).
[0097] The work framework generation engine (160) analyzes the received information to determine whether there are risk factors at the construction site (S707), and can notify the manager terminal (200) of the risk factors based on the determination result (S708).
[0098] Next, examples of how the safety management system (100) operates in various types of work are described with reference to FIGS. 7 to 10. First, examples of how the safety management system (100) operates are described with reference to FIG. 7, using power distribution work as an example.
[0099] FIG. 7 is an exemplary diagram showing the operation of each component of a safety management system according to an embodiment of the present invention.
[0100] The safety management system (100) includes multiple engines that include a learned AI model and are executed based on environmental information for each type of work, and the executed multiple engines process information collected from the outside to determine whether there are risk factors at the construction site.
[0101] That is, when information indicating that a power distribution work is in progress at a construction site is input from the administrator terminal (200), a power distribution work SOP analysis engine, a customized complex risk recognition engine control unit, and a power distribution-related conversation engine, each including a previously trained AI model as shown in FIG. 7, are executed.
[0102] And it is assumed that when power distribution work is being carried out, the worker must wear a safety helmet, safety vest, and insulating gloves, as is set in the work-type configuration information stored in the work-type environment information database (150). The safety helmet linkage engine, safety vest linkage engine, and insulating glove linkage engine included in the safety equipment linkage engine (140) are also executed simultaneously.
[0103] The engines executed in this manner, namely the power distribution work type SOP analysis engine, the customized complex risk recognition engine control unit, the power distribution related conversation engine, the safety helmet linkage engine, the safety vest linkage engine, and the insulating glove linkage engine, are generated into a work framework for the power distribution work type through the work framework generation engine (160).
[0104] The executed power distribution work type SOP analysis engine analyzes the standard work procedures of power distribution work types to identify potential risk factors that may occur during the work. This explanation uses an example where the power distribution work type SOP analysis engine extracts the work type name and code included in the standard work procedures, as well as pre-set risk information corresponding to those work types.
[0105] At this time, the customized complex risk recognition engine control unit executes at least one engine according to the power distribution type among the plurality of engines constituting the power distribution type customized risk recognition engine, and transmits the information collected by the executed engines to the work framework generation engine (160).
[0106] In other words, the customized complex risk recognition engine control unit operates each engine of the distribution work-type customized risk recognition engine to receive construction footage or worker biometric information required for processing in the distribution work type, based on work type information.
[0107] And, the customized complex risk recognition engine control unit receives construction footage of the power distribution work type or biometric information of the worker collected from the information collection means (500) from each engine. The power distribution work type customized risk recognition engine is described as an example of receiving and analyzing biometric information collected through the safety helmet, safety vest, and insulating gloves worn by the worker from the equipment interlocking engine (140).
[0108] Furthermore, the received construction video and biometric information are analyzed to determine whether the worker is exposed to danger or if there are risk factors at the construction site. The method by which the customized complex risk recognition engine control unit receives construction video or biometric information analyzed by the power distribution type-specific risk recognition engine to identify risk factors is a known technology, and the embodiments of the present invention are not limited to any one of these methods.
[0109] The power distribution conversation engine can provide construction procedures corresponding to power distribution work types in voice or text to a worker terminal or a manager terminal. Additionally, when the customized complex risk recognition engine control unit (170) of the power distribution conversation engine confirms that there is a risk factor at the construction site or for a worker, it provides a notification of the risk factor to the worker terminal or the manager terminal.
[0110] Next, an example of how the safety management system (100) operates using a material center work type is explained with reference to FIG. 8.
[0111] When information indicating that a construction type for the material center is in progress at the construction site is input from the administrator terminal (200), the material center construction type SOP analysis engine, the customized complex risk recognition engine control unit, and the material center related conversation engine, each including a previously trained AI model as shown in FIG. 8, are executed.
[0112] And it is assumed that when a material center work process is underway, the worker must wear a safety helmet and a safety vest, as is set in the work-type configuration information stored in the work-type environment information database (150). The safety helmet linkage engine and the safety vest linkage engine included in the safety equipment linkage engine (140) are also executed simultaneously.
[0113] The engines executed in this manner, namely the material center work type SOP analysis engine, the customized complex risk recognition engine control unit, the material center related conversation engine, the safety helmet linkage engine, and the safety vest linkage engine, are generated into a work framework for the material center work type through the work framework generation engine (160).
[0114] The executed Material Center Work Item SOP Analysis Engine analyzes the standard work procedures of Material Center work items to identify potential risk factors that may occur during the work. This explanation uses an example where the Material Center Work Item SOP Analysis Engine extracts the work item name and code included in the standard work procedure, as well as pre-set risk information corresponding to those work items.
[0115] The customized complex risk recognition engine control unit (170) receives construction video or worker biometric information regarding the material center work type collected by at least one of the multiple engines constituting the material center work type customized risk recognition engine according to its control. Additionally, the customized complex risk recognition engine control unit (170) is described as an example of receiving and analyzing biometric information collected through the safety helmet and safety vest worn by the worker from the equipment linkage engine (140).
[0116] Then, the received construction video and biometric information are analyzed to determine whether the worker is exposed to danger or if there are risk factors at the construction site. The method by which the customized composite risk recognition engine control unit (170) analyzes the construction video or biometric information to identify risk factors is a known technique, and the embodiment of the present invention is not limited to any one of these methods.
[0117] The material center-related conversation engine can provide construction procedures corresponding to material center construction types in voice or text to a worker terminal or a manager terminal. Additionally, when the material center-related conversation engine control unit (170) confirms that there is a risk factor at the construction site or for a worker, it provides a notification of the risk factor to the worker terminal or the manager terminal.
[0118] Next, an example of how the safety management system (100) operates is described with reference to FIG. 9, using a transmission work type as an example.
[0119] When information indicating that a power transmission work is in progress at a construction site is input from the administrator terminal (200), a power transmission work SOP analysis engine, a customized complex risk recognition engine control unit (170), and a power transmission-related conversation engine each containing a pre-trained AI model as shown in FIG. 9 are executed.
[0120] And it is assumed that when the power transmission work is in progress, the worker must wear a safety helmet, safety vest, and insulating gloves, as is set in the work-type configuration information stored in the work-type environment information database (150). The safety helmet linkage engine, safety vest linkage engine, and insulating glove linkage engine included in the safety equipment linkage engine (140) are also executed simultaneously.
[0121] The engines executed in this manner, namely the transmission work type SOP analysis engine, the customized complex risk recognition engine control unit (170), the transmission-related conversation engine, the safety helmet linkage engine, the safety vest linkage engine, and the insulating glove linkage engine, are generated into a work framework for the transmission work type through the work framework generation engine (160).
[0122] The executed transmission work type SOP analysis engine analyzes the standard work procedures of transmission work types to identify potential risk factors that may occur during the work. This explanation uses an example where the transmission work type SOP analysis engine extracts the work type name and code included in the standard work procedures, as well as pre-set risk information corresponding to those work types.
[0123] The customized complex risk recognition engine control unit (170) receives construction footage of the transmission work type or biometric information of the worker collected from the information collection means (500) from the engines within the transmission process customized risk recognition engine. The customized complex risk recognition engine control unit (170) is described by example by receiving and analyzing biometric information collected through the safety helmet, safety vest, and insulating gloves worn by the worker from the equipment interlocking engine (140).
[0124] Then, the received construction video and biometric information are analyzed to determine whether the worker is exposed to danger or if there are risk factors at the construction site. The method by which the customized composite risk recognition engine control unit (170) analyzes the construction video or biometric information to identify risk factors is a known technique, and the embodiment of the present invention is not limited to any one of these methods.
[0125] The transmission-related conversation engine can provide construction procedures corresponding to the transmission work type in voice or text to a worker terminal or a manager terminal. In addition, when the customized complex risk recognition engine control unit (170) of the transmission-related conversation engine confirms that there is a risk factor at the construction site or for a worker, it provides a notification of the risk factor to the worker terminal or the manager terminal.
[0126] Next, an example of how a safety management system (100) operates is described with reference to FIG. 10, using a substation construction type as an example.
[0127] When information indicating that a substation work is in progress at the construction site is input from the administrator terminal (200), a substation work SOP analysis engine, a customized complex risk recognition engine control unit (170), and a substation-related conversation engine each containing a pre-trained AI model as shown in FIG. 10 are executed.
[0128] And it is assumed that when the substation work is being carried out, the worker must wear a safety helmet and a safety vest, as is set in the work-type configuration information stored in the work-type environment information database (150). The safety helmet linkage engine and the safety vest linkage engine included in the safety equipment linkage engine (140) are also executed simultaneously.
[0129] The engines executed in this manner, namely the substation work type SOP analysis engine, the customized complex risk recognition engine control unit (170), the substation-related conversation engine, the safety helmet linkage engine, and the safety vest linkage engine, are generated into a substation work type work framework through the work framework generation engine (160).
[0130] The executed substation work type SOP analysis engine analyzes the standard work procedures of substation work types to identify potential risk factors that may occur during the work. This explanation uses an example where the substation work type SOP analysis engine extracts the work type name and code included in the standard work procedures, as well as pre-set risk information corresponding to those work types.
[0131] The customized complex risk recognition engine control unit (170) receives construction footage of a substation work type or biometric information of a worker from engines activated according to the control of the customized complex risk recognition engine control unit (170) among a plurality of engines included in the substation work type customized risk recognition engine. The customized complex risk recognition engine control unit (170) is described as an example of receiving and analyzing biometric information collected through a safety helmet and safety vest worn by a worker from the equipment interlocking engine (140).
[0132] Then, the received construction video and biometric information are analyzed to determine whether the worker is exposed to danger or if there are risk factors at the construction site. The method by which the customized composite risk recognition engine control unit (170) analyzes the construction video or biometric information to identify risk factors is a known technique, and the embodiment of the present invention is not limited to any one of these methods.
[0133] The substation-related conversation engine can provide construction procedures corresponding to the substation construction type in voice or text to a worker terminal or a manager terminal. In addition, when the custom complex risk recognition engine control unit (170) of the substation-related conversation engine confirms that there is a risk factor at the construction site or for a worker, it provides a notification of the risk factor to the worker terminal or the manager terminal.
[0134] Next, an example of the operation of a conversation engine (130) according to an embodiment of the present invention will be described with reference to FIG. 11.
[0135] FIG. 11 is an example diagram of the operation of a conversation engine according to an embodiment of the present invention.
[0136] As illustrated in FIG. 11, the conversation engine (130) includes a voice receiving module (131), a conversational AI interface (132), and a main module (133). Additionally, in the embodiment of the present invention, the conversation engine (130) is illustrated as an example including an AI computer system (134), an STT server (135), and a TTS server (136), but the AI computer system (134), the STT server (135), and the TTS server (136) may be implemented as separate physical devices.
[0137] When power is first applied to the safety management system (100), the voice receiving module (131) of the conversation engine (130) starts operating (①). Then, the main module (133) registers a service component (②), and the conversational AI interface (132) also performs connection initialization with the AI computer system (134).
[0138] After the initialization phase is completed, the administrator utters a call command to execute the conversation engine (130) (④). In an embodiment of the present invention, the call command "safety construction" is uttered as an example.
[0139] After the manager utters a call command, if a command such as “start construction” is uttered, the main module (133) transmits the “start construction” command uttered by the manager to the STT server (135) (⑤). In FIG. 11, the part indicated by the solid line is the path through which the voice signal is transmitted, and the part indicated by the dotted line is the path through which the text is transmitted.
[0140] The STT server (135) converts the command received from the main module (133) into text and returns it to the conversational AI interface (132) through the main module (133) (⑥). The conversational AI interface (132) transmits the returned text to the AI computer system (134) (⑦).
[0141] The AI computer system (134) generates a query text corresponding to the returned text and transmits it to the main module (133) via the conversational AI interface (132) (⑧). The main module (133) transmits the query text to the TTS server (136) (⑨), receives the query voice converted into speech from the TTS server (136), and outputs the query voice so that an administrator can verify it (⑩). In an embodiment of the present invention, the output of the query voice “Shall we really start?” is used as an example for explanation.
[0142] When the main module (133) receives a response voice to a question voice, namely “Let’s really start” (11), it repeats steps ⑨ and ⑩ described above. Through this, the conversation engine (130) can provide voice notifications to workers or managers regarding risk factors identified by the safety management system (100) for each type of work.
[0143] FIG. 12 is a structural diagram of a computing device according to an embodiment of the present invention.
[0144] As illustrated in FIG. 10, the safety management system (100) executes a program containing instructions described to execute the operation of the present invention in a computing device (600) operated by at least one processor.
[0145] The hardware of the computing device (600) may include at least one processor (610), memory (620), storage (630), and communication interface (640), and may be connected via a bus. In addition, hardware such as input devices and output devices may be included. The computing device (600) may be equipped with various software, including an operating system capable of running programs.
[0146] The processor (610) is a device that controls the operation of the computing device (600) and may be a various type of processor (510) that processes instructions included in a program, for example, a CPU (Central Processing Unit), MPU (Micro Processor Unit), MCU (Micro Controller Unit), GPU (Graphic Processing Unit), etc.
[0147] The memory (620) loads the corresponding program so that instructions described for executing the operation of the present invention are processed by the processor (610). The memory (620) may be, for example, ROM (read only memory), RAM (random access memory), etc. The storage (630) stores various data, programs, etc. required to execute the operation of the present invention. The communication interface (540) may be a wired / wireless communication module.
[0148] Although embodiments of the present invention have been described in detail above, the scope of the present invention is not limited thereto, and various modifications and improvements by those skilled in the art using the basic concept of the present invention as defined in the following claims also fall within the scope of the present invention.
Claims
Claim 1 A method of operation of a construction site-specific safety management system operated by at least one processor for recognizing risk factors at a construction site, comprising: mapping a plurality of engines for each type of work at the construction site and setting environment information so that the engines mapped in each type of work are executed; receiving information on a type of work in progress from a manager terminal, executing a plurality of engines mapped to the type of work based on environment information corresponding to the type of work information; and identifying risk factors at the construction site using the executed plurality of engines, wherein the plurality of engines include a plurality of engines among a standard work procedure analysis engine, a customized complex risk recognition engine, a dialogue engine, and an equipment linkage engine. Claim 2 A method of operation according to claim 1, wherein the above-mentioned environmental information is additionally mapped to standard work procedures corresponding to each work type and information on wearable equipment that a worker must wear according to each work type. Claim 3 delete Claim 4 A method of operation according to claim 1, further comprising, prior to the step of setting the environment information, a step of training AI models constituting the mapped plurality of AI engines using work type learning data collected for each work type. Claim 5 In claim 4, the step of training the AI models is a method of operation in which, when a construction video included in the construction type training data and biometric information of a worker performing the construction are input into the AI model of a customized complex risk recognition engine executed in the construction type, the AI model of the customized complex risk recognition engine is trained such that a risk factor is output. Claim 6 In claim 5, the step of training the AI models is a method of operation in which, when a standard work procedure included in the work type training data is input into the AI model of a conversational engine executed in the work type, the AI model of the conversational engine is trained to output a work sequence according to the standard work procedure or a risk factor of the worker or construction site as a notification. Claim 7 A method of operation of a construction site-specific safety management system operated by at least one processor for recognizing risk factors at a construction site, comprising: receiving construction type information of the construction site from a manager terminal; generating a work framework including a plurality of pre-learned engines based on construction type-specific environmental information corresponding to the construction type information; executing an equipment linkage engine, a standard work procedure analysis engine, a customized complex risk recognition engine, and a dialogue engine corresponding to the construction type information; and inputting information collected at the construction site into the work framework to identify risk factors at the construction site. Claim 8 delete Claim 9 A method of operation according to claim 7, comprising, after the step of executing the above, a step in which the equipment linkage engine collects video of the construction site and biometric information of a worker deployed to the construction site; a step in which the standard work procedure analysis engine analyzes a standard work procedure corresponding to the work type information to confirm the work information and the risk defined for the work type; and a step in which the customized complex risk recognition engine determines the risk level of the construction site based on the video and biometric information collected by the equipment linkage engine and the work information and risk level analyzed by the standard work procedure analysis engine. Claim 10 In claim 9, the above-described execution step further comprises the step of notifying the above-described work information or risk level to at least one of the worker terminal or manager terminal of the construction site. Claim 11 A construction site customized safety management system that determines the risk level according to the type of work at the construction site, comprising: an interface that receives video of the construction site and biometric information of a worker deployed to the construction site; a memory that stores environmental information including information of a plurality of engines mapped to be executed for each type of work occurring at the construction site; and a processor, wherein the processor sets the environmental information by mapping a plurality of engines to be executed for each type of work at the construction site, and upon receiving information on a work type in progress from a manager terminal, identifies risk factors of the construction site using a plurality of engines executed based on the environmental information corresponding to the work type information, wherein the plurality of engines include a standard work procedure analysis engine, a customized complex risk recognition engine, a dialogue engine, and an equipment linkage engine. Claim 12 In Clause 11, the above memory is a construction site-customized safety management system in which standard work procedures corresponding to each type of work and information on wearable equipment that a worker must wear according to each type of work are additionally mapped to the above environment information. Claim 13 In Clause 12, the above processor trains AI models constituting the mapped plurality of engines using work type learning data collected for each work type, in a construction site customized safety management system. Claim 14 A construction site customized safety management system according to claim 13, wherein the processor trains the AI model of the customized complex risk recognition engine so that risk factors are output when the processor inputs the construction video included in the construction type learning data and the biometric information of the worker carrying out the construction into the AI model of the customized complex risk recognition engine executed in the construction type. Claim 15 A construction site customized safety management system according to claim 14, wherein the processor inputs a standard work procedure included in the work type learning data into an AI model of a conversational engine executed in the work type, and trains the AI model of the conversational engine to output a work sequence according to the standard work procedure or a risk factor of the worker or construction site as a notification. Claim 16 A construction site customized safety management system according to claim 15, wherein the processor generates an equipment linkage engine, a standard work procedure analysis engine, a customized complex risk recognition engine, and a dialogue engine executed in response to the above work type information as a work framework, and determines the risk level of the construction site based on the collected video and biometric information and the risk defined for the above work type.
Citation Information
Patent Citations
Intelligent safety management system for reducing damages of construction site
KR101985161B1
Apparatus and method for generating learning engine based on deep learing of LEGO type
KR1020200027096A
Smart safety-integration control system
KR1020220057132A
