Emergency bell management system using an interactive artificial intelligence model

Abstract

An emergency bell management system using a conversational artificial intelligence model is disclosed. The emergency bell management system using a conversational artificial intelligence model may include: a plurality of emergency bell modules installed at different locations, each equipped with a microphone capable of receiving a user's voice and a speaker capable of outputting voice; and a control server that receives the user's voice signal and location type information of the emergency bell module from each of the emergency bell modules, converts the user's voice signal into text, and transmits response data for the user's voice signal generated through a conversational artificial intelligence model using the text and the location type information as input data to the emergency bell modules.

Description

Technology Field

[0001] The present invention relates to an emergency bell management system using an interactive artificial intelligence model, and more specifically, to an emergency bell management system capable of providing accurate answers to questions or requests made by users or performing appropriate responses depending on the location where the emergency bell is installed. Background Technology

[0002] Fourth Industrial Revolution technologies and artificial intelligence are bringing about significant changes across industry, society, and the economy. In particular, conversational AI models, including ChatGPT (Generative Pre-trained Transformer), are providing customized solutions for human industries and daily life. I believe that the development of nations and societies driven by the Fourth Industrial Revolution will vary depending on how AI is utilized.

[0003] Conversational AI models provide intelligent systems capable of performing tasks that typically require human intelligence, such as problem-solving, decision-making, and learning. The Transformer model used in ChatGPT is one of the most popular deep learning models in the field of natural language processing. Unlike existing models such as RNNs (Recurrent Neural Networks) and LSTMs (Long Short Term Memory), it utilizes an attention mechanism and a parallel computation structure employing matrix multiplication to provide faster and more accurate results.

[0004] Currently developed AI STT (Speech to Text) models for natural language processing have an accuracy of 85% in Korean recognition. AI STT models apply a method of collecting large-scale data via the internet and automatically refining and learning from it.

[0005] Meanwhile, local governments in South Korea are installing and operating emergency bells throughout the country to ensure the safety of residents. However, responding to urgent violent crimes in real time is difficult. Furthermore, it is challenging to address issues such as personal and psychological counseling during emergencies.

[0006] Therefore, it is necessary to research alternatives for situations such as emergency dispatch by applying artificial intelligence technology to emergency bells operated by local governments to provide real-time responses to urgent violent crimes or cases requiring psychological intervention, such as counseling.

[0007] When a resident of a local government presses an emergency bell and verbally transmits the emergency situation, there is a need to develop a system in which a conversational AI model responds verbally on behalf of the CCTV control center operator before the operator does, and an AI STT model converts the speech into text to coordinate with emergency response services 119 and 112. Additionally, since emergency bells are used only during emergencies, their utilization is very low, so new measures are required to increase their utility during normal times. The problem to be solved

[0008] According to the present invention, an emergency bell management system using an interactive artificial intelligence model capable of providing more accurate responses to user safety and requests is provided.

[0009] In addition, according to the present invention, an emergency bell management system using an artificial intelligence model capable of rapidly transmitting an emergency situation occurrence signal to an emergency response agency is provided. means of solving the problem

[0010] An emergency bell management system using an interactive artificial intelligence model according to an embodiment of the present invention may include: a plurality of emergency bell modules installed at different locations, each equipped with a microphone capable of receiving a user's voice and a speaker capable of outputting voice; and a control server that receives the user's voice signal and location type information of the emergency bell module from each of the emergency bell modules, converts the user's voice signal into text, and transmits response data for the user's voice signal generated through an interactive artificial intelligence model that uses the text and the location type information as input data to the emergency bell modules.

[0011] In addition, the conversational artificial intelligence model can extract keywords from the text, vectorize previously stored data into sentence, paragraph, and document units based on the keywords, vectorize the location type information, and generate the answer data by connecting the vectorized data of the sentence, paragraph, and document units with the vectorized data of the location type information.

[0012] In addition, the conversational artificial intelligence model determines whether there is an emergency situation using the keyword, the user's voice signal, and the location type information as input data, and the control server can transmit an emergency situation occurrence signal to an emergency response agency when an emergency situation is determined.

[0013] In addition, the above location type information may include any one of park identification information, school identification information, bus stop identification information, hiking trail identification information, dense area alley identification information, tourist attraction identification information, and restroom identification information.

[0014] In addition, the above location type information may include location identification information of the point where the emergency bell module is installed within a certain area. Effects of the invention

[0015] According to the present invention, more accurate answer data can be generated by using a conversational artificial intelligence model that uses a user's voice signal and location type information of an emergency bell module as input data.

[0016] In addition, according to the present invention, the user's psychological state is determined to be in a state requiring emergency through the user's voice signal, keywords extracted therefrom, and location type information of the emergency bell module, and if it is determined to be in a state requiring emergency, the emergency situation occurrence signal and the location information of the emergency bell module can be transmitted to the emergency situation response agency (20). Brief explanation of the drawing

[0017] FIG. 1 is a drawing showing an emergency bell management system using an artificial intelligence model according to an embodiment of the present invention. FIG. 2 is a drawing showing an emergency bell module according to an embodiment of the present invention. FIG. 3 is a diagram showing the configuration of a control server according to an embodiment of the present invention. FIG. 4 is a structural diagram showing an interactive artificial intelligence model according to an embodiment of the present invention. FIG. 5 is a flowchart illustrating an example of operation of an emergency bell management system according to one embodiment of the present invention. Specific details for implementing the invention

[0018] Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the attached drawings. However, the technical concept of the present invention is not limited to the embodiments described herein and may be embodied in other forms. Rather, the embodiments introduced herein are provided to ensure that the disclosed content is thorough and complete and to ensure that the concept of the present invention is sufficiently conveyed to those skilled in the art.

[0019] In this specification, when a component is described as being on another component, it means that it may be formed directly on the other component or that a third component may be interposed between them. Additionally, in the drawings, the thicknesses of the films and regions are exaggerated for the effective description of the technical content.

[0020] Additionally, although terms such as first, second, third, etc., have been used to describe various components in the various embodiments of this specification, these components should not be limited by such terms. These terms are used merely to distinguish one component from another. Accordingly, what is referred to as the first component in one embodiment may be referred to as the second component in another embodiment. Each embodiment described and illustrated herein also includes its complementary embodiment. Furthermore, in this specification, "and / or" is used to mean including at least one of the components listed before and after it.

[0021] In the specification, singular expressions include plural expressions unless the context clearly indicates otherwise. Furthermore, terms such as "include" or "have" are intended to specify the existence of the features, numbers, steps, components, or combinations thereof described in the specification, and should not be understood as excluding the existence or addition of one or more other features, numbers, steps, components, or combinations thereof. Additionally, in this specification, "connection" is used to include both indirectly connecting multiple components and directly connecting them.

[0022] Furthermore, in describing the present invention below, if it is determined that a detailed description of related known functions or configurations could unnecessarily obscure the essence of the invention, such detailed description will be omitted.

[0024] FIG. 1 is a drawing showing an emergency bell management system using an artificial intelligence model according to an embodiment of the present invention.

[0025] Referring to FIG. 1, the emergency bell management system (10) using an artificial intelligence model provides an emergency bell management system that can provide accurate answers to questions or requests made by users or perform appropriate responses depending on the location where the emergency bell is installed.

[0026] An emergency bell management system (10) using an artificial intelligence model includes an emergency bell module (100), a repeater (200), and a control server (300).

[0027] The emergency bell module (100) is installed in various locations and can be used by users to communicate with the facility control center or to report an emergency situation. Additionally, the emergency bell module (100) can be used to search for various information using the user's voice signal.

[0028] Multiple emergency bell modules (100) are provided, and each can be installed in different zones (100a, 100b, 100c). According to an embodiment, the emergency bell modules (100) can be installed in parks, schools, bus stops, hiking trails, alleys in densely populated urban areas, tourist attractions, public restrooms, etc. When the operator of the emergency bell management system (10) is a local government, multiple of the aforementioned management zones (100a, 100b, 100c) are located in the management area of ​​the said local government, and an emergency bell module (100) is individually installed in each of the management zones (100a, 100b, 100c).

[0029] Additionally, multiple emergency bell modules may be individually installed in different locations within a single management area (100a, 100b, 100c). For example, if the management area (100a) is a park, walking paths, public restrooms, exercise facilities, public facilities, etc. may be provided within the park, and emergency bell modules (100) are individually installed in these locations. If the management area (100b) is a building, emergency bell modules (100) may be individually installed in floor corridors, floor staircases, elevators, floor restrooms, underground parking lots, etc.

[0030] The emergency bell module (100) includes location type information according to the management area and installation location. Location type information is information regarding the installation location, rather than location information (e.g., latitude and longitude) of the emergency bell module (100), and is related to the use of the facility.

[0031] According to one example, an emergency bell module (100) installed in a public restroom may generate an emergency bell event of a different type than an emergency bell module (100) installed in a walking path, exercise facility, underground parking lot, etc. For example, if there is no toilet paper in the restroom, the restroom door is locked, or the cleanliness of the restroom is poor, a user may generate an emergency bell event through the emergency bell module (100).

[0032] In the case of walking paths, unlike other places, emergency bell events may be triggered to inquire about the path, report malfunctions in surrounding facilities such as streetlights, or indicate the presence of dangerous objects on the path.

[0033] As such, since the intended use of the emergency bell module (100) varies depending on the management area and installation location where the emergency bell module (100) is installed, each emergency bell module (100) is assigned location type information suitable for its intended use.

[0035] FIG. 2 is a drawing showing an emergency bell module according to an embodiment of the present invention.

[0036] Referring to FIG. 2, the emergency bell module (100) can be installed on a pillar or a building wall. The emergency bell module (100) includes an emergency button (110), a microphone (120), a speaker (130), a warning alarm (140), a media board (150), a surveillance camera (160), and a communication unit (170).

[0037] The emergency button (110) is a button that the user can press with their hand, and when the user presses the emergency button (100), the microphone (120) and speaker (130) are activated.

[0038] The microphone (120) receives the user's voice signal.

[0039] The speaker (130) outputs the response data transmitted from the control server (300) as voice.

[0040] The warning alarm (140) may be activated when the user presses the emergency button (110) or by a control signal transmitted from the control server (300). The warning alarm (140) may output a warning sound or a warning light to the outside.

[0041] The media board (150) is provided as a display capable of displaying various information. The media board (150) can be kept active at all times and can display local event information, advertising information, etc. When the user presses the emergency button (110), the screen switches and information received from the control server (300) can be displayed on the screen. Additionally, when the user converses with the manager of the control facility through the microphone (120) and speaker (130), the manager's video information can be displayed on the media board (150).

[0042] The surveillance camera (160) films the area around the point where the emergency bell module (100) is installed. The filming range of the surveillance camera (160) may include users using the emergency bell module (100).

[0043] The communication unit (170) transmits the user's voice signal and location type information received by the microphone (120) to the repeater (200) and receives the response data and consultation data transmitted from the control server (300). The communication unit (170) can be connected to the repeater (200) via wired or wireless connection.

[0044] Referring again to FIG. 1, the repeater (200) connects the communication unit (170) and the control server (300). The repeater (200) can be connected to a plurality of emergency bell modules (100). According to an embodiment, a plurality of emergency bell modules (100) are installed in each management area (100a, 100b, 100c), and the repeater (200) can be connected to them. Additionally, the emergency bell modules (100) can be managed in a plurality of groups, and the repeater (200) can be provided for each group.

[0045] The control server (300) receives the user's voice signal from each of the emergency bell modules (100), video information captured by the surveillance camera (160), and location type information of the emergency bell module (100) through the relay (200), and converts the user's voice signal into text. Then, using a conversational artificial intelligence model that takes the text and the location type information of the emergency bell module (100) as input data, it generates response data and consultation data for the user's voice signal. Then, it transmits the generated response data and consultation data to the relay (200).

[0046] FIG. 3 is a diagram showing the configuration of a control server according to an embodiment of the present invention, and FIG. 4 is a structural diagram showing an interactive artificial intelligence model according to an embodiment of the present invention.

[0047] Referring to FIGS. 3 and 4, the control server (300) includes a data communication unit (310) and a data processing unit (320).

[0048] The data communication unit (310) communicates with the repeater (200) via wired and wireless means.

[0049] The data processing unit (320) converts the user's voice signal into text, generates response data for the user's voice signal using a conversational artificial intelligence model that uses text and location type information of the emergency bell module (100) as input data, and converts the response data into a voice signal.

[0050] The conversational AI model extracts keywords from text and vectorizes the stored data into sentences, paragraphs, and documents based on the keywords. The stored data refers to document data containing various situations that may occur in the management areas (100a, 100b, 100c) and corresponding information, and each document contains text.

[0051] Conversational AI models vectorize sentences, paragraphs, contexts, and documents contained in document data in a direction that approaches keywords. To this end, conversational AI models can use the Doc2Vec (Document Embedding with Paragraph Vectors) algorithm. Doc2Vec is an embedding method that extends Word2Vec and treats a Document ID as a word appearing in all contexts. In other words, the vector corresponding to a Document ID moves in a direction that approaches every word appearing in that document. Therefore, even if the words appearing in different documents are different, the word vectors become similar.

[0052] Additionally, the conversational AI model vectorizes the aforementioned location type information. Due to the vectorization of the location information type, data is searched in a direction that is closer to the purpose of the facility where the emergency bell module is installed. To this end, the conversational AI model may use the Loc2Vec (Location Embedding with paragraph Vectors) algorithm. Loc2Vec can search for data by vectorizing the location type information of the emergency bell module (100) to match the purpose for which the local government installed the emergency bell module (100).

[0053] The conversational AI model generates response data for the user's voice signal by combining words, sentences, paragraphs, and context contained in document data extracted in a direction closer to keywords with words, sentences, paragraphs, and context contained in document data extracted in a direction closer to the purpose of the facility where the emergency bell module is installed. As a result, the conversational AI model can generate more accurate response data regarding the user's safety and requests.

[0054] In addition, the conversational AI model can generate consultation data. The conversational AI model can generate consultation data by combining words, sentences, paragraphs, and context contained in document data extracted in a direction closer to keywords, and words, sentences, paragraphs, and context contained in document data extracted in a direction closer to the purpose of the facility where the emergency bell module is installed.

[0055] Conversational AI models can maximize rewards to improve the accuracy of generated answer data. To this end, the mathematical framework Markov Decision Process (MDP) can be used. MDP is an algorithm focused on enhancing the value of the reward function, and as shown in Equation 1 below, it consists of a state (S), action (A), state transition probability (P), reward function (R), and depreciation rate (γ), and is represented as the sum of rewards received considering the environment.

[0056] [Formula 1]

[0057] G t =R t+1 +γR t+2 +γ 2 R t+3 +γ 3 R t+4 ...

[0058] Pr{R t+1 =r, S t+1 =s'|S0,A0,R1,..., S t-1 , A t-1 ,Rt ,S t ,A t}

[0059] Here, MDP<S, A, P, R, γ> It finds the optimal policy through cumulative rewards, and by considering the characteristics that always occur at a fixed location, it can obtain the optimal result through rewards.

[0060] The present invention can achieve the effect of performance improvement by increasing the value of the compensation function (R) through the addition of a variable called location type information.

[0061] In addition, the conversational AI model determines whether an emergency situation exists by using keywords, the user's voice signal, and location type information as input data.

[0062] According to an embodiment, the conversational AI model generates a compressed signal by compressing the user's voice signal and text, respectively, to different sizes. A Convolutional Neural Network (CNN) may be used for generating the compressed signal.

[0063] A convolutional neural network analyzes speech signals to train a speech recognition model, converts the trained data into a vector form using convolution operations, and generates a feature map by applying multiple filters to the converted vector data. The size of the generated feature map is reduced using pooling, but by calculating the representative values ​​of the corresponding pixels, the effects such as changes in size, warping, and distortion of the feature map are minimized.

[0064] Pooling reduces the size of the convolutional neural network by shrinking the pixels of the feature map into a single representative value, thereby reducing the horizontal and vertical spatial dimensions. Pooling is divided into max pooling and average pooling depending on how the representative value is set; max pooling sets the maximum value among the pixels of the feature map as the representative value, while average pooling sets the average value of the feature map pixels as the representative value.

[0065] A conversational AI model can adjust the degree of compression of a user's voice signal and text by adjusting the number of pooling cycles of a convolutional neural network, and the voice signal and text can be reduced to 1 / 2, 1 / 4, 1 / 8, or 1 / 16 each time pooling is applied 1, 2, 3, or 4 times. Either max pooling or average pooling may be applied.

[0066] Conversational AI models extract feature points by applying weights to compressed signals. Specifically, conversational AI models calculate attention corresponding to the compressed speech signal and the compressed text signal, and extract feature points by assigning different weights to each compressed signal based on the said attention. In this case, a higher weight may be assigned to the compressed speech signal than to the compressed text signal.

[0067] The conversational AI model determines the user's psychological state by assigning weights to the above feature points. According to an embodiment, the conversational AI model determines whether the user's psychological state is in a state requiring urgency. If the user's psychological state is determined to be in a state requiring urgency, the conversational AI model generates an emergency situation occurrence signal. The emergency situation occurrence signal is transmitted to an emergency response agency (20) along with location information of the emergency bell module (100) through the data communication unit (310).

[0068] When a user is injured or threatened by others, their voice tone, speed, and tremors differ from usual, and these states are derived as feature points through the aforementioned analysis of voice signals and text. Therefore, even if the user does not directly describe their condition, the conversational AI model can infer the user's state and perform an immediate response.

[0070] Hereinafter, the operation process of an emergency bell management system using an interactive artificial intelligence model according to an embodiment of the present invention will be described in detail.

[0071] When a user presses the emergency button (110) of the emergency bell module, the microphone (120) and speaker (130) are activated, and the media board (150) switches to conversation mode. Video information of the manager of the control facility can be displayed on the media board (150). When a user speaks into the microphone (120), the voice signal, video information captured by the surveillance camera (160), and location type information of the emergency bell module (100) are received by the data communication unit (310) of the control server (300) through the repeater (200).

[0072] The data processing unit (320) of the control server (300) converts the user's voice signal into text. The conversational AI model generates response data for the user's voice signal using the text and the location type information of the emergency bell module (100) as input data. Specifically, the conversational AI model extracts keywords from the text, vectorizes the previously stored data into sentences, paragraphs, and documents based on the keywords, and vectorizes the location type information. As a result, sentences, paragraphs, contexts, and documents included in the document data are searched in a direction closer to the keywords, and data is searched in a direction closer to the purpose of the facility where the emergency bell module is installed.

[0073] The conversational artificial intelligence model generates response data for the user's voice signal by combining the words, sentences, paragraphs, and context included in the document data extracted in a direction closer to the keyword, and the words, sentences, paragraphs, and context included in the document data extracted in a direction closer to the use of the facility where the emergency bell module (100) is installed.

[0074] In addition, conversational AI models maximize rewards to improve the accuracy of generated answer data, and can use the mathematical framework MDP for this purpose.

[0075] Through the process described above, the conversational artificial intelligence model can generate more accurate answer data. The answer data is converted into voice data and then transmitted to the communication unit (170) of the emergency bell module (100) via the data communication unit (310), and is output as voice through the speaker (130). Additionally, the answer data is transmitted to the communication unit (170) via the data communication unit (310) and is output to the media board (150) as text or an image.

[0076] In contrast, when a user requests response data as text, the response data may be displayed on the media board (150) in the form of text, web pages, images, or Link URLs. For example, when a user requests information about nearby facilities or weather information via voice signal, a web page or Link URL containing information about nearby facilities may be displayed on the media board (150), or text or a web page providing current weather information may be displayed on the media board (150).

[0077] Meanwhile, the conversational AI model analyzes the user's voice signal and words contained in the text to determine whether it is an emergency situation.

[0078] Conversational AI models utilize convolutional neural networks to generate compressed signals by compressing the user's voice signal and text, respectively, to different sizes, and extract feature points from each compressed signal. Conversational AI models can extract feature points by applying different weights to the compressed signals. In this case, a higher weight may be assigned to the compressed voice signal than to the compressed text signal.

[0079] The conversational AI model assigns weights to the above feature points to determine whether the user's psychological state is in a state requiring emergency. If the user's psychological state is determined to be in a state requiring emergency, the conversational AI model generates an emergency situation occurrence signal and transmits it to the emergency situation response agency (20) along with the location information of the emergency bell module through the data communication unit (310).

[0081] FIG. 5 is a flowchart illustrating an example of operation of an emergency bell management system according to one embodiment of the present invention.

[0082] Referring to FIG. 5, when a local government resident presses the emergency button (110) and says, "Please help me, a scary person is chasing me," the resident's voice signal and the location type information of the emergency bell module (100) are received by the data communication unit (310) of the control server (300) through the relay (200).

[0083] The data processing unit (320) of the control server (300) converts the user's voice signal into text and generates response data for the user's voice signal using a conversational artificial intelligence model. At the same time, it determines whether the user's psychological state is in a state requiring urgency and generates an emergency situation occurrence signal. In the case of the above resident, the conversational artificial intelligence model generates the response data "I will dispatch a 112 patrol car. Please wait without moving under the bright light under the emergency bell for a moment," modulates this into a voice signal, and outputs it through the speaker (130) of the emergency bell module (100). At the same time, it determines the user's psychological state as a state requiring urgency and transmits the emergency situation occurrence signal to the emergency situation response agency (20).

[0084] Additionally, the conversational AI model generates information about the current progress, modulates it into a voice signal, and transmits it to the emergency bell module (100). The speaker (130) of the emergency bell module (100) may output a voice saying, "A 112 patrol car is on its way. Please wait in a bright location near the emergency bell without moving for a moment."

[0085] The conversational AI model can provide counseling to ensure the psychological stability of residents while they wait for the patrol car to arrive. The conversational AI model generates counseling data such as, "Please describe the appearance, clothing color, and brief characteristics of the scary person," converts it into speech, and transmits it to the emergency bell module. The generation of counseling data can continue until the 119 patrol car and police arrive.

[0087] The emergency bell management system using the aforementioned conversational AI model uses the user's voice signal and the location type information of the emergency bell module as input data to generate response data and consultation data regarding the user's voice signal through the conversational AI model, and maximizes rewards using an MDP to improve the accuracy of the generated data. By iteratively training this process, accurate response data can be generated that is suitable for the installation location and purpose of the emergency bell module and can respond to various types of emergency bell events.

[0088] Through this, the emergency bell management system using an interactive artificial intelligence model according to an embodiment of the present invention can solve the problem of existing emergency bell management systems in which an administrator must directly respond to user requests because user requests vary depending on the installation location of the emergency bell module and the content of the emergency bell event differs.

[0089] Furthermore, the emergency bell management system using an interactive artificial intelligence model according to an embodiment of the present invention does not require the setting of a separate response manual after the installation of the emergency bell module. This can solve the problem of existing systems where an emergency response manual must be individually set according to the installation location after the installation of the emergency bell module.

[0091] Although the present invention has been described in detail using preferred embodiments, the scope of the invention is not limited to specific embodiments and should be interpreted by the appended claims. Furthermore, those skilled in the art will understand that many modifications and variations are possible without departing from the scope of the invention. Explanation of the symbols

[0092] 10: Emergency bell management system using an artificial intelligence model 100: Emergency bell module 110: Emergency button 120: Microphone 130: Speaker 140: Warning alarm 150: Media Board 160: Surveillance camera 170: Communications Department 200: Repeater 300: Control Server

Claims

Claim 1 A plurality of emergency bell modules installed at different locations, each equipped with a microphone capable of receiving a user's voice and a speaker capable of outputting voice; and a control server that receives the user's voice signal and location type information of the emergency bell module from each of the emergency bell modules, converts the user's voice signal into text, and transmits response data for the user's voice signal generated through a conversational artificial intelligence model using the text and the location type information as input data to the emergency bell modules, wherein the conversational artificial intelligence model is an MDP<S, A, P, R, γ> Compensation is performed using, wherein the state (S), action (A), state transition probability (P), reward function (R), and depreciation rate (γ) are composed of the following formula, [Formula] G t =R t+1 +γR t+2 +γ 2 R t+3 +γ 3 R t+4 ...Pr{R t+1 =r, S t+1 =s'|S0,A0,R1,..., S t-1 , A t-1 ,R t ,S t ,A t The above MDP<S, A, P, R, γ> An emergency bell management system using an interactive artificial intelligence model that finds an optimal policy through cumulative rewards and increases the value of the reward function (R) by adding the location type information as a variable. Claim 2 An emergency bell management system using an AI model according to claim 1, wherein the conversational AI model extracts keywords from the text, vectorizes previously stored data into sentences, paragraphs, and documents based on the keywords, vectorizes location type information, and generates answer data by connecting the vectorized data of the sentences, paragraphs, and documents with the vectorized data of the location type information. Claim 3 In claim 2, the conversational artificial intelligence model determines whether an emergency situation exists by using the keyword, the user's voice signal, and the location type information as input data, and the control server transmits an emergency situation occurrence signal to an emergency response agency when an emergency situation is determined, an emergency bell management system using an artificial intelligence model. Claim 4 In claim 1, the location type information is an emergency bell management system using an artificial intelligence model that includes any one of park identification information, school identification information, bus stop identification information, hiking trail identification information, densely populated alley identification information, tourist attraction identification information, and restroom identification information. Claim 5 In claim 1, the location type information is an emergency bell management system using an artificial intelligence model that includes location identification information of a point where the emergency bell module is installed within a certain area.

Images