Smart Banner System Having Intelligent Disaster Safety Control and Self-Protection Functions and Method for Controlling the Same

Abstract

The present invention includes a communication unit that collects disaster or public interest events generated by analyzing and processing data received from an external server through an AI cloud server, a data processing unit that normalizes data into a standard schema and performs hash-based duplicate filtering, a priority calculation unit that determines content interrupts and layout variations according to a priority score, and a content control unit that displays three-dimensional disaster information in conjunction with a camera module and a speaker module. In particular, the present invention relates to an intelligent smart display stand system equipped with a self-protection function that optimizes the operating environment of the device itself by including a sensor unit that detects the six major meteorological elements (temperature, relative humidity, atmospheric pressure, wind direction, wind speed, and solar radiation) and illuminance, and an environment control unit that drives a heater and a fan inside the enclosure. Through this, public safety information is rapidly and three-dimensionally disseminated, while simultaneously significantly improving the durability of the device in extreme weather environments and quantitatively compensating for the advertiser's losses.

Description

Technology Field

[0001] The present invention relates to an AI smart display stand system and a control method thereof, and more specifically, to an intelligent smart display stand system equipped with an interrupt function that analyzes and processes information received from an external public data server through an AI cloud server based on Internet of Things (IoT) technology to generate real-time disaster and emergency safety events, and dynamically controls content being transmitted according to the priority calculation result based on a unique algorithm.

[0002] In addition, the present invention relates to a self-protection technology that actively manages power control and the internal environment (heater, fan) of an enclosure for device protection by real-time sensing of six meteorological elements—temperature, relative humidity, atmospheric pressure, wind direction, wind speed, and solar radiation—and illuminance data, and to a three-dimensional disaster response technology that simultaneously performs surrounding situation monitoring and voice guidance by linking a camera module and a speaker module in the event of a disaster.

[0003] Furthermore, the present invention relates to a compliance filtering technology that detects regulatory items under the Outdoor Advertising Act and related administrative guidelines in real time through the analysis of image and text data, and an intelligent integrated operation platform technology that realizes service value in line with the ESG management environment by quantifying advertising exposure value based on pedestrian traffic and environmental data and intelligently scheduling to replenish lost advertising time. Background Technology

[0004] With the recent implementation of smart cities and the acceleration of digital transformation, the outdoor advertising industry is actively transitioning from traditional analog methods to smart display stands utilizing digital signage.

[0005] Traditional banner stands have operated by manufacturing PVC banners and having workers at heights manually install and replace them. However, this method exposes significant limitations from the perspective of carbon neutrality and ESG management. Not only is there a constant risk of safety accidents, such as falls, during installation and removal, but the production and disposal processes also generate large amounts of plastic waste, which releases serious environmental pollutants like dioxins and microplastics upon incineration. In particular, despite the need to rapidly deliver safety information to citizens during national disasters or local emergencies, the analog method poses a critical weakness in securing the "golden time" for disaster response because real-time information updates are physically impossible.

[0006] Digital display boards introduced in the early stages to address these issues also remain in a passive operational mode, primarily relying on the simple sequential playback of pre-stored advertising content. In other words, they suffer from limitations, such as a lack of interrupt logic to flexibly control content delivery based on situational urgency by analyzing data received in real-time from public APIs of the Ministry of the Interior and Safety or local governments, as well as insufficient technical means to actively comply with the brightness and flashing rate regulations under the Outdoor Advertising Act.

[0007] A more serious technical bottleneck lies in the 'environmental durability and operational stability' of outdoor digital display boards. Existing devices suffer from chronic issues where the lifespan of precision components, such as LED modules and control units, is drastically shortened or they malfunction when exposed to internal enclosure temperature rises due to direct sunlight, high humidity in summer and condensation in winter, or strong winds like typhoons. Furthermore, as they are limited to merely conveying visual information, they struggle to deliver three-dimensional safety information to the visually impaired or pedestrians, and they lack the functionality of an integrated safety hub that allows managers to remotely monitor the surroundings of the display board in real time.

[0008] Therefore, there is an urgent need to develop an advanced intelligent smart display stand system capable of three-dimensional disaster response by combining camera modules and voice guidance, while ensuring data reliability through precise analysis of external public data via AI cloud servers and self-protecting the device by sensing ever-changing weather conditions (wind speed, solar radiation, etc.) in real time. The problem to be solved

[0009] The technical problem that the present invention aims to solve is to provide an intelligent smart display stand system and a control method thereof that can protect the main body of the device by actively responding to changes in the weather environment while realizing optimal content display control according to the urgency of each situation by intelligently standardizing unstructured disaster and public interest data collected from the outside through an AI cloud server and ensuring data reliability.

[0010] More specifically, the first problem that the present invention aims to solve is to provide technical means to prevent computational load on the system and maximize the accuracy of data processing by normalizing heterogeneous events collected from various external servers into a Standard Event Schema and performing duplicate filtering using a hash algorithm.

[0011] The second task is to provide a three-dimensional safety infrastructure capable of securing the "golden time" for responding to national disasters by applying a unique priority calculation algorithm based on objective indicators such as disaster severity, proximity to the site, and the timeliness of information to actively control (interrupt) currently playing content or dynamically vary the screen layout, while simultaneously eliminating blind spots in visual information and monitoring the surrounding situation in real time by integrating camera and speaker modules.

[0012] The third task is to significantly improve the durability of outdoor installed equipment and increase maintenance efficiency by realizing self-protection logic that controls display power or drives heaters and fans inside the enclosure in extreme weather conditions such as strong winds or overheating by sensing the six major weather elements of temperature, relative humidity, atmospheric pressure, wind direction, wind speed, and solar radiation in real time.

[0013] The fourth task aims to provide an innovative operational platform that automatically determines compliance by monitoring content brightness and flashing rates restricted by relevant regulations, such as the Outdoor Advertising Act, in real time, and intelligently compensates for lost advertising time by quantifying the value of advertising exposure through integration with surrounding pedestrian traffic data. means of solving the problem

[0014] An AI smart display stand system according to an embodiment of the present invention for solving the above problem comprises: a communication unit that receives in real time event data including disaster or public interest information generated by AI analysis processing of data received from an external server on a cloud server; a data processing unit that analyzes the received data and normalizes it into a Standard Event Schema in which disaster grade and occurrence information are defined; a priority calculation unit that determines whether there is an interrupt and the display priority of content currently being played based on the normalized data; and a content control unit that controls the display of the event data by varying the screen layout of a display device according to the determination of the priority calculation unit, and simultaneously controls the capture of surrounding conditions through a camera module and the output of voice guidance through a speaker module when disaster information is displayed.

[0015] According to one feature of the present invention, the data processing unit receives normalized standard event schema data from a cloud server (400), inputs the source information and publication time information of the data from a local terminal into a hash function to generate a unique identification value, and is configured to prevent unnecessary computational load on the system by filtering duplicate received data in real time based on the identification value. Through this, a division of roles is achieved in which high-computational tasks such as data normalization are performed on the cloud server, and lightweight tasks such as hash-based duplicate verification are processed by an edge device inside the display stand. The priority calculation unit [determines] the severity of the disaster ( ), Proximity to the place of origin( ), timeliness of information( ) and the credibility of the information issuing institution( By applying a calculation formula that includes ) as an independent variable, the priority score ( It calculates ) and is characterized by the threshold value, which is the criterion for determining whether to interrupt, being dynamically varied according to the population density by time period measured through the camera sensor.

[0016] According to another feature of the present invention, the communication unit performs an edge response function that retrieves and maintains data within a valid timestamp cached in a local storage when a network failure with an external server occurs. The content control unit selects and executes one of a full-screen occupancy mode, an always-exposed area overlay mode through screen splitting, and a forced insertion mode between existing content slots according to a priority score, and in particular, automatically generates a supplementary transmission schedule by reflecting the mobile population weight at that time in the packet of non-exposure time of the advertisement content interrupted by an interrupt.

[0017] In addition, the present invention is equipped with a filtering module that automatically determines compliance with regulations by detecting brightness information and flicker rate of video content in real time. Furthermore, the system includes six weather sensors and an illuminance sensor that measure temperature, relative humidity, atmospheric pressure, wind direction, wind speed, and solar radiation, and performs self-protection logic that controls display power or transmits a protection screen when the collected wind speed or temperature data exceeds a device protection threshold. In particular, it features an intelligent feedback control mechanism that optimizes operational stability in an outdoor environment by actively controlling the prevention of condensation and heat exhaust inside the enclosure through an environment control unit including a thermometer, a heater, a fan, and a controller installed inside the enclosure.

[0018] An intelligent content control method performed by a smart display stand system according to an embodiment of the present invention comprises: a communication unit receiving disaster or public interest event data generated by AI analysis processing of data received from an external server in real time on a cloud server; a data processing unit performing hash function-based duplicate filtering on normalized event data received from the cloud server; a priority calculation unit calculating a priority score through a calculation formula including the severity of the disaster, proximity to the source of occurrence, the timeliness of information, and the credibility of the information issuing agency as independent variables; a priority calculation unit determining whether to interrupt the content currently being played by comparing the priority score with a preset threshold; and a content control unit displaying the event data by varying the screen layout of a display device according to the interrupt determination, wherein when displaying disaster information, the camera module captures the surrounding situation and the speaker module simultaneously controls voice guidance output. Effects of the invention

[0019] According to the AI ​​smart display stand system equipped with intelligent disaster safety control and self-protection functions and the control method thereof according to the present invention, the following effects can be expected.

[0020] First, in the event of a national disaster or emergency, optimal safety information can be transmitted within seconds by linking in real-time with precise analysis data from an AI cloud server, thereby significantly securing the golden time for disaster response. In particular, by performing real-time monitoring via a camera module and voice guidance via a speaker module simultaneously with automated interrupt control, it eliminates blind spots in information delivery and has the effect of establishing a three-dimensional safety infrastructure that protects the lives and property of citizens.

[0021] Second, there is a technical advantage in that it can maximize the durability of outdoor installed equipment and drastically reduce maintenance costs through a self-protection logic that, based on data from six major weather sensors measuring temperature, relative humidity, atmospheric pressure, wind direction, wind speed, and solar radiation, the device itself executes a protection mode in extreme weather conditions such as strong winds or heat waves, and drives the heater and fan inside the enclosure to prevent condensation and overheating.

[0022] Third, by combining hash function-based data filtering technology and edge computing technology, redundant information among a large amount of heterogeneous data is removed in real time, thereby minimizing system load and enabling the continuous display of valid data stored in the local cache even in the event of a network failure, thus ensuring high reliability and continuity as a disaster management system.

[0023] Fourth, by performing intelligent compensation scheduling that reflects time-based pedestrian traffic weighting for advertising losses that are interrupted during commercial advertising broadcasts, reasonable and quantitative compensation can be provided to advertisers, thereby simultaneously optimizing the economic efficiency and reliability of system operations.

[0024] Fifth, by fundamentally blocking the emission of plastic waste and hazardous substances generated during the production and disposal of existing analog banners, it is possible to realize carbon reduction and ESG management values. Furthermore, it has the effect of simultaneously enhancing the safety and efficiency of administrative operations by fundamentally preventing safety accidents caused by working at heights during manual banner replacement and automatically determining compliance with the Outdoor Advertising Act. Brief explanation of the drawing

[0025] FIG. 1 is a front perspective view of an AI smart display stand device equipped with a self-protection and multimedia safety module according to one embodiment of the present invention. FIG. 2 is a rear perspective view of an AI smart display stand device according to one embodiment of the present invention. FIG. 3 is a schematic diagram of the entire system linked with an AI cloud analysis server and a central management system according to one embodiment of the present invention. FIG. 4 is a flowchart illustrating a content control and interrupt algorithm including a weather threshold-based self-protection logic according to an embodiment of the present invention. FIG. 5 is a priority score according to an embodiment of the present invention ( It is a dynamic layout variable state diagram of a display based on ) and multimedia safety response. FIG. 6 is a block diagram illustrating a feedback loop for internal enclosure environment control and brightness adjustment based on sensing of six major weather elements according to one embodiment of the present invention. Specific details for implementing the invention

[0026] Hereinafter, an AI smart display stand system equipped with intelligent disaster safety control and self-protection functions and a control method thereof according to a preferred embodiment of the present invention will be described in detail with reference to the attached drawings. The embodiments of the present invention are described in technical detail along with specific component reference numerals so that those skilled in the art can easily implement them.

[0027] Referring to FIGS. 1 and 2, the physical configuration and installation structure of a smart display stand system (100) according to one embodiment of the present invention is a device installed in public places, roadsides, intersections, etc., to visually display digital content. In terms of hardware, it is largely composed of a display module (110) on which digital content is projected, a frame (120) which is an enclosure structure that stably holds the display module (110) and protects it from external impact, and a support member (130) that supports the frame (120) at a certain height from the ground.

[0028] The above display module (110) is composed of a high-brightness LED panel or an LCD panel to ensure high visibility even during the day or night when direct sunlight is strong, and by adopting a variable frame structure, it has the flexibility to operate as a multi-stage screen that can be freely divided from 1 to 7 stages in accordance with existing banner specifications.

[0029] As can be seen in the rear perspective view of FIG. 2, a plurality of display inspection doors (115) for maintenance are arranged in a grid pattern on the back of the frame (120) to provide access to internal components, and the support member (130) can be transformed into a single central column or a plurality of support structures depending on the installation environment and the weight of the display, and various variations of materials can be applied, such as not only corrosion-resistant steel but also aluminum alloy or reinforced plastic materials for lightweighting.

[0030] In particular, a sensor unit (140) is positioned at the top of the device and includes a camera module (150) for capturing surrounding conditions in real time, an illuminance sensor (143) for detecting ambient illuminance, and six weather sensors (141) for measuring temperature, relative humidity, atmospheric pressure, wind direction, wind speed, and solar radiation. Additionally, a speaker (160) for voice guidance is installed on the front of the support unit (130) to provide three-dimensional safety information that eliminates blind spots in visual information.

[0031] Inside the device, an internal control unit (not shown) for optimizing the operating environment is built in to actively control a heater (not shown) for preventing condensation and a cooling fan (not shown) for discharging internal heat, and an embedded control box is built in to function as a single independent edge device.

[0032] The above heater (not shown) may incorporate an overheating prevention function by adopting, for example, a self-regulating heating element of the PTC (Positive Temperature Coefficient) type, and when the internal temperature of the enclosure reaches the set temperature (e.g., 5) in winter or high-humidity environments It is automatically activated when the relative humidity drops below a certain level or rises above a critical threshold (e.g., 85%), thereby fundamentally blocking condensation.

[0033] The above fan (not shown) is linked with a temperature sensor so that when the internal temperature of the enclosure reaches an upper threshold (e.g., 60°C) in summer or high-temperature environments When the value exceeds ), it switches to a high-speed rotation mode to rapidly expel internal heat to the outside, and is designed to enable low-noise operation through speed control using Pulse Width Modulation (PWM). These environmental control elements are continuously monitored and controlled 24 hours a day by an independent microcontroller (MCU), and minimize component degradation and maximize energy efficiency by performing a predictive control algorithm in conjunction with real-time weather data collected from the sensor unit (140).

[0034] With reference to FIG. 3, the system architecture and network linkage system of the present invention will be described in detail. The present system forms a three-layer structure in which an external server group (300) which is a source of data, a cloud server (400) which intelligently processes collected data, a central management system (CMS, 200) which manages the entire infrastructure, and a smart display system (100) installed at an actual location are organically combined.

[0035] The above external server group (300) includes a public API (310), a disaster safety API (320), and a local government API (330), and the received unstructured source data is transmitted to the CMS (200) in the form of refined events through an AI analysis processing node (410) within the cloud server (400).

[0036] The core patentability of the present invention lies in the optimized division of roles and organic combination of each layer in this three-layer structure. Specifically, the data received from the external server group (300) is unstructured data in various formats such as JSON and XML, and there is a problem in that each issuing agency uses different schema structures and data field names. For example, the Ministry of the Interior and Safety's disaster text broadcasting API (320) transmits the disaster level with the field name "disaster_level," whereas a specific local government API (330) may use the field name "severity," and the severity level is also inconsistent, such as being expressed as levels 1 to 5, grades A to E, or text such as "urgent" or "caution." If such heterogeneous data is transmitted as is to the display board system (100), a critical problem arises in which the computational load of the edge device increases rapidly and real-time response is delayed because each display board must be equipped with complex parsing logic.

[0037] To solve this, the present invention automatically normalizes all heterogeneous data into a Standard Event Schema by applying Natural Language Processing (NLP) technology and a machine learning-based Schema Mapping algorithm at the AI ​​analysis processing node (410) of the cloud server (400).

[0038] The above standard event schema includes essential fields such as disaster type, severity, latitude / longitude of the location of occurrence, timestamp, authority, and expiry, and each field is standardized into a normalized format according to international standards (ISO 8601, etc.). For example, severity is converted into a normalized real value between 0.0 and 1.0, the location of occurrence is standardized into a latitude / longitude pair based on the WGS84 coordinate system, and the time information is standardized into the ISO 8601 format based on UTC (e.g., 2026-02-11T14:30:00Z).

[0039] By processing this normalization process in batches on a cloud server (400), hundreds or thousands of display terminals can share and utilize the same high-performance AI analysis resources, which produces an economic effect that improves cost-efficiency by more than tens of times compared to individually equipping each terminal with AI processing capabilities.

[0040] Furthermore, the AI ​​analysis processing node (410) is distinguished in that it performs a credibility assessment function of the received text information, going beyond simply converting the data format.

[0041] For example, when false disaster information or exaggerated rumors propagated from social media or unofficial channels are introduced, the node (410) applies a multi-layered credibility evaluation algorithm, including checking the source domain of the information, querying the credibility database of the issuing organization, cross-validating multiple sources regarding the same event, and detecting exaggerated expressions through sentiment analysis of the text content. Through this, a credibility score is calculated, which is directly utilized as a credibility variable in the priority calculation unit described later, thereby simultaneously achieving the dual effects of automatic filtering of false information and priority display of genuine emergency situations. This AI-based credibility verification mechanism holds significant technical value in the field of disaster safety, given that the accuracy of information is directly linked to life.

[0042] The communication unit (not shown) equipped in the smart display system (100) performs independent computation and data caching in a local area based on edge computing technology, thereby reducing data processing delay time and performing a safe mode function that maintains display by immediately retrieving valid data stored in a local storage even in the event of a network failure. Specifically, the communication unit ensures high reliability by adopting a redundant network configuration including a wireless communication module such as 5G or LTE and a wired Ethernet interface, and is responsible for real-time bidirectional communication with a cloud server (400) and a central management system (200). In addition, the communication unit detects transmission errors by applying a hash-based checksum mechanism to verify the integrity of received event data, and guarantees the security of public safety information by maintaining a secure data transmission path encrypted through the TLS (Transport Layer Security) protocol.

[0043] As described above, the data processing unit (not shown) within the display system (100), which receives normalized standard event data from the cloud server (400), is responsible for hash function-based duplicate filtering, which is a lightweight operation that can be performed on a local terminal. This is intended to handle cases where the same disaster event is received redundantly through multiple paths (e.g., the Ministry of the Interior and Safety and local governments simultaneously), or where the same data packet is retransmitted due to network instability. Specifically, the data processing unit inputs a string formed by concatenating the source information (e.g., "ministry_of_interior", "seoul_city") and the publication time information (e.g., "2026-02-11T14:30:00Z") of the received event data into a hash function such as SHA-256 or MD5 to generate a unique identification value (e.g., "a3f5c9d2e8b1..."). The above identifier is compared with a list of the last N event identifiers (e.g., 1,000 processed within the last hour) stored in local memory, and if a matching value is found, the data is determined to be a duplicate and immediately discarded. This hash-based duplicate filtering is Since it operates with time complexity, real-time processing is possible, and it provides a technical effect that prevents unnecessary computational load on the system, allowing the priority calculation unit and content control unit to focus solely on truly new urgent information.

[0044] In addition, the data processing unit of the present invention is equipped with an edge response mechanism to prepare for network failure situations, which gives it excellent reliability as a disaster management system. In a disaster situation, the communication infrastructure itself may be damaged and the connection with the cloud server (400) may be severed; however, existing cloud-dependent systems have a fatal weakness in that case, becoming completely useless. Accordingly, the present invention automatically caches event data within a valid timestamp received within the last 24 hours in the local storage (e.g., SSD or eMMC) of the display system (100), and when the communication unit detects a failure in connection with the cloud server (400), it immediately switches to Safe Mode and continuously displays information among the cached data that is still within the expiry period as of the current time.

[0045] For example, if a typhoon warning is valid from 2 PM to 6 PM, even if the network is disconnected at 3 PM, the display board continues to display the relevant typhoon warning stored locally until 6 PM, thereby ensuring the continuity of citizen safety information delivery. This distinguishes itself as the first application of the core benefits of Edge Computing technology—'reduced dependency on central authority' and 'independent operation in the event of failure'—to the field of disaster safety.

[0046] Referring to FIG. 4, the operation process of the data normalization and intelligent control algorithm is examined in detail. First, in step 400, data and weather information are collected, and in step 410, it is determined whether the wind speed or temperature data measured from the six weather sensors (141) exceeds a danger threshold. If extreme weather conditions such as a typhoon or a heatwave are detected and device protection is required, the self-protection mode of step 420 is executed to cut off the panel power or drive the heater and fan to ensure hardware durability.

[0047] If weather conditions are within the normal range, the disaster severity in Step 440 is determined through normalization and hash filtering in Step 430 as shown in the formula below ( ), proximity to the place of origin ( ), timeliness of information( ), credibility of the issuing institution( Priority score with ) as the independent variable ( Produces ).

[0049]

[0051] In the above mathematical formula 1 inside represents a weight constant according to the operation policy, and if the score is above the threshold in step 450, an interrupt is executed to immediately stop the transmission of commercial advertisements and forcibly transmit disaster safety information according to step 460, and if it is below the threshold, the existing content transmission schedule is maintained through step 470.

[0052] In addition, the severity of the disaster, which is an individual independent variable included in the above mathematical formula 1 ( ), proximity to the place of origin ( ), timeliness of information( ) and credibility of the source( ...etc. use values ​​normalized to between 0 and 1 respectively to ensure consistency and objectivity of calculations, and the weights assigned to each variable ( inside It is desirable to set the total sum of ) to 1 (100%).

[0053] As a specific operational example, we assume a scenario where an imminent disaster event, such as a major fire or earthquake, occurs near the installation location of the display stand. We assume that the display stand is installed at the Teheran-ro Intersection in Gangnam-gu, Seoul (Latitude 37.5010, Longitude 127.0394), and that at 2:30 PM, a major fire broke out in a commercial building in Yeoksam-dong (Latitude 37.4995, Longitude 127.0380), which is within a 500m radius. In this case, the proximity of the event ( The ) score spikes to converge to 1, and this is the total priority score ( It acts as a decisive factor in raising the display threshold above a preset level. Consequently, the system immediately interrupts the commercial advertisement currently being broadcast based on intelligent judgment and forcibly broadcasts the most urgent evacuation route and safety information to the residents of the area on the full screen.

[0054] The above example can be quantitatively elaborated as follows.

[0055] The priority calculation unit calculates the straight-line distance between the display stand location and the fire origin using the Haversine Formula, obtaining a result of approximately 170m. This is used as a proximity function (e.g., When substituted into ), the proximity score is calculated to be approximately 0.84. At the same time, the severity of the fire information received from the Fire Department API is classified as "Grade 3 fire (total building combustion)" and a severity score of 0.9 is assigned; since the time of publication of the information is within 1 minute of the current time, the recency score is close to 1.0, and since the issuing agency is the Fire Department, an official national agency, a credibility score of 1.0 is also assigned.

[0056] weights (Severity), (Proximity), (latest), When set to (credibility), the final priority score is calculated as (0.3X0.9) + (0.4X0.84) + (0.2X1.0) + (0.1X1.0) = 0.27 + 0.336 + 0.2 + 0.1 = 0.906. Assuming the interrupt threshold set in the system is 0.7, the score of 0.906 significantly exceeds this, so an immediate interrupt is executed, the cosmetics advertisement currently being broadcast is stopped within 1 second, and emergency safety information stating "Large fire within 500m radius - No evacuation towards Teheran-ro - Quickly evacuate towards Seolleung Station" takes up the entire screen.

[0057] The priority calculation algorithm of the present invention has additional patentability in that it goes beyond simple single threshold comparison and dynamically varies the threshold according to the density of the floating population by time period.

[0058] For example, during business hours at 2:00 PM on weekdays, the pedestrian traffic around the Teheran-ro intersection is measured in real-time through the camera module (150) of the sensor unit (140) and counted as approximately 500 people per minute, whereas at 3:00 AM, it decreases to less than 10 people per minute. This pedestrian traffic density information is transmitted to the content control unit in real-time, and intelligent adaptive control is performed by lowering the threshold to 0.6 to respond more sensitively during times of high pedestrian traffic and raising the threshold to 0.8 during times of low pedestrian traffic to suppress unnecessary interruptions. This reflects the reality that the risk of citizen exposure differs even for disasters of the same severity, and by differentiating the response between a situation where 500 people are exposed during the day and a situation where 10 people are exposed at dawn, it maximizes the efficiency of safety information delivery while simultaneously minimizing advertising losses, thereby producing a dual effect.

[0059] Conversely, we also examine specific scenarios where the priority score does not exceed the threshold. For example, let us assume that a heavy snow advisory was issued in Sokcho-si, Gangwon-do (latitude 38.2070, longitude 128.5910) rather than Gangnam-gu, Seoul. The distance from the Teheran-ro display stand is approximately 185 km, so the proximity score is only 0.000001. The heavy snow advisory is calculated as having a low severity score of 0.3 for fire or earthquake preparedness, a recency score of 0.7 for 1 hour since it was issued, and a credibility score of 0.8 for the local government. In this case, the final priority score is calculated as (0.3X0.3) + (0.4X0.000001) + (0.2X0.7) + (0.1X0.8) = 0.09 + 0.0000004 + 0.14 + 0.08 = 0.31, which is significantly lower than the threshold of 0.6, so the interrupt is not executed and the existing ad delivery continues normally.

[0060] However, the present invention utilizes the split transmission mode (510) of FIG. 5 for such intermediate priority information to overlay brief information such as "Heavy snow warning issued for Sokcho City, Gangwon Province" in the form of a ticker in the bottom 10% area of ​​the screen, thereby realizing a balanced operation that fulfills the obligation to deliver public information while maintaining the transmission of advertisements.

[0061] Referring to FIG. 5, the dynamic variable state of the display layout according to the priority calculation result is explained as follows: during normal times, full-screen advertising is performed in the general broadcast mode (500), but when the score is in the intermediate stage, a public information ticker is overlaid in the bottom area through the split broadcast mode (510), and in the emergency broadcast mode (520) which is the danger stage, the entire screen is occupied by a disaster evacuation notice. At this time, the voice guidance operation icon (522) and the real-time control operation icon (524) are activated, and the speaker (160) and camera module (150) are simultaneously operated to eliminate blind spots in visual information and activate a three-dimensional response system that transmits the situation of the flow of people at the site in real time.

[0062] One of the key patentability features of the present invention is that when the emergency transmission mode (520) is operated, the camera module (150) and the speaker (160) are organically combined with the display screen to produce a synergistic effect. Specifically, the moment the emergency transmission mode (520) is activated, the content control unit transmits an automatic shooting command to the camera module (150), and the camera module (150) captures the road and pedestrian path in front of the display stand and transmits real-time video to the central management system (CMS, 200) at a speed of 5 frames per second (5 FPS) or more.

[0063] At the same time, the AI ​​video analysis algorithm embedded in the camera module (150) detects human objects in real time from the captured video, counts the number of citizens currently staying around the display stand, and tracks their direction of movement and speed.

[0064] For example, in the fire situation described above, if the camera detects that there are approximately 80 citizens within a radius of 50m of the display stand and captures that 30 of them are moving in the direction of the fire, the system determines this as a "wrong evacuation direction" and automatically transmits a customized voice guidance through the speaker (160) at maximum volume (e.g., 95dB) saying, "The fire is in the direction of Teheran-ro. Please evacuate in the opposite direction, towards Seolleung Station."

[0065] Unlike existing systems that simply play pre-recorded messages, this video analysis-based adaptive voice guidance clearly distinguishes itself as an intelligent safety system that responds to on-site situations in real time.

[0066] Furthermore, the real-time video transmitted by the camera module (150) is displayed on the integrated control console of the central management system (CMS, 200), allowing a disaster management officer to simultaneously monitor video collected from multiple display boards. Unlike the existing system where the CCTV control system and the disaster information dissemination system were operated separately, the smart display board system (100) of the present invention integrates the two-way functions of 'information dissemination' and 'on-site control' into a single infrastructure, thereby reducing installation and operating costs while dramatically improving the efficiency of disaster response, which gives it great practical value as a public safety infrastructure.

[0067] In addition, voice guidance through a speaker (160) is provided to vulnerable groups who have difficulty perceiving the display screen, such as the visually impaired or the elderly, thereby ensuring accessibility to disaster safety information and achieving a social effect of realizing an inclusive safety infrastructure.

[0068] After the situation ends, the commercial loss of the advertiser is technically compensated by calculating the ad time packets that were not exposed due to the interrupt and automatically generating a supplementary transmission schedule that reflects pedestrian traffic weighting.

[0069] Specifically, as in the previous fire example, it is assumed that the cosmetics advertisement was interrupted for 15 minutes from 2:30 PM to 2:45 PM. The content control unit retrieves pedestrian traffic data (500 people per minute) for that time period from the database and calculates it using a conversion factor of 2.5 (=500 / 200) relative to the standard time period (e.g., 6 AM to midnight, average 200 people per minute). Therefore, it is determined that the actual interrupted time of 15 minutes corresponds to a loss of advertising value of 15 minutes × 2.5 = 37.5 minutes when the pedestrian traffic weight is applied. The system automatically searches for early morning hours with relatively low pedestrian traffic (e.g., 2 AM to 4 AM, 50 people per minute) among the broadcast schedules within the next 7 days and schedules 75 minutes (37.5 minutes × 0.5 = 75 minutes) of that time period as supplementary cosmetics advertisements, thereby ensuring that the total sum of advertising exposure value (total number of exposed people) is equally preserved.

[0070] Unlike existing methods that simply "add more broadcasts for the interrupted time," this pedestrian traffic-based advertising compensation scheduling algorithm is patentable as a convergence technology in the advertising industry and public safety sectors, as it provides fair and reasonable compensation to advertisers by quantifying and rewarding the 'exposure population,' which represents the actual value of the advertisement.

[0071] Finally, referring to FIG. 6, the environment sensing feedback control loop is described, and the data measured from the six weather sensors and illuminance sensors (600) is processed in real time by the AI ​​cloud and intelligent controller (610).

[0072] The above controller performs a display output optimization step (620) in response to illuminance data to adjust brightness and flashing rate, thereby minimizing visual fatigue and complying with regulations, and extends hardware lifespan and maximizes energy efficiency by completing a closed-loop system that actively controls internal heaters and fans through an enclosure environment optimization step (630) based on temperature and humidity data.

[0073] The self-protection logic of the present invention is described in detail with specific numerical values ​​as follows. First, if the instantaneous wind speed measured by the wind speed sensor records 25 m / s or more (typhoon-level strong wind) or if the average wind speed for 3 minutes continues to be 20 m / s or more, the system determines that there is a high risk of physical damage to the display panel and the housing due to wind and automatically executes a protection mode.

[0074] Specifically, the content control unit first minimizes power consumption by rapidly reducing the display brightness from 100% to 10% or less, or turning it off completely, while simultaneously displaying a protection mode guidance message such as "Strong wind detected - Pause for safety" on the screen at low brightness. This provides a dual effect of not only saving power but also preventing panel deformation or joint damage caused by thermal expansion by suppressing the heat generated by the LED module during high-brightness emission, thereby preventing an increase in internal temperature. When the wind speed returns to a safe range (e.g., 15 m / s or less), the system automatically returns to normal operation mode after a stabilization period of 5 minutes.

[0075] In the case of temperature sensor-based self-protection logic, if the internal temperature of the enclosure exceeds 60 degrees Celsius due to direct sunlight during the summer, the lifespan of the LED module is drastically shortened and thermal damage to the control circuit may occur. To prevent this, when the temperature sensor detects 55 degrees Celsius, a primary alarm is triggered, and a fan (not shown) is automatically driven to expel internal heat to the outside by gradually increasing the rotation speed using a PWM method. If the temperature continues to rise despite the fan operation and reaches 60 degrees Celsius, secondary protection logic is activated to reduce heat generation by limiting the display brightness to 70%. When it reaches 65 degrees Celsius, as a final protection measure, the display is completely turned off, and only the "High temperature detected - Cooling" message is displayed in low-power mode.

[0076] Conversely, during the winter season, if the internal temperature of the enclosure drops below 5 degrees Celsius or the relative humidity rises above 85% and a risk of condensation is detected, a PTC heater (not shown) is automatically activated to maintain the interior at 10 to 15 degrees Celsius, thereby preventing moisture condensation at the source and preventing wet corrosion of electronic components. This temperature and humidity-based self-protection logic dramatically improves the durability of outdoor equipment operating unmanned 24 hours a day, 365 days a year, and demonstrates economic benefits by reducing annual maintenance costs by more than 40% compared to existing methods.

[0077] The method of utilizing the solar irradiance sensor also demonstrates the unique patentability of the present invention. When the solar radiation energy measured by the solar irradiance sensor records more than 1,000W per square meter (strong direct sunlight at midsummer noon), the system utilizes this as a high-temperature risk prediction signal to perform predictive control, which preemptively activates the fan before the temperature sensor reaches a critical threshold. For example, even if the internal temperature of the enclosure is only 50 degrees Celsius under conditions of high solar irradiance, if analysis of accumulated past data predicts a high probability of reaching 60 degrees Celsius within the next 30 minutes, the fan is operated at a low speed in advance to gently suppress the temperature rise. This has been experimentally verified to improve energy efficiency by 25% and reduce the rate of component degradation by 30% compared to simple reactive control.

[0078] In addition, the display brightness is dynamically optimized by utilizing both an illuminance sensor and a solar radiation sensor. For example, under direct sunlight during the day, the brightness is maximized to 100% to ensure visibility, while at night, it is automatically reduced to 30% to reduce visual fatigue and save 70% on power consumption. This illuminance-adaptive brightness control also provides a compliance effect by automatically ensuring compliance with nighttime brightness regulations under the Outdoor Advertising Act (e.g., 400 cd / ㎡ or less).

[0079] Although this embodiment is configured and operates as described above, it is not limited thereto and various variations are possible. For example, the display module (110) is exemplified as an LED panel in this embodiment, but can be replaced with other display technologies such as LCD, OLED, or MicroLED, and can be implemented as an aesthetically pleasing display stand that harmonizes with the background building by applying a transparent display.

[0080] The communication unit (not shown) can be configured with various wireless communication protocols such as Wi-Fi 6, LoRaWAN, and NB-IoT in addition to 5G / LTE, and can be expanded into an ultra-high reliability system that operates even in remote mountainous areas or when terrestrial communication networks are cut off during disasters by adding satellite communication.

[0081] The weights (w1~w4) of the priority calculation algorithm can be adjusted according to the operational policy; for example, display boards installed in tourist attractions or parks can be tuned to display daily public information (traffic congestion, weather, etc.) more frequently by lowering the severity weight and increasing the recency weight.

[0082] The sensor unit (140) can be expanded into an environmental monitoring platform by adding a fine dust sensor, a noise meter, a vibration detector, etc., in addition to the six weather sensors, and the collected environmental data can be linked with the local government's smart city integrated platform and used for urban environment management.

[0083] The above speaker (160) can be configured as a multi-channel speaker array instead of a single-directional speaker to apply directional audio technology that concentrates sound in a specific direction. In this case, it is possible to minimize noise complaints from residential areas around the display stand while achieving the effect of delivering voice guidance intensively only to citizens in the direction requiring evacuation.

[0084] In addition, the camera module (150) can be improved to accurately detect human objects even at night or in foggy conditions by incorporating a thermal camera in addition to a visible light camera, and the accuracy of counting the flow of people can be improved to over 95% by adding a 3D depth sensor.

[0085] The ad reward scheduling algorithm can provide advertisers with more precise ROI (Return on Investment) analysis results by integrating actual ad effectiveness measurement data (e.g., number of QR code scans, growth rate of product search volume, etc.) rather than being based on simple time.

[0086] In addition to commercial power, the power supply method can be implemented as an energy-independent display by applying a hybrid power system that combines solar panels and lithium-ion batteries, and in this case, ultra-high reliability can be secured, enabling independent operation for more than 72 hours even in the event of a power outage caused by a disaster.

[0087] It is obvious to those skilled in the art that these variations are included within the scope of the present invention, as they all modify the specifications of the components or algorithm parameters while maintaining the organic combination structure of 'AI-based data normalization - priority calculation - interrupt control - self-protection,' which is the core technical concept of the present invention.

[0088] The present invention is not limited to the aforementioned embodiments, and it is obvious to those skilled in the art that various sensors may be added or algorithms modified within the scope of the technical concept. Explanation of the symbols

[0089] 100: Smart Display System 110: Display module 115: Display inspection door 120: Frame 130: Landlord 140: Sensor section 141: 6 major weather sensors 143: Light sensor 150: Camera module 160: Speaker 200: Central Management System (CMS) 300: External server group 310: Public API 320: Disaster Safety API 330: Local Government API 400: Cloud Server 410: AI Analysis Processing Node 500: Normal broadcast mode 510: Split transmission mode 520: Emergency Transmission Mode

Claims

Claim 1 A communication unit that receives disaster or public interest event data generated by AI analysis processing of data received from an external server on a cloud server; a data processing unit that normalizes the received data into a Standard Event Schema including fields for disaster type, severity, location, time of occurrence, issuing agency, and validity period by applying Natural Language Processing (NLP) technology and a machine learning-based schema mapping algorithm, and filters duplicate received data in real time based on a unique identification value generated by inputting a string formed by concatenating the source information and issuance time information of the data into a hash function; a sensor unit including six weather sensors measuring temperature, relative humidity, atmospheric pressure, wind direction, wind speed, and solar radiation, an illuminance sensor, and a camera module; a priority score (Score_pr) that calculates a priority score through a calculation formula including the severity of the disaster, proximity to the site of occurrence, the timeliness of the information, and the credibility of the information issuing agency as independent variables, and if the priority score (Score_pr) is greater than or equal to a preset threshold, the currently playing content A priority calculation unit that executes an interrupt, wherein the threshold is dynamically variable according to the time-based population density measured through the camera module;An AI smart display stand system comprising: a content control unit that controls the display device to display disaster or public interest event data by varying the screen layout according to the priority score (Score_pr); when displaying disaster information, tracks the direction of movement of citizens by detecting surrounding objects in real-time through the camera module; automatically outputs customized voice guidance through a speaker to guide the correct evacuation route if a citizen moving in the wrong evacuation direction is detected; executes self-protection logic to cut off power to the display device or transmit a protection mode screen if wind speed or temperature data collected from the sensor unit exceeds a preset device protection threshold; and performs predictive control to preemptively drive a cooling fan before the temperature data reaches the device protection threshold if the solar radiation measurement value collected from the sensor unit is above a preset prediction standard. Claim 2 delete Claim 3 delete Claim 4 The AI ​​smart display board system according to claim 1, characterized in that the communication unit performs an edge response function that, in the event of a communication failure with the cloud server, retrieves data within a valid timestamp cached in a local storage and maintains display for a valid period. Claim 5 An AI smart display stand system according to claim 1, wherein the content control unit selects and executes one of the following modes based on the priority score: a full-screen display mode, an always-on exposure area overlay mode through screen splitting, and a forced insertion mode between existing content slots. Claim 6 The AI ​​smart display board system according to claim 1, wherein the content control unit calculates identification information and non-exposure time packets of advertising content whose transmission was interrupted by the interrupt and stores them in a database, and automatically generates a supplementary transmission schedule reflecting the pedestrian traffic weight at the time of non-exposure. Claim 7 In claim 6, the content control unit is equipped with a filtering logic that detects the brightness and flicker rate of video content to determine compliance with outdoor advertising regulations, and collects and analyzes surrounding pedestrian traffic data through the camera module to reflect in the determination of priority for the supplementary transmission schedule, thereby forming an AI smart display stand system. Claim 8 delete Claim 9 The AI ​​smart display stand system according to claim 1 further comprises an environment control unit including a thermometer / hygrometer, a heater, a fan, and a controller for controlling the same, wherein the environment control unit drives the heater to prevent condensation based on an internal humidity measurement or drives the fan to discharge heat based on an internal temperature measurement to maintain a constant operating environment inside the enclosure. Claim 10 In an intelligent content control method performed by an AI smart display system, the method comprises: (a) a communication unit receiving disaster or public interest event data generated by AI analysis processing of data received from an external server on a cloud server; (b) a data processing unit applying Natural Language Processing (NLP) technology and a machine learning-based schema mapping algorithm to normalize the received data into a Standard Event Schema including fields for disaster type, severity, location of occurrence, time of occurrence, issuing agency, and validity period, and filtering duplicate received data in real time based on a unique identification value generated by inputting a string formed by concatenating the source information and issuance time information of the data into a hash function; (c) a priority calculation unit calculating a priority score (Score_pr) through a calculation formula including independent variables for the severity of the disaster, proximity to the location of occurrence, the timeliness of information, and the credibility of the information issuing agency; (d) the priority calculation unit determining whether to interrupt the currently playing content by comparing the priority score (Score_pr) with a preset threshold, wherein the A step in which the threshold is dynamically varied according to the time-based population density measured through the camera module of a sensor unit comprising six weather sensors measuring temperature, relative humidity, atmospheric pressure, wind direction, wind speed, and solar radiation, an illuminance sensor, and a camera module; (e) a step in which a content control unit varies the screen layout of a display device according to the interrupt determination to display the disaster or public interest event data, wherein when displaying disaster information, surrounding objects are detected in real-time through the camera module to track the direction of movement of citizens, and if a citizen moving in the wrong evacuation direction is detected, a customized voice guidance is automatically output through a speaker to guide the correct evacuation route;and (f) a step in which the content control unit executes self-protection logic to cut off power to the display device or transmit a protection mode screen when wind speed or temperature data collected from the sensor unit exceeds a preset device protection threshold, and performs predictive control to preemptively drive a cooling fan before the temperature data reaches the device protection threshold when the solar radiation measurement value collected from the sensor unit is greater than or equal to a preset prediction reference value; comprising an intelligent content control method.

Images