Smart Predictive External Impact-Mitigation Safety Execution System for Industrial Mobility

Abstract

A smart accident prediction-based external buffer safety execution system for industrial mobility is disclosed. The smart accident prediction-based external buffer safety execution system for industrial mobility according to an embodiment of the present invention comprises: an accident prediction detection module that includes at least one of a radar sensor, an ultrasonic sensor, an image-based detection sensor, or a proximity detection sensor to detect the driving path and surrounding environment of industrial mobility, and determines the probability of an accident by analyzing the distance, relative speed, and approach direction to a potential collision target; an external buffer airbag module that is positioned at at least one of the front frame, side protection structure, or upper protection frame of industrial mobility and has a plurality of buffer airbags that are selectively deployed according to an accident prediction signal; a safety execution control unit that determines a risk level based on a signal input from the accident prediction detection module and controls the deployment time and deployment position of the external buffer airbag module; and an accident record generation module that automatically stores operational data including an accident prediction detection signal, the airbag deployment time, the movement speed of industrial mobility, the duration of impact, and safety execution history.

Description

Technology Field

[0001] The present invention relates to the field of industrial safety technology, and more specifically, to a smart accident prediction-based industrial mobility external cushioning safety execution system that predicts in advance collision, entrapment, rollover, and worker contact accidents of various industrial mobilitys operated at industrial sites, and automatically deploys a cushioning structure placed outside the industrial mobility at a stage prior to the occurrence of an accident to reduce impact energy. Background Technology

[0002] Various industrial mobility devices, such as forklifts, automated guided vehicles (AGVs), and autonomous logistics robots operated in industrial sites, are increasingly becoming faster and more automated to improve work efficiency. Since these industrial mobility devices are often operated in confined workspaces alongside multiple workers and equipment, there is a reality where collisions or entrapment accidents occur frequently during movement. However, conventional industrial mobility devices rely primarily on the attention of drivers or operators to ensure safety, which presents a problem in that it is difficult to respond immediately to unexpected situations.

[0003] In environments where industrial mobility systems operate in close proximity to workers, changes in worker positions occur very rapidly and irregularly. In such environments, even if sensors perceive the surrounding situation, they often remain at the level of simple detection, which limits their ability to lead to actual accident prevention. In other words, while conventional technology partially implements the function of detecting risk factors, it suffers from a lack of a structure to perform proactive safety measures based on that information.

[0004] Furthermore, collision accidents in industrial mobility often go beyond mere contact and lead to serious accidents; in particular, crushing or rollover accidents pose a very high risk of causing casualties. Nevertheless, conventional technology primarily focuses on designs centered on internal protection structures or frame reinforcement to reduce damage after an accident occurs, resulting in insufficient preventive response at the pre-accident stage.

[0005] As industrial automation technology advances, environments where multiple mobility devices operate simultaneously are increasing, leading to a greater risk of mutual interference and collisions between equipment. However, conventional technologies focus on controlling individual devices, resulting in a problem where risk prediction and response considering dynamic interactions with the surrounding environment are not sufficiently achieved.

[0006] Furthermore, the driving paths of industrial mobility vehicles change frequently depending on the work environment, and the locations of obstacles or workers also change in real time. In such dynamic environments, simple distance-based detection or static safety zone setting methods have limitations in effectively preventing accidents. In particular, there is a problem in accurately assessing collision risk when relative speed and approach direction are not comprehensively considered.

[0007] While some conventional technologies have attempted to apply airbags or cushioning structures to industrial equipment, most are implemented by being limited to internal installation or operating only after a collision. Since these structures fail to secure sufficient cushioning space prior to the moment of impact, they have limitations in effectively mitigating actual shock. Furthermore, in the case of structures that are not externally deployable, there is a problem with the limited scope of worker protection.

[0008] Furthermore, existing shock absorbers are configured to operate only under specific conditions, making it difficult to flexibly respond to complex hazardous situations occurring in various work environments. For instance, despite the fact that required shock absorbing characteristics vary depending on movement speed, the type of collision target, and the direction of approach, there is a lack of variable control structures that reflect these factors.

[0009] Furthermore, the inadequate system for recording and analyzing data related to accidents limits the ability to objectively analyze the causes of accidents or reflect them in future safety policies. Conventional technologies often stop at the level of simple event recording, posing a problem in that it is difficult to comprehensively reconstruct the situation before and after an accident.

[0010] Despite the increasing stringency of industrial safety regulations and certification standards, conventional industrial mobility often lacks a systematic safety implementation structure to meet these requirements. In particular, the absence of accident-prevention-oriented active safety systems acts as a disadvantage in regulatory compliance and safety assessments.

[0011] As such, conventional technology focuses on responding after an accident occurs, and there is a lack of integrated safety implementation technology that predicts risks before an accident occurs and actively deploys external cushioning structures to mitigate impact. Therefore, there is a need for technology capable of predicting various accidents that may occur during the operation of industrial mobility in advance and automatically deploying external cushioning structures before an accident occurs to simultaneously protect workers and reduce equipment damage, and the present invention aims to solve the problems of conventional technology. Prior art literature

[0012] Korean Registered Patent Publication No. 10-2926236 (Registration Date: February 06, 2026) The problem to be solved

[0013] It is required to predict collisions, entrapment, and rollover accidents occurring during the operation of industrial mobility in advance, and to ensure worker safety and minimize equipment damage by actively deploying external cushioning structures prior to the accident. Furthermore, it is necessary to implement an intelligent safety execution structure capable of precisely determining risk levels based on various sensor data and variably controlling deployment positions and pressures according to the situation. In addition, the task is to provide a system that automatically records operational data before and after an accident, which can be utilized for industrial safety management and accident analysis. means of solving the problem

[0014] A smart accident prediction-based industrial mobility external buffer safety execution system according to one aspect of the present invention for achieving such objectives is a system composed of software function blocks executed in a computer device including a processor, memory, storage device, input / output interface and network interface,

[0015] The configuration may include: an accident prediction detection module that includes at least one of a radar sensor, an ultrasonic sensor, an image-based detection sensor, or a proximity detection sensor to detect the driving path and surrounding environment of the industrial mobility, and determines the probability of an accident by analyzing the distance, relative speed, and approach direction to a potential collision target; an external cushioning airbag module that is positioned at at least one of the front frame, side protection structure, or top protection frame of the industrial mobility and has a plurality of cushioning airbags that are selectively deployed according to an accident prediction signal; a safety execution control unit that determines a risk level based on a signal input from the accident prediction detection module and controls the deployment timing and deployment position of the external cushioning airbag module; and an accident record generation module that automatically stores operational data including an accident prediction detection signal, the airbag deployment timing, the movement speed of the industrial mobility, the duration of impact, and safety execution history.

[0016] In one embodiment of the present invention, the accident prediction detection module may comprise: a data detection module that includes at least one of a radar sensor, an ultrasonic sensor, an image-based detection sensor, or a proximity detection sensor to detect the driving path and surrounding environment of industrial mobility, acquires detection data in real time, and transmits it to a data analysis module; a data analysis module that receives the data transmitted from the data detection module in real time and analyzes the distance, relative speed, and approach direction to a potential collision target; and an accident occurrence determination module that determines the possibility of an accident based on the results analyzed by the data analysis module and transmits an accident prediction signal to an external cushioning airbag module, a safety execution control unit, and an accident record generation module.

[0017] In one embodiment of the present invention, the external cushioning airbag module is configured to be stored on the outer surface of an industrial mobility vehicle by a hinge-coupled or sliding storage structure and to be deployed outward by a gas generator or a compressed gas cartridge, wherein the deployment pressure or deployment volume is variably controlled according to the magnitude of the collision risk or the moving speed of the industrial mobility vehicle, and may be configured to be deployed upward in response to a change in the angle of inclination of the industrial mobility vehicle or a rollover risk signal to secure a driver protection space.

[0018] In one embodiment of the present invention, the external cushioning airbag module may be configured to include: a first cushioning chamber having a structure that absorbs an initial shock; a second cushioning chamber having a structure that sprays absorbed shock energy and is mounted in a stacked form with respect to the first cushioning chamber; and a pressure relief chamber having a structure that stacks and arranges the first cushioning chamber and the second cushioning chamber together and discharges pressure in stages.

[0019] In one embodiment of the present invention, the safety execution control unit may be configured to include: a drive power cutoff module that automatically cuts off the drive power of an industrial mobility according to a preset standard; a deceleration warning control module that controls the industrial mobility to perform deceleration control or output an alarm signal according to a preset standard; and an accident spread prevention module that prevents the spread of an accident by controlling the operation of surrounding industrial facilities in conjunction through wireless or wired communication. Effects of the invention

[0020] According to the present invention, by detecting the surrounding environment in real time while industrial mobility is in motion and determining the possibility of a collision in advance, an external cushioning airbag is automatically deployed prior to the occurrence of an accident, thereby providing the effect of simultaneously protecting workers and equipment. Furthermore, by variably controlling the deployment position and pressure according to the collision risk and driving conditions, stable impact reduction performance can be secured even in various work environments. In addition, the spread of accidents can be effectively prevented through power cutoff, deceleration and alarm output, and control linked with surrounding equipment. Moreover, by automatically recording data generated during the accident prediction and execution process, it provides a highly reliable data base that can be utilized for industrial safety management, accident cause analysis, and regulatory compliance. Brief explanation of the drawing

[0021] FIG. 1 is a block diagram showing a smart accident prediction-based industrial mobility external buffer safety execution system according to one embodiment of the present invention. FIG. 2 is a flowchart illustrating a smart accident prediction-based industrial mobility external buffer safety execution method according to an embodiment of the present invention. Specific details for implementing the invention

[0022] Preferred embodiments of the present invention will be described in detail below with reference to the drawings. Prior to this, terms and words used in this specification and claims should not be interpreted as being limited to their ordinary or dictionary meanings, but should be interpreted in a meaning and concept consistent with the technical spirit of the present invention.

[0023] Throughout this specification, when it is stated that one component is located "on" another component, this includes not only cases where one component is in contact with another component, but also cases where another component exists between the two components. Throughout this specification, when it is stated that a part "includes" a component, this means that, unless specifically stated otherwise, it does not exclude other components but may include additional components.

[0024] FIG. 1 shows a block diagram illustrating a smart accident prediction-based industrial mobility external buffer safety execution system according to one embodiment of the present invention, and FIG. 2 shows a flowchart illustrating a smart accident prediction-based industrial mobility external buffer safety execution method according to one embodiment of the present invention.

[0025] The smart accident prediction-based industrial mobility external buffer safety execution system according to the present invention can be extended to industrial mobility as well as drones, military equipment, robots, vehicles, etc.

[0026] Referring to these drawings, the smart accident prediction-based industrial mobility external cushioning safety execution system (100) according to the present embodiment can prevent worker entrapment and collision accidents by actively operating the external cushioning structure in the stage prior to the occurrence of an accident, and can also minimize damage to industrial mobility and surrounding equipment through the dispersion of impact energy, and can strengthen the industrial safety management system by utilizing record data generated during the accident response process, and can also provide advantageous effects for responding to industrial safety regulations, smart safety equipment certification, public safety technology evaluation, and industrial safety policy response.

[0027] Hereinafter, with reference to FIGS. 1 and 2, each component constituting the smart accident prediction-based industrial mobility external buffer safety execution system (100) according to the present embodiment will be described in detail.

[0028] Detailed description of the accident prediction detection module (110)

[0029] The accident prediction detection module (110) is composed of a core functional block that recognizes the interaction with the surrounding environment in real time while driving industrial mobility and determines the possibility of an accident in advance by integrally analyzing data acquired from various sensors. The accident prediction detection module (110) may be configured to include a plurality of heterogeneous sensors, such as radar sensors, ultrasonic sensors, image-based detection sensors, and proximity detection sensors, and is designed to acquire distance, speed, and direction information complementarily by considering the characteristics of each sensor.

[0030] The accident prediction detection module (110) is configured to minimize errors that may occur in single-sensor-based perception by integrating and processing data received from each sensor through time synchronization and coordinate matching. For example, by obtaining relative speed and long-distance information through a radar sensor, recognizing the shape and type of an object through an image-based detection sensor, and supplementing short-distance precision distance information through an ultrasonic sensor, more reliable environmental perception results can be obtained.

[0031] The accident prediction detection module (110) is configured to track the position change pattern of a potential collision target based on analyzed data and to predict a potential collision point by determining whether it intersects with the driving path of industrial mobility. In this process, risk judgment based on dynamic prediction rather than simple distance-based judgment is made by considering the acceleration, direction change characteristics, and expected movement trajectory of the moving object.

[0032] The accident prediction detection module (110) is configured to classify the likelihood of an accident into risk grades and output them, and the risk grade is calculated by comprehensively considering the expected collision time, relative speed, approach angle, and object type. Through this, a quantitative criterion is provided that can determine whether to actually deploy a buffer airbag beyond a simple warning level.

[0033] The accident prediction detection module (110) can be applied to an autonomous transport vehicle operated in a logistics warehouse as an example of actual application in an industrial site, and contributes to preventing accidents by detecting in real time when a worker enters the vehicle's path and determining the possibility of a collision in advance, and transmitting a signal to the external cushioning airbag module (120) and the safety execution control unit (130).

[0034] Detailed description of the data detection module (111)

[0035] The data detection module (111) is a sub-component of the accident prediction detection module (110) and performs the function of directly detecting physical signals regarding the surrounding environment of industrial mobility and converting them into digital data. The data detection module (111) may be configured to include a radar transceiver, an ultrasonic transducer, a camera sensor, and a proximity sensor, and each sensor is designed to acquire environmental information based on different detection ranges and resolutions.

[0036] The data detection module (111) is configured to generate highly reliable detection data by performing filtering and noise removal processing according to the signal characteristics of each sensor. For example, in the case of radar signals, a signal processing algorithm to remove noise caused by multiple reflections may be applied, and in the case of image data, correction processing for lighting changes and shadow effects may be performed.

[0037] The data detection module (111) samples the detected data at regular intervals and transmits it to the data analysis module (112), and may include a high-speed data processing structure to minimize time delay during this process. In addition, considering industrial environments where real-time performance is important, a dedicated interface or high-speed communication protocol may be applied to reduce data transmission delay.

[0038] The data detection module (111) can be configured so that the placement of sensors is optimized according to the installation location and operating environment of the industrial mobility, and may include a multi-sensor array structure to simultaneously obtain environmental information from various directions such as the front, side, and rear. This minimizes blind spots and enables more accurate risk detection.

[0039] The data detection module (111) can be applied to a forklift operating in a narrow passage as an example of actual application, and detects the approach of surrounding workers from various angles and transmits this to the data analysis module (112), thereby enabling early detection of collision risks. Through this structure, it can contribute to preventing various types of accidents that may occur in industrial sites in advance.

[0040] Detailed description of the data analysis module (112)

[0041] The data analysis module (112) is a core processing unit that receives multi-type sensor data transmitted from the data detection module (111) and performs integrated analysis, and performs the function of enabling precise situational awareness of the surrounding environment of industrial mobility. The data analysis module (112) is configured to generate an integrated environment model by aligning radar-based distance and speed information, image-based object recognition information, ultrasonic-based short-range distance information, etc., into a single coordinate system.

[0042] The data analysis module (112) applies a sensor fusion algorithm to combine the advantages of each sensor and compensate for its disadvantages, thereby producing highly reliable analysis results. For example, by classifying the type of object through an image sensor, measuring relative speed through a radar sensor, and supplementing proximity information through an ultrasonic sensor, the location, direction of movement, and risk level of the object can be calculated more accurately.

[0043] The data analysis module (112) includes a function to track the movement trajectory of an object over time and generates a dynamic model that predicts future positions by analyzing changes in the object's speed, acceleration, and direction based on continuous sensor data. This prediction model is used to evaluate the possibility of a collision after a certain period of time in advance, rather than simply determining static distance.

[0044] The data analysis module (112) is configured to calculate intersection points where a collision is possible by comparing and analyzing the driving path information of industrial mobility and the movement trajectory of surrounding objects. At this time, the risk index is quantified by comprehensively considering the expected collision time, collision angle, and characteristics of the collision target, and this is used as a key judgment criterion transmitted to the accident occurrence judgment module (113).

[0045] The data analysis module (112) can be applied to an environment in which multiple AGVs cross-travel in an automated logistics system as an actual application example, and can maximize the accident prevention effect by simultaneously analyzing the movement path of each vehicle and the location information of the worker to derive potential collision situations in advance and inducing a response of the safety execution control unit (130) based on this.

[0046] Detailed explanation of the accident occurrence judgment module (113)

[0047] The accident occurrence determination module (113) performs the function of finally determining the possibility of an accident based on the analysis results transmitted from the data analysis module (112) and generating an accident prediction signal accordingly. The accident occurrence determination module (113) calculates the risk level by comprehensively considering the expected collision time, relative speed, approach direction, and object type, and is configured to determine whether an accident is possible by comparing this with a pre-set reference value.

[0048] The accident occurrence judgment module (113) can be configured to output risk levels classified into multiple stages, and each stage can be subdivided into warning level, caution level, and emergency response level. Through this stage-based judgment structure, precise safety execution linked to actual buffer airbag deployment and drive control is possible, rather than just a simple warning signal.

[0049] The accident occurrence judgment module (113) can be configured to dynamically adjust judgment criteria according to various situational variables, and the risk judgment sensitivity can be variably set by reflecting the driving speed of industrial mobility, the congestion of the work environment, and the density of workers. Through this, stable and flexible accident prediction is possible even in various industrial environments.

[0050] The accident occurrence judgment module (113) is configured to simultaneously transmit the judgment result to the external cushioning airbag module (120), the safety execution control unit (130), and the accident record generation module (140), thereby enabling integrated safety execution in which airbag deployment, equipment control, and data recording occur simultaneously. In particular, a high-speed computation structure and real-time processing function may be applied to minimize judgment delay.

[0051] As an actual application example, the accident occurrence judgment module (113) can realize worker protection by determining the possibility of a collision within a short time and generating an emergency response signal to immediately deploy the external cushioning airbag module (120) in a situation where a worker rapidly enters the path of industrial mobility. This structure plays an important role in implementing an active response in the pre-accident stage, which was lacking in conventional technology.

[0052] Five embodiments according to the judgment criteria of the accident occurrence judgment module (113)

[0053] First embodiment: Time-to-Collision based judgment module (113)

[0054] The accident occurrence determination module (113) is configured to determine the risk level by calculating the Time-To-Collision based on relative speed and distance information between the industrial mobility and the collision target. The accident occurrence determination module (113) determines a high-risk situation when the TTC value decreases below a preset threshold and controls the execution of safety actions, including the deployment of the external cushioning airbag module (120).

[0055] The accident occurrence judgment module (113) is configured to make a more precise prediction by reflecting not only speed relative to distance but also acceleration and deceleration patterns when calculating TTC. For example, even if a moving worker suddenly changes direction, the TTC can be recalculated by reflecting this to reduce misjudgment.

[0056] The accident occurrence judgment module (113) can be utilized as an actual application example in a logistics warehouse to detect when the TTC rapidly decreases while the AGV is moving and a worker enters the front, and if it is determined to be below a certain time, to immediately decelerate and deploy an airbag to prevent a collision.

[0057] Second embodiment: Relative speed and approach direction-based determination module (113)

[0058] The accident occurrence judgment module (113) is configured to determine the risk level by analyzing the relative speed and approach direction vector between the industrial mobility and the collision target. It prevents unnecessary malfunctions by evaluating the risk level as high only when the approach direction matches the collision path, rather than judging based solely on distance.

[0059] The accident occurrence judgment module (113) defines a high-risk state when the relative speed is above a certain level and the approach direction is included within a certain angle range, and is configured to generate an emergency response signal only when the conditions are simultaneously satisfied. Through this, accurate risk judgment is possible even in a complex environment where multiple objects exist in the surroundings.

[0060] The accident occurrence judgment module (113) can be used as an example of actual application in which, when a worker approaching from the side crosses the direction of travel of the vehicle at a speed greater than a certain speed while the forklift is moving through the workspace, the airbag in that direction is selectively deployed to prevent a collision.

[0061] Third embodiment: Object type and risk weight-based judgment module (113)

[0062] The accident occurrence judgment module (113) is configured to analyze the type of detected object and determine the likelihood of an accident by applying risk weights for each object. For example, different risk coefficients are applied according to the type of object, such as people, vehicles, and fixed structures, so that a differentiated judgment is made even under the same distance conditions.

[0063] The accident occurrence judgment module (113) classifies objects using image-based detection data, assigns a high risk weight to people, and applies a relatively low weight to simple obstacles. This enables safety control that prioritizes worker protection.

[0064] The accident occurrence judgment module (113) can be configured to immediately decelerate and deploy an airbag when a worker approaches the equipment in a factory as an example of actual application, and to respond with a warning and deceleration level when a simple pallet or box approaches under the same conditions.

[0065] Fourth embodiment: Multi-sensor reliability-based judgment module (113)

[0066] The accident occurrence judgment module (113) is configured to determine the risk level by evaluating the reliability of data obtained from multiple sensors. It calculates the reliability based on the status of each sensor, environmental conditions, and data consistency, and improves the judgment accuracy by giving greater weight to data with high reliability.

[0067] The accident occurrence judgment module (113) is configured to perform a judgment based on radar and ultrasonic sensor data, for example, when the image sensor has low reliability due to insufficient lighting. Conversely, when there is a lot of noise in the radar signal, it operates by prioritizing image-based data.

[0068] The accident occurrence judgment module (113) can be utilized as an actual application example to accurately detect the risk of collision through radar-based judgment even when camera recognition is difficult in a night work environment, and to ensure safety by deploying an airbag when necessary.

[0069] Fifth embodiment: Composite risk index integrated-based judgment module (113)

[0070] The accident occurrence determination module (113) is configured to calculate a single composite risk index by integrating distance, relative speed, approach direction, object type, and sensor reliability, and to determine the possibility of an accident based on this. Each element is summed with weights applied to derive an overall risk level.

[0071] The accident occurrence judgment module (113) determines a high-risk state when the composite risk index exceeds a preset threshold, and controls deceleration, warning, and airbag deployment to be performed step by step according to the risk level. This integrated judgment method provides higher accuracy compared to a single standard.

[0072] The accident occurrence judgment module (113) can be utilized in an environment where multiple mobile equipment and workers are mixed in a complex automation process as an example of actual application, and by comprehensively considering various factors, it provides the effect of reducing unnecessary malfunctions while enabling a quick and accurate response in actual dangerous situations.

[0073] Detailed description of the external cushioning airbag module (120)

[0074] The external cushioning airbag module (120) is a core physical safety structure that is positioned on the outside of the industrial mobility and is actively deployed prior to the occurrence of an accident, thereby preemptively reducing impact energy from the outside in the event of a collision or entrapment. The external cushioning airbag module (120) is configured to be distributed in various locations, such as the front frame, side protection structure, and upper protection frame, and is selectively deployed according to signals transmitted from the accident prediction detection module (110) and the safety execution control unit (130).

[0075] The external cushioning airbag module (120) is configured to be stored in close contact with the outer surface of the industrial mobility under normal conditions, including a hinged cover structure or a sliding storage structure. This storage structure is designed to minimize damage from external impacts or the working environment, while allowing it to quickly protrude outward when deployed. In the case of a hinge structure, a rotation axis and a return spring may be combined to enable instantaneous deployment, and in the case of a sliding structure, it is deployed in a straight direction through a guide rail and a linear drive unit.

[0076] The external cushioning airbag module (120) operates by instantaneously inflating the internal chamber using a gas generator or a compressed gas cartridge. The gas generator can be configured with an ignition method or an electric opening valve method, and the compressed gas cartridge is configured to ensure a rapid deployment speed using high-pressure nitrogen or carbon dioxide. Through this structure, a sufficient cushioning space can be formed within a short time prior to an accident.

[0077] The external cushioning airbag module (120) is configured to variably control the deployment pressure and volume according to the magnitude of the accident risk and the driving speed of industrial mobility. To this end, a multi-stage valve structure or an electronic flow control valve may be applied, and it is controlled to deploy slowly at low pressure in low-speed environments and to inflate rapidly at high pressure in high-speed collision risk situations. This variable control structure contributes to securing optimal cushioning performance in various working environments.

[0078] The external cushioning airbag module (120) can be applied to the side of an autonomous transport vehicle operated in a logistics warehouse as an example of actual application, and when a worker suddenly approaches, the airbag immediately deploys in that direction to form a cushioning space between the worker and the equipment. In addition, when a change in the angle of inclination is detected, the airbag deploys upward to secure a driver protection space in the event of a rollover.

[0079] Detailed description of the first buffer chamber (121)

[0080] The first cushioning chamber (121) is a structure located at the outermost side of the external cushioning airbag module (120) that performs the role of directly absorbing the initial impact. The first cushioning chamber (121) is positioned in the area that first contacts the external collision target and is designed to absorb impact energy by rapidly deforming when an impact occurs. Through this, the transmission of impact to the internal structure can be primarily reduced.

[0081] The first buffer chamber (121) may be formed from a high-strength fiber material such as high-elasticity polyurethane, nylon-based fabric, or aramid fiber, and is maintained with air or gas filled inside. The outer shell may be formed in a multi-layer structure to ensure both wear resistance and heat resistance, and reinforcing ribs or woven patterns may be applied to minimize damage even under repeated impacts.

[0082] The first buffer chamber (121) is configured to deform as its internal pressure rapidly increases upon impact, and to absorb impact energy in the form of internal air compression during this process. Additionally, a channel structure that induces fine airflow may be formed inside, so that upon impact, the pressure is designed to be evenly distributed rather than concentrated in a specific area.

[0083] The first buffer chamber (121) is configured to operate in conjunction with the second buffer chamber (122) and the pressure discharge chamber (123), and after absorbing the initial shock, it transfers some of the pressure to subsequent chambers to enable multi-stage buffering. This structure prevents the problem of excessive pressure concentration that can occur in a single chamber and enables the stepwise dispersion of shock energy.

[0084] The first cushioning chamber (121) can be installed on the front of a forklift as an example of actual application, and in a situation where a collision with a worker is expected, the first cushioning chamber (121) makes contact first after the airbag deploys to absorb the impact, thereby providing the effect of significantly reducing the impact transmitted to the worker. In addition, it can be designed to maintain stable cushioning performance even in a repetitive work environment.

[0085] Detailed description of the second buffer chamber (122)

[0086] The second buffer chamber (122) is a secondary buffer structure that additionally disperses the impact energy primarily absorbed by the first buffer chamber (121), and is arranged in a stacked form inside the first buffer chamber (121). The second buffer chamber (122) is configured to prevent the concentration of impact energy by performing the function of dispersing the transmitted pressure in various directions, rather than merely absorbing the impact.

[0087] The second buffer chamber (122) may include a plurality of partition structures or cell structures inside, and each partition deforms sequentially upon impact, thereby dispersing energy in stages. This multi-cell structure improves shock absorption efficiency compared to a single space and can effectively suppress the phenomenon of impact being concentrated at a specific point.

[0088] The second buffer chamber (122) is configured to receive pressure transmitted from the first buffer chamber (121) while simultaneously directing some of the pressure to the pressure discharge chamber (123). To this end, a channel or connecting valve may be formed inside, and the pressure is designed to move naturally according to the impact intensity. Through this structure, impact energy is continuously dispersed rather than remaining in a single chamber.

[0089] The second buffer chamber (122) may be formed of composite materials with different elastic moduli so as to adjust the deformation speed and deformation range according to the collision situation. For example, the outer side may be composed of a material with relatively high elasticity, and the inner side may be composed of a material with viscoelastic properties, so that shock absorption and energy dispersion can be achieved simultaneously.

[0090] The second buffer chamber (122) can be applied to the side protection structure of an industrial AGV as an actual application example, and in the event of a side collision, it takes over the impact absorbed by the first buffer chamber (121) and disperses the impact energy inward to reduce the load transmitted to the equipment frame and internal parts. This minimizes equipment damage and further improves the safety of the operator.

[0091] Detailed description of the pressure relief chamber (123)

[0092] The pressure relief chamber (123) is a structure that controls the rise in internal pressure and maintains the stability of the entire buffering system by gradually releasing the pressure transmitted from the first buffering chamber (121) and the second buffering chamber (122). The pressure relief chamber (123) is positioned between or inside the two buffering chambers and is responsible for the core buffering control function of controlling the pressure flow.

[0093] The pressure relief chamber (123) may include a plurality of discharge paths and flow control valves inside, and is configured so that the discharge speed is automatically adjusted according to the impact intensity. For example, it is designed so that gradual pressure relief occurs at low impacts and rapid relief occurs at high impacts, thereby preventing an excessive rise in internal pressure.

[0094] The pressure relief chamber (123) can perform the function of converting energy into heat or air flow and releasing it to the outside during the pressure relief process, thereby preventing impact energy from accumulating inside. In addition, some structures may be configured to allow for gradual pressure relief by applying porous materials or micro-ventilation holes.

[0095] The pressure relief chamber (123) completes a multi-stage buffering mechanism through coordinated operation with the first buffer chamber (121) and the second buffer chamber (122), and maintains pressure balance between each chamber. This enables stable buffering performance to be maintained even under repeated impact situations and contributes to preventing damage to the chamber structure.

[0096] The pressure relief chamber (123) can be applied to industrial logistics equipment requiring high-speed movement as an actual application example, and performs the function of rapidly releasing internal pressure that increases instantaneously upon collision to the outside to prevent damage to the airbag structure and simultaneously minimize the impact transmitted to the worker. This structure contributes to effectively improving the problems of pressure concentration and damage that occurred in conventional technology.

[0097] Five embodiments of the external cushioning airbag module (120)

[0098] First embodiment: Side hinge deployable multi-stage chamber external cushioning airbag module (120)

[0099] The external cushioning airbag module (120) is hinged to the side frame of the industrial mobility and is normally stored in close contact with the outer surface. When an accident prediction signal is input, it rotates outward and expands, causing the first cushioning chamber (121), the second cushioning chamber (122), and the pressure relief chamber (123) to expand simultaneously. The first cushioning chamber (121) is positioned at the outermost side to absorb the initial impact, the second cushioning chamber (122) is stacked inside to disperse energy, and the pressure relief chamber (123) is positioned to relieve internal pressure in stages.

[0100] The external cushioning airbag module (120) is configured so that the hinge rotation angle and the chamber inflation pressure are controlled in conjunction, and operates with a wider deployment range and higher pressure as the risk level increases. Each chamber is connected by a flow path so that the pressure generated in the first cushioning chamber (121) is transmitted to the second cushioning chamber (122) and then sequentially discharged through the pressure discharge chamber (123).

[0101] The external cushioning airbag module (120) can be applied in a forklift side collision situation and, upon contact with a worker, the first cushioning chamber (121) immediately absorbs the impact, the second cushioning chamber (122) disperses the energy, and the pressure relief chamber (123) releases the residual pressure, thereby providing the effect of minimizing impact transmission.

[0102] Second embodiment: Sliding pull-out multi-stage chamber external cushioning airbag module (120)

[0103] The external cushioning airbag module (120) is housed in a sliding structure within the outer surface of the industrial mobility and is configured such that, upon receiving an accident prediction signal, it is withdrawn in a straight direction, and simultaneously, the first cushioning chamber (121), the second cushioning chamber (122), and the pressure relief chamber (123) are sequentially inflated. The first cushioning chamber (121) is positioned at the leading edge of the withdrawal, the second cushioning chamber (122) is stacked at the rear, and the pressure relief chamber (123) is connected through an internal flow path.

[0104] The external cushioning airbag module (120) is configured such that the withdrawal distance and the chamber inflation amount are linked, so that partial withdrawal and low-pressure inflation are performed when the risk is low, and maximum withdrawal and high-pressure inflation are performed simultaneously when the risk is high. The pressure relief chamber (123) includes a valve that automatically opens as the internal pressure rises, thereby performing stable pressure control.

[0105] The external cushioning airbag module (120) can be applied to the AGV in a narrow logistics passage and is withdrawn only as much as necessary when a worker approaches, thereby maintaining space utilization while effectively reducing impact through a multi-stage chamber structure.

[0106] Third embodiment: Upper pop-up deployable multi-stage chamber external cushioning airbag module (120)

[0107] The external cushioning airbag module (120) is embedded inside the upper frame of the industrial mobility and pops up upward according to an accident prediction signal, and is configured with a structure in which the first cushioning chamber (121), the second cushioning chamber (122), and the pressure relief chamber (123) expand three-dimensionally. The first cushioning chamber (121) is positioned on the upper outer side to absorb the initial impact in the event of a rollover, the second cushioning chamber (122) provides structural support from the inside, and the pressure relief chamber (123) releases pressure in the upward direction.

[0108] The external cushioning airbag module (120) is controlled so that when a rollover risk signal is detected, the pop-up structure operates first, and then each chamber inflates sequentially. In particular, the second cushioning chamber (122) is designed to maintain structural rigidity so that the protective space does not easily collapse, and the pressure relief chamber (123) maintains internal pressure stably.

[0109] The external cushioning airbag module (120) can be applied to a rollover risk situation in a heavy industrial vehicle, and forms an upper protection space to secure a survival space for the driver and worker, and provides the effect of dispersing impact energy in multiple stages.

[0110] Fourth embodiment: Front-extended multi-stage chamber external cushioning airbag module (120)

[0111] The external cushioning airbag module (120) is stored in a foldable structure in the front of the industrial mobility and is configured such that the first cushioning chamber (121), the second cushioning chamber (122), and the pressure relief chamber (123) are continuously inflated as it expands forward according to an accident prediction signal. The first cushioning chamber (121) is positioned at the very front to make direct contact with the collision target, the second cushioning chamber (122) disperses the impact from the rear, and the pressure relief chamber (123) discharges internal pressure to the outside.

[0112] The external cushioning airbag module (120) is designed so that the front surface area expands widely when deployed, thereby increasing the contact area with the collision target and reducing the impact per unit area. Each chamber is interconnected so that pressure is transmitted in stages, and the pressure release chamber (123) automatically opens in an overpressure state to maintain structural stability.

[0113] The external cushioning airbag module (120) can be applied in a frontal collision situation of a logistics vehicle and plays a role in effectively reducing impact energy through a wide cushioning area and a multi-stage chamber structure when colliding with an obstacle or a worker.

[0114] Fifth embodiment: Module split independent deployment type multi-stage chamber external cushioning airbag module (120)

[0115] The external cushioning airbag module (120) is divided into a plurality of independent modules and arranged in the front, side, and corner portions of the industrial mobility, and is configured to selectively deploy only specific areas according to an accident prediction signal. Each module has an independent multi-stage cushioning structure including a first cushioning chamber (121), a second cushioning chamber (122), and a pressure relief chamber (123).

[0116] The external cushioning airbag module (120) is controlled so that only the necessary modules are activated according to the direction of the accident, thereby minimizing energy consumption and preventing unnecessary deployment. Each chamber can be individually pressure-controlled, and the overall cushioning performance is optimized through data linkage between modules.

[0117] The external cushioning airbag module (120) can be applied in complex industrial environments where there is a risk of multi-directional collision, and for collisions occurring in a specific direction, only the module in that direction is deployed to provide efficient protection, and provides the effect of effectively absorbing and dispersing impact through a multi-stage chamber structure.

[0118] Detailed description of the safety execution control unit (130)

[0119] The safety execution control unit (130) is a core control block that controls the actual physical response of the industrial mobility based on accident prediction signals transmitted from the accident prediction detection module (110) and the accident occurrence judgment module (113). The safety execution control unit (130) determines the deployment time, deployment position, and deployment pressure of the external cushioning airbag module (120), and is configured to control the driving state of the industrial mobility at the same time to minimize accident damage.

[0120] The safety execution control unit (130) is designed to selectively perform multiple control scenarios according to the input risk level. For example, at the warning level, deceleration and alarm output are performed first, and at the emergency response level, the airbag is deployed and the driving power is cut off and peripheral equipment interlocking control is performed simultaneously. This stepwise control structure prevents unnecessary malfunctions while enabling a rapid response in actual dangerous situations.

[0121] The safety execution control unit (130) may be implemented based on a high-speed processor and a real-time operating system to ensure real-time control performance, and may include a priority-based task scheduling structure to minimize sensor data processing delay. In addition, it is configured to be linked with an external system through a communication interface to perform integrated safety control of the entire workspace.

[0122] The safety execution control unit (130) is configured to continuously monitor various state information of industrial mobility and dynamically modify the control strategy when the likelihood of an accident changes. For example, when situations such as a sudden change in movement speed, an increase in the angle of inclination, or an increase in the density of surrounding workers are detected, the control parameters are adjusted so that more conservative safety control is performed.

[0123] The safety execution control unit (130) can be applied to an unmanned transport vehicle operated in an automated factory as an actual application example, and performs the role of immediately deploying an airbag and simultaneously decelerating and stopping the vehicle when a worker enters the vehicle's path, and transmitting an alarm signal to surrounding equipment to simultaneously control the danger to the entire workspace. Through this, the spread of accidents can be prevented and worker protection can be maximized.

[0124] The safety execution control unit (130) according to the present invention can, of course, expand the control method to include AI-based judgment, external system linkage, cloud-based control, etc.

[0125] Detailed description of the drive power cutoff module (131)

[0126] The drive power cutoff module (131) is a component that performs the function of immediately cutting off the drive power of the industrial mobility according to the control command of the safety execution control unit (130) to stop physical movement. The drive power cutoff module (131) includes an electronic switch or relay structure installed on the power supply line and is designed to cut off the power with a fast response speed when a dangerous situation occurs.

[0127] The drive power cutoff module (131) may include a soft cutoff function or a step-by-step current reduction control function to prevent electrical shock or system damage that may occur when the power is cut off. This allows for the implementation of a safe stop state while preventing equipment damage caused by sudden power cutoff.

[0128] The drive power cutoff module (131) is configured to be applicable to an electric motor, a hydraulic system, or a mixed drive system depending on the drive method of the industrial mobility. In the case of an electric drive device, a method of cutting off the inverter control signal or cutting off the battery power line may be applied, and in the case of a hydraulic system, the drive of the hydraulic pump may be cut off to induce an immediate cessation of operation.

[0129] The drive power cutoff module (131) may include an auxiliary detection function capable of determining its own safety conditions in addition to the signal transmitted from the safety execution control unit (130), and is configured to independently perform power cutoff when an exception occurs. This allows for maintaining a minimum level of safety function even in the event of a system error or communication failure.

[0130] The drive power cutoff module (131) can be applied to an industrial forklift moving at high speed as an actual application example, and provides the effect of preventing a secondary collision between the worker and the equipment by immediately suppressing further movement of the vehicle by cutting off the power simultaneously with the deployment of the airbag in situations with a high risk of collision. This structure effectively complements the active stop control function that was lacking in the prior art.

[0131] Detailed description of the deceleration warning control module (132)

[0132] The deceleration warning control module (132) performs the function of actively reducing the movement speed of industrial mobility according to the control command of the safety execution control unit (130) and simultaneously outputting a warning signal to alert surrounding workers and drivers to a dangerous situation. The deceleration warning control module (132) is configured to go beyond a simple speed reduction function and to variably control the deceleration pattern and warning method according to the risk level.

[0133] The deceleration warning control module (132) is configured to gradually reduce the speed of the industrial mobility by adjusting the electric motor control signal or controlling the flow rate of the hydraulic system. In this process, to prevent the falling of loads or equipment instability caused by sudden deceleration, the deceleration curve is designed to be performed according to a preset profile.

[0134] The deceleration warning control module (132) is configured to clearly convey dangerous situations by utilizing a combination of visual, auditory, and tactile warning means. For example, the warning light, warning sound output, and vibration notification device operate simultaneously so that the operator can intuitively recognize the danger. In addition, the intensity and frequency of the warning signal are set to change in stages according to the level of danger.

[0135] The deceleration warning control module (132) can be configured to control the direction of the warning by reflecting the surrounding environment and the location information of the worker. For example, by focusing a stronger warning signal on a worker approaching from a specific direction, it is possible to reduce unnecessary warnings while increasing the effectiveness of actual risk response.

[0136] The deceleration warning control module (132) can be applied to an AGV operated in a logistics center as an example of actual application, and provides the effect of gradually reducing the speed and simultaneously outputting a warning sound when a worker approaches the vehicle's path of travel, thereby inducing the worker to avoid the collision, thereby reducing the possibility of a collision in advance.

[0137] Detailed explanation of the accident spread prevention module (133)

[0138] The accident spread prevention module (133) performs the function of preventing secondary accidents and chain accidents by controlling the surrounding industrial equipment and other mobility beyond the response of industrial mobility alone in situations where the probability of an accident is high. The accident spread prevention module (133) is configured to exchange information with various equipment within the workspace via wired or wireless communication.

[0139] The accident spread prevention module (133) transmits a deceleration or stop command to adjacent industrial mobility when a dangerous situation is detected, and controls the operation to temporarily stop in conjunction with conveyor systems, robot arms, automation equipment, etc. This effectively prevents a single accident from spreading to the entire work process.

[0140] The accident spread prevention module (133) is configured to dynamically set safety zones in units of work zones and to transmit collective control signals to all equipment within the zone. For example, if a collision risk occurs in a specific zone, the entire zone can be switched to safety mode to suppress the occurrence of additional risks.

[0141] The accident spread prevention module (133) may include a distributed control structure or a local control function to maintain stable control even in the event of communication failure or abnormal operation of some equipment. This reduces dependence on the central control system and enables reliable safety control in various industrial environments.

[0142] The accident spread prevention module (133) can be applied to an environment in a smart factory where multiple autonomous logistics equipment is operated simultaneously as an example of actual application, and when an accident risk is detected in a specific piece of equipment, it provides the effect of preventing chain collisions by simultaneously decelerating or stopping surrounding equipment.

[0143] Detailed description of the accident record generation module (140)

[0144] The accident record generation module (140) collects and stores various operational data transmitted from the accident prediction detection module (110), the safety execution control unit (130), and the external cushioning airbag module (120), and performs the function of systematically recording the situation before and after an accident occurs. The accident record generation module (140) is configured as a core data generation device for industrial safety management and accident analysis.

[0145] The accident record generation module (140) is configured to record various data in high resolution, including the accident prediction time, the airbag deployment time, the movement speed of the industrial mobility, the distance to the collision target, the duration of the impact, and the control signal history. This data is arranged by time to enable precise reproduction of the accident flow.

[0146] The accident record generation module (140) can be configured to transmit stored data to an external server or cloud system so that long-term storage and analysis are possible. Through this, multiple accident data can be integrated and analyzed to derive risk patterns, which can then be utilized for future safety policies and system improvements.

[0147] The accident record generation module (140) may include encryption and tamper-proof functions to ensure data reliability and is designed to be used as objective evidence in legal disputes or accident liability analysis. Additionally, data security can be enhanced through an access control management function.

[0148] The accident record generation module (140) can be used to analyze the cause of a collision accident that occurred at an industrial site as an example of actual application, and can contribute to establishing measures to prevent recurrence by clearly identifying the circumstances of the accident based on sensor data and control history immediately before the accident occurred. This structure effectively complements the systematic accident data management function that was lacking in conventional technology.

[0149] Below, an industrial mobility external buffer safety execution method (S100) using a smart accident prediction-based industrial mobility external buffer safety execution system (100) is described in detail.

[0150] The industrial mobility external cushioning safety execution method (S100) is a series of control processes designed to predict risks and respond proactively prior to an accident. First, in the data detection step (S110), information regarding the surrounding environment and objects is acquired in real time through radar sensors, ultrasonic sensors, image-based detection sensors, etc. Subsequently, in the data analysis step (S120), the distance, relative speed, and approach direction are analyzed to quantify the probability of a collision. Then, in the accident occurrence judgment step (S130), a risk level is calculated based on the analyzed data, and the possibility of an accident occurring is determined. In the safety execution step (S140), the deployment position, deployment time, and deployment pressure of the external cushioning airbag module are controlled according to the judgment result, and safety controls such as deceleration or power cutoff are performed simultaneously. Additionally, in the accident spread prevention step (S150), control is performed to prevent further spread of the accident by linking with surrounding industrial facilities and adjacent mobility. Finally, in the accident record generation step (S160), accident prediction signals, control history, and impact-related data are stored so that they can be utilized for future analysis and safety management.

[0151] Detailed explanation of the data detection step (S110)

[0152] The data detection step (S110) is the starting step of the industrial mobility external buffer safety execution method (S100), and is a step of securing basic data in real time regarding the driving state of the industrial mobility and various objects, equipment, workers, and obstacles present in the surrounding work environment. In the data detection step (S110), the data detection module (111) included in the accident prediction detection module (110) acts as the center and acquires external environment information from various angles using one or more of a radar sensor, an ultrasonic sensor, an image-based detection sensor, an infrared-based proximity detection sensor, a LiDAR-based distance detection sensor, or a non-contact detection means equivalent thereto. At this time, it is preferable that the data detection step (S110) is configured to collect multidimensional data that includes the location, distance, relative approach speed, direction of movement, size, occupied area, and change amount between consecutive frames of each object, rather than merely confirming the presence of an object.

[0153] In the data detection step (S110), detection environment correction considering the specific characteristics of the industrial site may be performed. Since factors such as dust, water vapor, reflected light, low light intensity, multiple reflections caused by metal structures, vibrations of moving equipment, and noise interference frequently occur in actual industrial sites, using raw data directly acquired from the sensor as is may result in false detection or non-detection. Accordingly, the data detection step (S110) includes preprocessing operations corresponding to the characteristics of each sensor; for example, clutter removal and reflected signal matching are performed for radar data, abnormal reflected wave removal and distance value stabilization are performed for ultrasonic data, and exposure correction, distortion correction, backlight correction, object contour enhancement, and noise removal are performed for image data. Through this structure, the data detection step (S110) can reliably generate highly reliable detection data that can be utilized in subsequent steps.

[0154] The data detection step (S110) may also include a process for ensuring temporal consistency and spatial alignment of data acquired from multiple sensors. Since there is a possibility that the same object may be recognized as a different object in the subsequent data analysis step (S120) when the radar sensor and the camera sensor detect the same worker at different points in time and based on different coordinates, the data detection step (S110) may be configured to synchronize the sampling cycle of each sensor with a reference time and perform coordinate transformation based on the sensor mounting position and installation angle so that the same object is represented in a single integrated coordinate system. In addition, the data detection step (S110) minimizes blind spots by distributing sensors at multiple locations of the industrial mobility, such as the front, side, rear, and upper frame, and enables rapid detection even in corner sections where pinch accidents are frequent or side areas where workers frequently access.

[0155] In the data detection step (S110), information on the operational status of the industrial mobility itself can also be acquired. In other words, if internal operational data such as the industrial mobility's current speed, steering angle, driving direction, acceleration / deceleration status, loading status, weight change, body inclination angle, and braking capability status are detected together with external object information, the precision of the actual risk assessment can be increased even under the same external conditions. For example, even if a worker approaches from the same distance, the level of accident risk differs significantly depending on whether the industrial mobility is moving straight at low speed or turning at high speed; therefore, it is effective for the data detection step (S110) to be configured to transmit this internal driving status along with external detection information to subsequent steps. Such a detection structure goes beyond simple environmental detection and performs the role of generating a basic situational model for assessing the possibility of an accident.

[0156] As an actual application example of the data detection step (S110), consider a situation where an AGV operating in an automated logistics warehouse moves along a narrow passage. In this case, the data detection step (S110) can continuously detect the distance to a front obstacle using a front radar sensor, measure the proximity to a rack structure using a side ultrasonic sensor, identify the presence of a worker and the direction of movement using a front camera, and acquire signs of overturning risk due to uneven loading of the load using a vehicle body tilt sensor. Additionally, if a worker suddenly enters a cross passage, the possibility of crossing the movement path can be detected early by rapidly capturing changes in position between consecutive frames. As such, the data detection step (S110) is an important step that captures complex and variable risk factors in the industrial site in real time, thereby providing basic data that allows for the specific calculation of the possibility of an accident in the subsequent data analysis step (S120).

[0157] Detailed explanation of the data analysis step (S120)

[0158] The data analysis step (S120) is a step that integrates and interprets various sensor data and operational status information obtained in the data detection step (S110) to analyze the interaction relationship between industrial mobility and surrounding objects and quantitatively derives the possibility of an accident occurring. The data analysis step (S120) can be performed by a data analysis module (112) and includes a prediction-based analysis structure that calculates not only the change in relative position between objects, approach direction, relative speed, movement trajectory, expected intersection point, and expected collision time, rather than simply reading the current distance value. Accordingly, the data analysis step (S120) functions as a key step that goes beyond static space occupancy judgment and identifies in advance the possibility of collision, pinching, entrapment, side contact, or rollover risk that may occur within hundreds of milliseconds to several seconds.

[0159] In the data analysis step (S120), multiple pieces of information regarding the same object are first integrated through sensor fusion. For example, if a camera recognizes the shape of a person, a radar measures the approach speed of the object, and an ultrasonic sensor provides a precise value of the proximity distance, the data analysis step (S120) can integrate this information into a single object profile to determine whether the object is a worker, a moving piece of equipment, a stationary obstacle, or a temporary load. Additionally, the data analysis step (S120) can apply risk weights differentially according to the type of object; for instance, the possibility of contact with a worker can be reflected as the highest risk, while the possibility of gentle contact with a temporarily stopped pallet can be reflected as a relatively low risk. Such classification by object type and reflection of weights allows for a reduction in unnecessary safety procedures while enabling a priority response to situations with a high actual risk of accident.

[0160] The data analysis step (S120) may include a function to track the movement trajectory of an object and predict its future location based on continuously acquired time-series data. To this end, practical analysis techniques such as Kalman filters, extended Kalman filters, particle filters, moving average-based speed prediction, trajectory regression models, and simple neural network-based path prediction models may be applied. For example, if a worker crossing from the right to the left of an industrial mobility is detected, the data analysis step (S120) calculates not only the current location but also the predicted location of movement at the next point in time, and determines whether it overlaps with the current steering path of the industrial mobility. In addition, even if the objects are within the same distance, the probability of collision varies significantly depending on whether the object is moving away from the industrial mobility, moving in parallel in the same direction, or approaching head-on; therefore, the data analysis step (S120) calculates the risk more precisely by considering the approach direction and the speed vector together.

[0161] In the data analysis step (S120), the dynamic state and braking capability of the industrial mobility itself may also be reflected. For example, in situations where the current speed is high, the road surface is slippery, or the braking distance increases due to the load, the actual risk of an accident can be significantly higher even under the same object distance conditions. Therefore, the data analysis step (S120) may have a structure that corrects the risk index by considering the vehicle body speed, acceleration, steering angle, load, vehicle body center of gravity shift, inclination angle, and braking response capability together. Furthermore, in the case of rollover risk analysis, since the risk can increase if sharp turns and uneven loading occur simultaneously even without an external collision object, it is desirable for the data analysis step (S120) not to be limited to simple collision analysis but to generate complex risk judgment information combined with driving stability analysis.

[0162] As an example of the actual application of the data analysis step (S120), a situation can be assumed where a forklift transports loaded cargo while turning a corner in a large logistics center. In this case, the data analysis step (S120) detects the movement of a worker approaching from behind the corner through a front camera and a corner side sensor, calculates the relative approach speed through radar data, and, by reflecting the forklift's current turning radius and speed together, can conclude that there is a high probability of a collision within approximately 1 second. As another example, in the case of industrial mobility descending a slope with a heavy load and a side obstacle approaching, the data analysis step (S120) can assess a high probability of a complex accident involving rollover rather than a simple contact. In this way, the data analysis step (S120) quantifies the probability of an accident by reflecting the actual operational context of the industrial site and provides key judgment grounds so that the subsequent accident occurrence judgment step (S130) can quickly and rationally decide whether to deploy an external cushioning airbag and whether to decelerate or stop.

[0163] Detailed explanation of the accident occurrence judgment stage (S130)

[0164] The accident occurrence judgment step (S130) is a key step that finally determines the likelihood of an actual accident occurring based on the integrated risk index derived in the data analysis step (S120) and decides the corresponding response level. The accident occurrence judgment step (S130) calculates the risk level by synthesizing complex factors such as distance, relative speed, approach direction, object type, driving status, and braking possibility, and determines whether an accident is likely to occur by comparing this with a pre-set threshold standard. This step functions not merely as a warning judgment, but as a decision-making step intended to induce an actual physical response.

[0165] The accident occurrence judgment stage (S130) can be configured to classify and judge the level of risk into multiple stages, for example, subdivided into a caution stage, a warning stage, and an emergency response stage. In the caution stage, preparation for deceleration and warning output may be performed; in the warning stage, deceleration and direction control are carried out; and in the emergency response stage, it is configured to link the immediate deployment of the external cushioning airbag module (120) and the cutoff of the driving power. This stepwise judgment structure enables rapid response in actual dangerous situations while minimizing unnecessary malfunctions.

[0166] The accident occurrence judgment step (S130) may include time-based judgment elements, and in particular, improves judgment accuracy by utilizing time-space composite indicators such as the estimated time to collision (TTC) or the estimated collision distance. For example, even under the same distance conditions, the risk level is configured to increase sharply when the relative speed is fast, and conversely, to decrease the risk level when the relative speed is low or the distance increases. Additionally, a method to increase the risk judgment sensitivity may be applied when the object's movement pattern is irregular or a sudden change in direction is detected.

[0167] The accident occurrence judgment step (S130) is configured so that judgment criteria can be dynamically adjusted according to the operating environment of industrial mobility. In environments with high worker density, conservative criteria are applied to make risk judgments more sensitive, whereas in environments with restricted access, judgment criteria may be relaxed to reduce unnecessary alarms. Additionally, in situations where sensor reliability is low, multi-sensor matching conditions can be strengthened to prevent unnecessary safety actions caused by false detections.

[0168] As an actual application example of the accident occurrence determination step (S130), a situation may be cited where a worker rapidly approaches from the side while the AGV is passing through an intersection section. In this case, the accident occurrence determination step (S130) determines the emergency response step by combining conditions such as the TTC decreasing below a threshold value, the approach direction matching the collision path, and the object type being determined to be a worker. Accordingly, a control signal is transmitted to the safety execution step (S140) immediately so that the external cushioning airbag deployment and deceleration control are performed simultaneously.

[0169] Detailed description of the safety execution step (S140)

[0170] The safety execution step (S140) is a step for performing actual physical safety measures according to the risk level determined in the accident occurrence judgment step (S130), and integrates the deployment of the external cushioning structure and driving control of the industrial mobility. The safety execution step (S140) operates in conjunction with the external cushioning airbag module (120), the drive power cutoff module (131), the deceleration warning control module (132), and the accident spread prevention module (133), and performs the function of reducing impact and preventing secondary damage before an accident occurs.

[0171] The safety execution step (S140) precisely controls the deployment position, deployment timing, and deployment pressure of the external cushioning airbag module (120) according to the risk level. For example, if the risk of a side collision is high, only the airbag in that direction is selectively deployed, and if the risk of a frontal collision is high, the front airbag is deployed preferentially. In addition, low-pressure inflation is performed in low-speed situations, and high-pressure inflation is performed in high-speed collision risk situations.

[0172] The safety execution step (S140) is configured to perform deceleration or stopping in conjunction with the driving control of industrial mobility. Gradual speed reduction is achieved through the deceleration warning control module (132), and in emergency situations, immediate stopping is achieved through the drive power cutoff module (131). During this process, a control profile is applied to prevent sudden impacts by considering the stability of the load and the structure of the equipment.

[0173] The safety execution step (S140) can simultaneously perform warnings and interlocking control for workers and surrounding equipment. It notifies workers of dangerous situations through warning lights, warning sounds, vibration signals, etc., and simultaneously controls risks in the entire work environment by stopping surrounding equipment or switching to safety mode through the accident spread prevention module (133). This integrated control structure contributes to ensuring safety of the entire workspace beyond the level of a single piece of equipment.

[0174] As an example of the actual application of the safety execution step (S140), a situation in a logistics center where there is a high probability that a forklift will collide with a front obstacle can be cited. In this case, the safety execution step (S140) immediately deploys the front external cushioning airbag module (120), simultaneously rapidly reduces the vehicle speed, and outputs a warning signal to nearby workers. In addition, control to stop adjacent AGVs is simultaneously performed, so that a single collision accident can be effectively prevented from spreading into a chain of accidents.

[0175] Detailed explanation of the accident spread prevention stage (S150)

[0176] The accident spread prevention step (S150) is a step of performing additional interlocking control to prevent the spread of a dangerous situation to surrounding equipment, adjacent industrial mobility, worker-dense areas, or subsequent processes after primary safety measures for the industrial mobility itself have been performed in the safety execution step (S140). Since a single risk of collision or entrapment in an industrial site is not limited to a single piece of equipment but can lead to sudden stopping of adjacent equipment, falling of cargo, collision of continuous driving equipment, or malfunction of automation equipment, the accident spread prevention step (S150) plays the role of simultaneously readjusting the operating state of the entire surrounding system around the point where the danger occurs. To this end, the accident spread prevention step (S150) can be configured to be linked with other equipment, alarm devices, access control equipment, and process control systems within the work area via a wired or wireless communication network centered on the accident spread prevention module (133).

[0177] In the accident spread prevention stage (S150), a dynamic safety zone is first established based on the location where the risk occurred, and differential control commands can be output for equipment and moving objects included within the safety zone. For example, if other industrial mobility is present within the radius of a specific AGV or forklift, a command to immediately decelerate or stop is transmitted; in areas connected to automatic doors or shutters, an access blocking signal is output; and in cell processes connected to conveyors or robot arms, control can be exercised to perform a temporary stop or return to a safe position. This structure has technical significance in that it blocks a chain reaction to prevent the accident from spreading to surrounding processes, rather than merely stopping a single piece of equipment involved in the accident. In particular, in smart logistics warehouses or assembly production lines where workers and equipment are mixed, the organic interlocking control of the accident spread prevention stage (S150) can significantly improve the safety of the entire site.

[0178] It is desirable that the accident spread prevention step (S150) be configured to apply different spread prevention scenarios depending on the level of risk and the type of accident. For example, in the case of a simple low-speed proximity risk, response may be limited to illuminating warning lights around the equipment and broadcasting instructions to restrict access; however, if there is a high probability of worker entrapment or a high probability of high-speed collision, more active spread prevention measures may be performed, such as stopping all surrounding equipment, activating process interlocks, restricting power to the area, and transmitting emergency alerts. Additionally, in situations involving a risk of overturning, methods to restrict access to the upper transfer device or surrounding loading racks in the area may also be applied, taking into account the possibility of falling objects from above. As such, the accident spread prevention step (S150) can be configured as a multi-layered control step that considers not only the type of accident cause but also the type of secondary accident that may spread after the accident.

[0179] The accident spread prevention step (S150) can be configured to operate reliably even in exceptional situations such as communication failures, non-response of some equipment, or control delays. That is, even if the connection status with the central control server is unstable, it must be possible to perform minimal safety stop control through the local controller, and if a specific facility does not respond to a control command, supplementary actions such as bypassing the stop of equipment surrounding that facility, expanding the warning range, and enhancing worker notifications may be performed. Additionally, the accident spread prevention step (S150) may include a return control logic that determines when to return surrounding equipment to normal mode by re-detecting whether the risk has been resolved. This prevents unnecessarily long work interruptions while maintaining a protection state until safety is fully ensured.

[0180] As an actual application example of the accident spread prevention step (S150), consider a situation where an emergency safety execution occurs when a worker enters the front of an industrial mobility vehicle traveling at high speed in an automated logistics center. In this case, the accident spread prevention step (S150) can deploy the front cushioning airbag of the equipment, simultaneously transmit a stop command to another AGV traveling in an adjacent passage, temporarily stop the conveyor in the area, and activate a ceiling warning light and alarm broadcast to block access by surrounding workers. As another example, if a risk of side rollover is detected in a heavy-loading forklift, the robot arm operation can be stopped along with a signal prohibiting access to the nearby loading area, and the danger zone display can be automatically updated on the work management server. In this way, the accident spread prevention step (S150) is a step that effectively implements active safety assurance functions at the entire site level, which were lacking in conventional technology, by rapidly eliminating factors that escalate a single accident into a complex disaster.

[0181] Detailed explanation of the accident record generation step (S160)

[0182] The accident record generation step (S160) is a step that systematically stores sensor data, analysis results, judgment information, and control history generated during the entire safety execution process from the data detection step (S110) to the accident spread prevention step (S150), and generates record data capable of objectively reproducing the situation before and after the accident. In conventional technology, there were limitations in clearly proving the actual circumstances of the accident, the appropriateness of the risk judgment, the timing of the operation of the safety device, and the legitimacy of the control process, as only simple alarm occurrences or equipment stop history were often recorded. In contrast, the accident record generation step (S160) provides a highly reliable recording system that can be utilized for future accident analysis, safety improvement, legal proof, and regulatory compliance by integrating and storing data generated from the accident prediction detection module (110), the safety execution control unit (130), the external cushioning airbag module (120), and the accident spread prevention module (133) based on the time axis.

[0183] In the accident record generation step (S160), raw sensor data and preprocessed data immediately before and after the accident may be stored together. Specifically, this may include radar distance values, ultrasonic reflection signals, video frames or object recognition results, proximity sensor outputs, the driving speed, steering angle, acceleration, inclination angle, loading status, and braking command timing of the industrial mobility. It is preferable that this data be stored along with timestamps in milliseconds. Additionally, the integrated risk index calculated in the data analysis step (S120), the estimated collision time, the approach direction vector, the object type classification result, and the risk grade and threshold comparison result determined in the accident occurrence judgment step (S130) may also be stored together. Through this, a structure is established that goes beyond merely recording the fact that an accident occurred, enabling the backtracking of why such a judgment was made.

[0184] The accident record generation step (S160) also includes a function to record control history related to safety execution in detail. For example, the timing of the deployment command of the external cushioning airbag module (120), the actual start of inflation, the deployed location, deployment pressure, deployment volume, whether the driving power is cut off, the warning output method of the deceleration warning control module (132), and the type and response status of the control command transmitted to surrounding equipment during the accident spread prevention step (S150) may be recorded together. This control history is very important for verifying whether accident prevention actions were performed in a timely manner, whether some devices were delayed, and whether the control command was executed normally. In addition, if the same type of dangerous situation is repeated, past records can be analyzed and utilized as learning data to improve the deployment pressure or deceleration algorithm.

[0185] The accident record generation step (S160) may include a structure to ensure the integrity and security of the stored data. Since records related to industrial accidents may be utilized for subsequent liability analysis, insurance processing, legal disputes, safety audits, and reporting to public institutions, it is desirable for the record data to include structures for encrypted storage, electronic signature assignment, hash-based verification of forgery, access control management, and external server backup. Additionally, the accident record generation step (S160) may be configured to enable double or triple storage by linking not only with internal storage devices of field equipment but also with cloud servers, central control systems, safety management databases, etc. Such a structure ensures that record data is stably preserved even if field equipment is damaged immediately after an accident, and enables long-term safety history analysis.

[0186] As an actual application example of the accident record generation step (S160), a situation can be cited where a forklift deploys a front external cushioning airbag due to a risk of collision with a worker in a corner section. In this case, the accident record generation step (S160) can record the entire series of processes, including the camera recognition result at the time of worker entry, the radar-based distance change value, the TTC calculation result, the time of risk level increase, the time of deceleration command output, the time of airbag deployment completion, whether a stop command for surrounding AGVs is transmitted, and the time of worker avoidance completion. As another example, even in a close accident case where there was a risk of rollover but no actual rollover occurred, the change in vehicle body inclination, load bias, whether the upper airbag deployed, and power cut-off history can be stored to enhance preventive design for similar situations in the future. As such, the accident record generation step (S160) is a step that significantly improves the reliability and reproducibility of industrial safety management by providing an objective and precise safety history management basis that was lacking in conventional technology.

[0187] As explained above, the smart accident prediction-based industrial mobility external cushioning safety execution system according to the present invention overcomes the limitations of conventional technology, which focused on responding after an accident occurs, and provides a structure capable of actively responding by predicting risks at a pre-accident stage, thereby having the effect of improving the fundamental safety level in industrial sites. While conventional methods primarily involved absorbing impact through internal structures after a collision, the present invention has the effect of shifting the technological paradigm toward reducing the risk of the collision itself by preemptively deploying an external cushioning structure.

[0188] Furthermore, it overcomes the limitations of sensor-based safety functions, which were limited to simple detection in conventional technology. By comprehensively analyzing distance, relative speed, and approach direction to precisely determine the likelihood of an accident, it effectively prevents hazardous situations that could lead to actual accidents. This enables the securing of effective safety in industrial settings by implementing a structure that links to actual physical responses beyond simple warnings.

[0189] In addition, the present invention is configured to selectively deploy external cushioning airbag modules on the front, sides, and top of industrial mobility, thereby solving the problem of limited protection range in conventional technology and effectively responding to collision and entrapment situations occurring from various directions. In particular, it has the advantage of securing a survival space for drivers and workers by forming a protective space in the upward direction even in situations of rollover risk.

[0190] Furthermore, by variably controlling the deployment pressure and volume of the airbag according to the magnitude of the collision risk and the movement speed of industrial mobility, it is possible to simultaneously resolve the problems of excessive impact transmission or insufficient cushioning caused by the uniform cushioning structure in conventional technology. Through this, stable impact reduction performance can be secured even in various work environments and operating conditions.

[0191] In addition, by enabling continuous initial shock absorption, energy dispersion, and stepwise pressure release through a multi-stage buffering structure composed of a first buffering chamber, a second buffering chamber, and a pressure release chamber, it is possible to effectively mitigate the shock concentration phenomenon that occurred due to a single buffering structure in conventional technology. This contributes to minimizing the impact transmitted to the human body and equipment during a collision.

[0192] In addition, by performing drive power cutoff, deceleration control, alarm output, and surrounding equipment interlocking control through the safety execution control unit, it is possible to effectively prevent secondary damage that spreads after an accident occurs in conventional technology. In particular, by controlling the entire workspace simultaneously through interlocking with surrounding equipment, high safety can be ensured even in complex industrial environments.

[0193] In addition, by automatically recording various operational data such as accident prediction detection signals, airbag deployment timing, movement speed, and impact duration, it is possible to systematically secure accident analysis and safety management data that was lacking in conventional technology. Through this, objective data that can be utilized for establishing industrial safety policies, identifying the causes of accidents, and legal responses can be provided, and the present invention can effectively solve the problems of conventional technology.

[0194] The above detailed description of the present invention describes only specific embodiments thereof. However, it should be understood that the present invention is not limited to the specific forms mentioned in the detailed description, but rather should be understood to include all variations, equivalents, and substitutions within the spirit and scope of the invention as defined by the appended claims.

[0195] In other words, the present invention is not limited to the specific embodiments and descriptions described above, and any person skilled in the art to which the present invention pertains can make various modifications without departing from the essence of the invention as claimed in the claims, and such modifications fall within the scope of protection of the present invention. Explanation of the symbols

[0196] 100: Smart Accident Prediction-Based Industrial Mobility External Buffer Safety Execution System 110: Accident Prediction Detection Module 111: Data detection module 112: Data Analysis Module 113: Accident Occurrence Judgment Module 120: External shock absorber airbag module 121: First buffer chamber 122: Second buffer chamber 123: Pressure relief chamber 130: Safe Execution Control Unit 131: Driving power cutoff module 132: Deceleration Warning Control Module 133: Accident Spread Prevention Module 140: Accident Record Generation Module S100: Industrial Mobility External Buffer Safety Execution Method S110: Data detection step S120: Data analysis stage S130: Accident occurrence determination stage S140: Safe execution phase S150: Accident Spread Prevention Stage S160: Accident record generation step

Claims

Claim 1 A system comprising software function blocks executed on a computer device including a processor, memory, storage device, input / output interface, and network interface, comprising at least one of a radar sensor, an ultrasonic sensor, an image-based detection sensor, or a proximity detection sensor to detect the driving path and surrounding environment of an industrial mobility, and an accident prediction detection module (110) that determines the probability of an accident by analyzing the distance, relative speed, and approach direction to a potential collision target; an external cushioning airbag module (120) having a plurality of cushioning airbags that are disposed at at least one location among the front frame, side protection structure, or upper protection frame of the industrial mobility and are selectively deployed according to an accident prediction signal; and a safety execution control unit (130) that determines a risk level based on a signal input from the accident prediction detection module and controls the deployment time and deployment position of the external cushioning airbag module (120). The accident record generation module (140) automatically stores operational data including an accident prediction detection signal, an airbag deployment time, the movement speed of the industrial mobility, the duration of impact, and a safety execution history; wherein the accident prediction detection module (110) includes a data detection module (111) that acquires detection data in real time and transmits it to a data analysis module (112) by including at least one of a radar sensor, an ultrasonic sensor, an image-based detection sensor, or a proximity detection sensor to detect the driving path and surrounding environment of the industrial mobility; and a data analysis module (112) that receives the data transmitted from the data detection module (111) in real time and analyzes the distance, relative speed, and approach direction to a potential collision target.The accident occurrence judgment module (113) operates by determining the likelihood of an accident based on the results analyzed from the data analysis module (112), transmitting an accident prediction signal to the external cushioning airbag module (120), the safety execution control unit (130), and the accident record generation module (140), analyzing the type of detected object based on the results analyzed from the data analysis module (112), and determining the likelihood of an accident by applying risk weights for each object, wherein a relatively high risk weight is assigned to people and a relatively low weight is applied to simple obstacles, and the risk is determined by evaluating the reliability of data obtained from multiple sensors, wherein reliability is calculated based on the state of each sensor, environmental conditions, and data consistency, and a greater weight is assigned to data with high reliability to improve judgment accuracy, and if the reliability of the image sensor is below a preset standard due to insufficient lighting, judgment is performed based on radar and ultrasonic sensor data, and conversely, if noise in the radar signal exceeds a preset standard, image-based data is reflected preferentially; and the external cushioning airbag module (120) is a hinge-coupled type or It is structured to be stored on the outer surface of an industrial mobility by a sliding storage structure and deployed outward by a gas generator or a compressed gas cartridge, and the deployment pressure or deployment volume is variably controlled according to the magnitude of the collision risk or the movement speed of the industrial mobility, and is configured to deploy upward according to a change in the inclination angle of the industrial mobility or a rollover risk signal to secure a driver protection space, and is divided and arranged into a plurality of independent modules on the front, side, and corner portions of the industrial mobility, and is configured to selectively deploy only a specific area according to an accident prediction signal, and the external cushioning airbag module (120) comprises: a first cushioning chamber (121) having a structure that absorbs initial impact; and a second cushioning chamber (122) having a structure that is mounted in a stacked form with the first cushioning chamber (121) and sprays the absorbed impact energy.The smart accident prediction-based industrial mobility external buffer safety execution system is characterized by comprising: a pressure discharge chamber (123) that discharges pressure in stages, which is stacked and arranged in a manner that binds the first buffer chamber (121) and the second buffer chamber (122) together; and the safety execution control unit (130) comprises: a drive power cutoff module (131) that automatically cuts off the drive power of the industrial mobility according to a preset standard; a deceleration warning control module (132) that controls the industrial mobility to perform deceleration control or output an alarm signal according to a preset standard; and an accident spread prevention module (133) that prevents the spread of accidents by controlling the operation of surrounding industrial facilities in conjunction through wireless or wired communication. Claim 2 delete Claim 3 delete Claim 4 delete Claim 5 delete

Images