Elvator fault detection method and server performing thereof

Abstract

An elevator fault detection server according to the present invention comprises: a data collection unit that collects elevator operation data from a plurality of sensors; an abnormality indication determination unit that determines whether there is an abnormality by comparing the operation data with a reference value; an inspection item matching unit that automatically matches an inspection item corresponding to the abnormality indication; an inspection guide generation unit that automatically generates an inspection guide for each inspection item; and a mobile inspection interface unit that provides the inspection guide to a user terminal and stores input information.

Description

Technology Field

[0001] The present invention relates to an elevator fault detection method and a server for executing the same. More specifically, it relates to an elevator fault detection method and a server for executing the same that enable proactive prevention-oriented maintenance rather than reactive response by analyzing operation data collected from elevator sensors to automatically detect signs of a fault or a state exceeding a standard. Background Technology

[0002] With modern urbanization and the rise of high-rise buildings, elevators have established themselves as an essential means of transportation in residential and industrial infrastructure; consequently, ensuring their safety, operational efficiency, and reliability has become a critical task. In particular, in environments where elevators must operate without interruption—such as hospitals, apartment complexes, large commercial facilities, and logistics centers—the proactive prevention of breakdowns and real-time status monitoring are key maintenance challenges.

[0003] Conventional elevator maintenance methods have largely relied on preventive maintenance, performed according to regular inspection cycles set by manufacturers or maintenance operators, or corrective maintenance, which involves responding to breakdowns after they occur. However, these traditional methods have the following problems and limitations.

[0004] First is the uniformity of the periodic inspection cycle. Although the likelihood of failure and the rate of component wear can vary depending on the installation environment, usage frequency, and structural characteristics, the standard periodic inspection criteria fail to account for these variables, resulting in the repetitive checking of the same items at fixed intervals. Consequently, unnecessary inspections are performed, or items that are actually on the verge of failure are overlooked.

[0005] Second, there are limitations to the reactive approach. If maintenance is carried out after an elevator breaks down, the equipment becomes unavailable during that period, leading to user inconvenience, complaints, and an increased risk of safety accidents. This can result in severe work delays or service disruptions, particularly in facilities where downtime due to a breakdown is critical, such as hospitals or logistics centers.

[0006] Third, there is the variation and non-standardization of maintenance quality. As inspections often depend on the proficiency of field technicians, quality variations exist among workers due to omissions in inspection items, differences in inspection levels, and inaccuracies in records. In particular, methods such as paper checklists, handwritten records, and manual-based inspections make it difficult to integrate with real-time analysis or data-driven maintenance systems, and the reliability of follow-up management is low.

[0007] Fourth, there is a disconnect in failure and maintenance history. Since information such as failure types, conditions of occurrence, and maintenance methods is not systematically accumulated, it is difficult to analyze the causes and take proactive measures when recurring problems occur. Furthermore, the lack of accumulation of unique abnormal sign patterns for each piece of equipment makes it difficult to establish a data-based predictive maintenance system.

[0008] To address this, some companies are introducing IoT sensors, monitoring systems, and anomaly diagnosis algorithms, but these have limitations. Specifically, they fail to precisely analyze failure patterns because anomalies are determined solely by whether threshold values ​​are exceeded. Furthermore, even after a failure is detected, relevant inspection items, corrective measures, and procedures are not automatically presented; ultimately, work relies on the judgment of technicians, resulting in inspection quality and reporting methods remaining largely manual.

[0009] Furthermore, even when a fault is detected, the methods of providing relevant information to field technicians are limited to phone calls, text messages, or simple notifications, resulting in a lack of systems that provide real-time guidance on specific inspection guides, necessary tools, risk factors, and inspection procedures. Prior art literature

[0010] Korean Registered Patent 10-0875778 The problem to be solved

[0011] The present invention aims to provide an elevator fault detection method and a server for executing the same, which enable proactive prevention-oriented maintenance rather than reactive response by analyzing operation data collected from elevator sensors to automatically detect signs of failure or conditions exceeding standards.

[0012] In addition, the present invention aims to provide an elevator fault detection method and a server for executing the same, which can standardize the quality of technician work and ensure safety by automatically deriving relevant inspection items based on detected abnormal signals and automatically generating an inspection guide for each item that includes inspection procedures, necessary tools, precautions, etc.

[0013] In addition, the present invention aims to provide an elevator fault detection method and a server that executes the same, which enable a technician to check inspection items and guides in real time through a mobile terminal at the site, and to electronically record and transmit inspection results to a server, thereby realizing the digital transformation and history tracking of maintenance work.

[0014] In addition, the present invention aims to provide an elevator fault detection method and a server for executing the same, which enables the establishment of a predictive maintenance strategy by performing pattern analysis on recurring anomaly items based on accumulated inspection history and fault logs, and automatically proposing inspection priorities based on risk levels. means of solving the problem

[0015] An elevator fault detection server for achieving this purpose includes: a data collection unit that collects elevator operation data from a plurality of sensors; an abnormality detection unit that determines whether there is an abnormality by comparing the operation data with a reference value; an inspection item matching unit that automatically matches inspection items corresponding to the abnormality; an inspection guide generation unit that automatically generates an inspection guide for each inspection item; and a mobile inspection interface unit that provides the inspection guide to a user terminal and stores input information.

[0016] In one embodiment, the abnormal sign determination unit can calculate an abnormality determination index by comparing data including at least one of door opening / closing time, current value, and travel time with a reference value.

[0017] In one embodiment, the inspection guide generation unit can automatically generate a text document including a work sequence, a tool list, precautions, and related drawings corresponding to each inspection item.

[0018] In one embodiment, the elevator fault detection server further includes a multiple authentication unit that performs multiple authentication means sequentially or in parallel for a user's elevator access request, and the multiple authentication unit can perform user authentication using at least two of RFID, smartphone-based NFC authentication, and biometric recognition.

[0019] In one embodiment, the elevator fault detection server may further include an AI-based user behavior pattern analysis unit that determines whether there is an abnormal access based on data such as the user's authentication history, time of entry and exit, destination floor number, and access frequency.

[0020] In addition, an elevator fault detection method executed on an elevator fault detection server for achieving this purpose includes the steps of: collecting elevator operation data from a plurality of sensors; determining whether there is an abnormality by comparing the operation data with a reference value; automatically matching inspection items corresponding to the abnormal signs; and automatically generating an inspection guide for each inspection item, providing the inspection guide to a user terminal, and storing input information.

[0021] In one embodiment, the step of determining whether there is an abnormality by comparing the operation data with a reference value may include the step of calculating an abnormality determination index by comparing data including at least one of door opening / closing time, current value, and travel time with the reference value.

[0022] In one embodiment, the step of providing the inspection guide to a user terminal and storing input information may include the step of automatically generating a text document containing a work sequence, a tool list, precautions, and related drawings corresponding to each inspection item.

[0023] In one embodiment, the elevator fault detection method may further include the step of performing at least two of the plurality of authentication means, such as RFID, smartphone-based NFC authentication, and biometric recognition, sequentially or in parallel in response to a user's request to access the elevator.

[0024] In one embodiment, the elevator fault detection method may further include a step of determining whether there is an abnormal access based on data such as the user's authentication history, time of entry and exit, destination floor number, and access frequency. Effects of the invention

[0025] According to the present invention as described above, there is an advantage in that maintenance centered on proactive prevention rather than reactive response is possible by analyzing operation data collected from elevator sensors to automatically detect signs of failure or conditions exceeding standards.

[0026] In addition, according to the present invention, there is an advantage in that the quality of work performed by technicians can be standardized and safety ensured by automatically deriving relevant inspection items based on detected abnormal signals and automatically generating an inspection guide for each item that includes inspection procedures, necessary tools, precautions, etc.

[0027] In addition, according to the present invention, there is an advantage in that the digital transformation and history tracking of maintenance work can be realized by enabling a technician to check inspection items and guides in real time through a mobile terminal at the site and to electronically record and transmit inspection results to a server.

[0028] In addition, according to the present invention, there is an advantage in that it is possible to establish a predictive maintenance strategy by performing pattern analysis on recurring anomaly items based on accumulated inspection history and failure logs, and automatically suggesting inspection priorities based on risk levels. Brief explanation of the drawing

[0029] FIG. 1 is a network configuration diagram for explaining an elevator fault detection system according to one embodiment of the present invention. FIG. 2 is a block diagram illustrating the internal structure of an elevator fault detection server according to an embodiment of the present invention. FIG. 3 is a block diagram illustrating the internal structure of an elevator fault detection server according to another embodiment of the present invention. FIG. 4 is a flowchart illustrating an embodiment of an elevator fault detection method according to the present invention. Specific details for implementing the invention

[0030] The aforementioned objectives, signatures, and advantages are described in detail below with reference to the attached drawings, thereby enabling those skilled in the art to easily implement the technical concept of the present invention. In describing the present invention, detailed descriptions of known technologies related to the present invention are omitted if it is determined that such descriptions would unnecessarily obscure the essence of the invention. Hereinafter, preferred embodiments according to the present invention will be described in detail with reference to the attached drawings. In the drawings, the same reference numerals are used to indicate the same or similar components.

[0032] FIG. 1 is a network configuration diagram for explaining an elevator fault detection system according to one embodiment of the present invention.

[0033] Referring to FIG. 1, the elevator fault detection system includes an elevator fault detection server (100), a plurality of sensors (200_1 to 200_N) and a user terminal (300).

[0034] The elevator fault detection server (100) collects status data, such as vibration, current, temperature, and location, received from multiple sensors (200_1 to 200_N) in real time. Based on the collected data, it analyzes whether there are signs of a fault, and if an abnormal condition is detected, it transmits the judgment result to the inspection item matching unit. The elevator fault detection server (100) maps corresponding inspection items according to the type of fault and links to generate work instructions through the inspection guide automatic generation unit. In addition, the abnormal judgment result is automatically transmitted to the manager system, and alarms and warnings are sent if necessary. This elevator fault detection server (100) will be described in detail below.

[0035] A plurality of sensors (200_1 to 200_N) detect signs of abnormalities in the operating status and components of the elevator in real time and provide the corresponding data to the fault detection server (100) so that a fault can be determined early. The plurality of sensors (200_1 to 200_N) are classified according to their respective functions and installation locations and may include vibration sensors, current sensors, temperature sensors, door opening / closing sensors, acceleration and gyroscope sensors, inter-floor position sensors, etc.

[0036] Vibration sensors are attached to mechanical structures such as elevator motors, control panels, and guide rails to measure vibration signals generated during operation, and are designed to identify abnormal vibration if the frequency exceeds a certain threshold or the RMS value exceeds a reference value. Vibration data is collected periodically from a server or transmitted on an event-based basis.

[0037] A current sensor is installed on a drive motor or a main power supply line and measures real-time current values ​​generated during motor operation. By detecting overcurrent, sudden current fluctuations, or abnormal current patterns during operation, the current sensor enables early detection of motor overload, circuit damage, or braking abnormalities.

[0038] A temperature sensor is attached to a motor, control panel, inverter module, or power element to detect the temperature of the corresponding part, and when the temperature rises above a certain level, it can determine the possibility of an abnormality due to deterioration, overload, or equipment aging.

[0039] The door opening / closing sensor detects the open and closed status of the elevator door in real time and transmits information such as door opening / closing time, the number of opening / closing failures, and whether the elevator stopped midway to the server. This data can be used to diagnose issues such as door link wear, sensor contamination, or drive unit malfunctions.

[0040] The acceleration and gyroscope sensor (IMU) measures acceleration, deceleration, vibration, and tilt during the vertical movement of the elevator, thereby enabling the real-time detection of abnormal signs such as sudden stops, shaking, and unbalanced movement during operation.

[0041] The inter-floor position sensor identifies the elevator's current location in reference floor units and is utilized to determine position sensor malfunctions or control inconsistencies by verifying response speed upon call, whether the target floor has been reached, and errors.

[0042] The above plurality of sensors measure the status of the elevator from various perspectives and transmit the data to an elevator fault detection server (100) through a local controller or relay device. The transmission method may include wired communication (RS485, Ethernet, etc.) or wireless communication (Wi-Fi, BLE, LoRa, etc.), and the data is collected via periodic collection or event triggering.

[0043] The user terminal (300) is a portable terminal device that provides an inspection guide display, item-by-item result input, supplementary question response, and server linkage functions so that an inspection performer can perform an on-site inspection according to the instructions of the elevator fault detection system, and can be implemented in the form of a smartphone, tablet, or portable PDA.

[0044] The user terminal (300) receives inspection item information and work instruction data and displays them sequentially on a touch-based interface screen. Each inspection item is configured with visual materials such as a work sequence, target area, inspection method, precautions, and reference images or drawings, so that the worker can intuitively perform the inspection step by step.

[0045] The user selects and inputs a status such as "Normal," "Abnormal," or "Not Applicable" for the inspection results of each item, and can add additional information if necessary by taking a photo of the site, entering a text memo, or registering a voice memo. The above input information is structured through the internal data processing unit of the user terminal (300) and is automatically classified and stored by inspection item.

[0046] Additionally, if an "abnormal" state is entered in a specific item, the user terminal (300) displays conditional supplementary questions or additional inspection requests transmitted from the fault detection server (100) or the inspection item matching unit on the screen in real time. The user may perform additional input or verification work regarding this, and during this process, an automatic correction instruction or an administrator call function may be activated.

[0047] The user terminal (300) transmits the input inspection results and related data in real time to the fault detection server (100) or the central control system, and the elevator fault detection server (100) performs subsequent processing such as classifying the fault type, storing maintenance history, and transmitting abnormal alarms based on the data. At the same time, feedback on the inspection results, administrator comments, and additional instructions are received by the user terminal (300), thereby maintaining a real-time workflow.

[0048] According to this configuration, the user terminal (300) not only assists field technicians in performing inspections but also provides technical effects such as maintenance automation, improved failure response speed, and improved accuracy of history tracking through real-time collection, automatic storage, and central linkage of inspection data.

[0050] FIG. 2 is a block diagram illustrating the internal structure of an elevator fault detection server according to an embodiment of the present invention.

[0051] Referring to FIG. 2, the elevator fault detection server (100) includes a data collection unit (110), an abnormal sign determination unit (120), an inspection item matching unit (130), an inspection guide generation unit (140), a mobile inspection interface unit (150), and a control unit (160).

[0052] The data collection unit (110) collects data in real-time or periodically from a plurality of sensors (200_1 to 200_N) installed in the control panel or operation-related components of the elevator, and subsequently provides base information for determining abnormal signs, matching inspection items, and automatically generating guides.

[0053] The data collection unit (110) receives various physical / electrical data related to elevator operation, such as door opening / closing status, motor current, inter-floor travel time, and vibration frequency, and normalizes the received data into a structured form according to internal processing rules.

[0054] For example, data received from a door opening / closing sensor is structured by separating it into items such as the start time (Start_Timestamp), end time (End_Timestamp), duration of the operation (Duration), and whether it exceeds a threshold value (Over_Threshold_Flag), corresponding to the door opening or closing event.

[0055] In addition, the data received from the motor current sensor includes the operating current value (Current_Value), the result of comparison with the reference current threshold (Current_Threshold), flags for determining whether a current spike has occurred, and abnormalities; all information is organized in the form of attribute-value pairs.

[0056] The data collection unit (110) does not merely store the collected data, but also generates information that can determine whether there is an abnormality by comparing it with a predefined standard value and tolerance. For example, if the door opening / closing time exceeds a standard value (e.g., 2.5 seconds), the data is indicated with a 'DOOR_SLOW_CLOSE' flag to indicate the possibility of an abnormality.

[0057] The collected data is subsequently transmitted to the abnormal sign judgment unit (120) and used for fault prediction, and at the same time, it can be accumulated in a storage within the server and used as training data for repetitive fault analysis or predictive maintenance algorithms.

[0058] The data collection unit (110) may support a streaming-based structure or an event-triggered collection mode for processing real-time data flow, and the communication protocol may be configured in the form of RS-485, CAN, or IoT gateway-based REST API, MQTT, etc.

[0059] In addition, the data collection unit (110) classifies data for each sensor and manages the measurement values ​​of each sensor in an integrated manner along with time information, units, and measurement locations according to data normalization rules. This data is used as highly reliable input information for other components in determining conditions, automatically generating inspection items, generating inspection guide descriptions, etc.

[0060] Accordingly, the data collection unit (110) according to the present invention has the technical effect of being able to provide accurate and meaningful status information to a subsequent judgment module by collecting and normalizing sensor-based status information in real time and including a reference value comparison and flag setting function for abnormality determination.

[0061] The abnormal sign determination unit (120) quantitatively determines whether there is an abnormal state for a component of the elevator or the entire system based on structured state information received from the data collection unit (110). The abnormal sign determination unit (120) compares actual measurement data with a pre-set reference value or dynamic threshold value, and if there is a possibility of failure, generates a determination result in the form of a binary flag or risk score so that it is utilized for matching inspection items and generating a maintenance guide.

[0062] First, the abnormal sign judgment unit (120) loads a threshold value corresponding to a monitoring item (e.g., door opening / closing time, inter-floor movement speed, current value, etc.). At this time, the threshold value may be a fixed value defined in advance, or a dynamic threshold value that reflects the average, standard deviation, etc. of recent operation data. For example, the threshold value for the door opening / closing time may be set to 1.5 times the average opening / closing time, or the threshold value for the motor current may be set to ±10% of the design current value.

[0063] In one embodiment, the abnormal sign determination unit (120) performs a comparison operation with a reference value on the structured actual measurement data input from the data collection unit (110). The abnormal sign determination unit (120) can individually determine whether each item exceeds the limit, or perform a comprehensive abnormality determination through the combination condition of multiple items.

[0064]

[0066] For example, if the average time required for the door closing operation is 2.3 seconds, the reference value is 2.0 seconds, and the tolerance range is ±0.2 seconds, since |2.3-2.0|=0.3>0.2, the item is determined to be abnormal and P fault =1.

[0067] In another embodiment, the abnormal sign determination unit (120) quantitatively determines the possibility of a complex failure by logically combining abnormality flags (P1, P2, ..., Pn) that are individually determined for a plurality of status items.

[0068] To this end, the abnormal sign judgment unit (120) comprehensively determines the signs of failure based on whether there are abnormalities in a plurality of status items, and selectively applies an AND condition or an OR condition depending on the type of failure and the purpose of system operation.

[0069] When the abnormal sign judgment unit (120) combines the results of determining whether each state item is abnormal, it selects an operation method according to a combination condition rule (condition mapping rule) defined for each failure type. The combination condition applied at this time can be set to a fixed logical operation method (AND, OR) or a weight-based combination method.

[0070] More specifically, the abnormal sign determination unit (120) determines the calculation method for each cause of failure by referring to the failure type-condition mapping table. [Table 1] below shows the abnormal items and applied calculation conditions according to the failure type.

[0071]

[0073] In this way, the abnormal sign judgment unit (120) selects an AND condition and an OR condition according to the combination of input items. At this time, the AND condition is suitable for fault judgment that prioritizes accuracy (True Positive Rate) when it is necessary to verify complex signs caused by the same cause, and the OR condition is suitable for an initial warning system that prioritizes sensitivity and is applied when preemptive response is required even when a single abnormal item occurs.

[0074] Additionally, the abnormal sign judgment unit (120) can dynamically reset the condition expression by considering the equipment's operation history, failure rate, and equipment usage period. For example, for equipment where the same failure has occurred three or more times, the judgment condition can be strengthened from OR to AND, and conversely, for equipment within three months of new installation, the OR condition can be applied first to increase sensitivity.

[0075] Accordingly, the abnormal sign judgment unit (120) can quantitatively determine the possibility of a complex failure by logically combining abnormality flags (P1, P2, ..., Pn) that are individually determined for a plurality of status items.

[0076] For example, the abnormal sign determination unit (120) detects a motor current abnormality (P) when the actual value received from the motor current sensor exceeds the current reference value. current It is set to =1), and at the same time, if the frequency response received from the vibration sensor exceeds the reference, it is a vibration anomaly (P vibration =1) is set. At this time, the abnormal sign judgment unit (120) derives a comprehensive fault probability judgment result based on AND condition-based combination judgment such as [Equation 2].

[0077]

[0079] Accordingly, when an abnormal sign judgment unit (120) detects that a motor current abnormality (Pcurrent=1) and a vibration abnormality (Pvibration=1) occur simultaneously, the comprehensive judgment result is It is set to =1, which is judged as "possibility of motor overload or bearing wear". On the other hand, if even one is normal, the overall judgment result is considered to be 0.

[0080] In addition, the abnormal sign judgment unit (120) derives a comprehensive failure probability judgment result based on OR condition-based combination judgment such as [Equation 3].

[0082]

[0084] In another embodiment, the abnormal sign judgment unit (120) assigns different weights according to the importance of each item, calculates the weighted average-based overall risk R according to [Equation 4], and determines a failure state only when it is above a set threshold.

[0085]

[0087] For example, w1=0.6 for motor current abnormality and w2=0.4 for vibration abnormality are assigned, and if both items are judged to be in an abnormal state, R=1.0, and since the threshold value Rth =0.8 or higher, it is finally judged as Ptotal=1.

[0088] This multi-item combined judgment technique reduces the possibility of misjudgment caused by noise or errors in single indicators and enables more precise and reliable fault indication judgment based on combinations of items that have a close correlation with actual failure occurrence.

[0089] As a result, the abnormal sign judgment unit (120) implements a multi-condition combination judgment algorithm based on logical operation and weight combination method for complex failure situations that are difficult to judge based solely on whether a simple reference value is exceeded, thereby increasing the accuracy of predictive maintenance, providing a more reliable input signal to the inspection item matching unit, preventing unnecessary waste of maintenance resources, and providing technical effects that maximize system availability and maintenance efficiency.

[0090] The inspection item matching unit (130) automatically derives a corresponding maintenance task item according to each failure type by referring to a failure-inspection response mapping table or an ontology-based rule knowledge graph.

[0091] In this context, the fault-inspection response mapping table is a reference information table designed to automatically match maintenance inspection items (Tasks) to be performed based on predefined rules. The fault-inspection response mapping table defines a list of corresponding inspection items for each fault type, unit by fault code.

[0092] The above fault-inspection response mapping table is used to determine which inspection item is required and which part or sub-module needs to be inspected according to the cause of the related fault when an abnormal condition is detected in the inspection item matching unit (130).

[0093] Additionally, the inspection item matching unit (130) generates the following structured inspection items (Task Set) based on the extracted list of work items. Each inspection item includes metadata such as a unique identifier (Task_ID), a component to be inspected (Component), an objective (Objective), an inspection method (Method), necessary equipment / tools (Tool), a level of work risk, and an estimated duration.

[0094] In one embodiment, the inspection item matching unit (130) may filter or supplement inspection items according to the equipment status or environmental conditions. For example, if the equipment that has failed is an externally installed type, an item for 'waterproof connector status check' may be added, or if the equipment is installed in a high-temperature environment, an item for 'temperature sensor calibration verification' may be included. This process is configured as a conditional logic by considering the equipment's operation log, metadata, field environmental conditions, etc.

[0095] In another embodiment, the inspection item matching unit (130) calculates an inspection priority score for each inspection item based on failure severity, frequency of occurrence, safety impact, etc., and sets a ranking so that important items are reflected in the inspection guide first. The priority score can be calculated as [Equation 5].

[0097]

[0099] The above results are placed at the top of the list starting from the highest priority items, and can be grouped by task stage if necessary.

[0100] The inspection guide generation unit (140) automatically generates a work inspection guide based on one or more sets of inspection items input from the inspection item matching unit (130), which includes information such as the purpose of the work, the order of inspection, precautions during the work, tools to use, and visual materials corresponding to each item. The inspection guide generation unit (140) simultaneously includes a fixed template-based configuration and a document search-based description supplementation function, and the generated inspection guide can be transmitted to a user terminal (200) in real time.

[0101] First, the inspection guide generation unit (140) automatically selects a template that matches the item from a pre-stored work inspection guide template library based on the meta-information (Task_ID, inspection target area, maintenance purpose, risk level, etc.) included in each input inspection item. The template includes fields such as work name, procedure sequence, required equipment, and time required, and the selected template is dynamically filled for each item.

[0102] Then, for items that are difficult to create using only a template, the inspection guide generation unit (140) searches for unstructured documents such as past work reports, equipment manuals, work logs, and maintenance history by linking with a RAG-based document generation module, extracts sentences or precautions similar to the inspection item, and automatically generates descriptive sentences. The generated sentences are inserted into the description, precaution, or warning section of the inspection item.

[0103] Additionally, the inspection guide generation unit (140) links with an ontology system or a vision system to automatically search for and select visual materials (e.g., drawings, 3D models, on-site images, etc.) associated with the inspection item, and inserts the corresponding images or visual information into each work step. At this time, the search query is generated based on key values ​​such as part ID, work location, and process name, and the visual materials are placed next to the description of each step.

[0104] Additionally, the inspection guide generation unit (140) automatically inserts warning messages or worker reference sentences based on past work results, problem history, and worker feedback regarding the same item, if there are matters requiring caution regarding a specific work step. For example, if there is a history of "rework occurred due to missing wire binding during past inspection," a warning message such as "check the wire fixing status at least twice" is automatically inserted.

[0105] In addition, the inspection guide generation unit (140) converts the automatically generated inspection guide into a set format (e.g., HTML, PDF, JSON, etc.) and transmits it to a user terminal (mobile app, web UI, etc.), an administrator portal, and a history storage within a server, and the document is used for pre-viewing before work, step-by-step reference during work, and recording results after work.

[0106] The configuration and linkage structure of the inspection guide generation unit (140) above provides a technical effect that improves work consistency, safety, and efficiency in that it can automatically generate and provide a high-quality work inspection guide suitable for the maintenance site using only the fault judgment result and inspection items, without relying on the worker's experience or document writing ability.

[0107] The mobile inspection interface unit (150) provides a work inspection guide generated by the inspection guide generation unit (140) to a user terminal and provides a mobile-based inspection execution interface that allows the user to perform an inspection and input results by referring to the inspection guide step-by-step at the site. In addition, the mobile inspection interface unit (150) is configured to transmit the input inspection results to a server in real time and to enable automatic saving and subsequent processing of the data by linking with a maintenance history system or an administrator module.

[0108] First, the mobile inspection interface unit (150) receives a work inspection guide transmitted from the inspection guide generation unit (140) in HTML, JSON, or PDF format and displays it on a user terminal (e.g., a mobile app or a web-based UI). Each step of the inspection guide is arranged sequentially on the screen according to the order of work, and each item includes information such as the name of the work, method of execution, precautions, necessary equipment, and estimated time required, and if necessary, a drawing image, a 3D model, or an actual captured image is displayed together.

[0109] After performing the tasks of each step, the user can directly input the result of the task as a status such as "Normal," "Abnormal," or "Not Performed," and if necessary, add detailed information by attaching on-site photos, entering text memos, or voice recordings.

[0110] In particular, when a user indicates a specific item as "abnormal," the mobile inspection interface unit (150) automatically presents conditional supplementary questions or additional inspection items related to the abnormal state. For example, if "damage to wire insulation" is entered, supplementary input fields such as "photographing the location of damage," "whether immediate action is taken," and "whether replacement is necessary" are automatically activated. This allows for the minimization of inspection omissions or judgment errors.

[0111] Subsequently, the inspection results entered by the user are automatically synchronized and stored in the maintenance history management system. The stored items include the inspection date and time, equipment ID, operator information, inspection results for each stage, attached files, and inspection completion status. This stored information can then be integrated with management systems, PMS, ERP, etc., to be utilized for statistical analysis of recurring failures, work quality evaluation, and the establishment of predictive maintenance policies.

[0112] In addition, the mobile inspection interface unit (150) of the present invention can transmit a notification in real time to an administrator terminal or an administrator system when a plurality of abnormal conditions occur consecutively during the inspection process or when a high-risk abnormality is detected. At this time, the notification is delivered through various means such as mobile push messages, SMS, and email, and allows for immediate action to be taken regarding whether an emergency inspection or maintenance is necessary.

[0113] According to one embodiment, when a manager or senior technician reviews the inspection results entered from the elevator fault detection server (100) and creates a feedback message, the mobile inspection interface unit transmits the feedback to the user terminal in real time and, if additional action is required, can create a follow-up work item or re-instruct.

[0114] The control unit (160) first monitors in real time whether a plurality of sensor data (e.g., vibration, current, temperature, location, door status, etc.) transmitted from the data collection unit (110) have been received and whether there is an abnormality, and verifies the integrity and consistency of the received data. The collected data is transmitted to the abnormal sign judgment unit (120), and the control unit (160) manages the execution timing of the judgment unit and the return of the analysis result.

[0115] When a specific abnormal condition is detected by the abnormal sign judgment unit (120), the control unit (160) transmits the judgment result to the inspection item matching unit (130) and instructs it to automatically map inspection items corresponding to the fault sign. The mapped inspection items are transmitted to the inspection guide generation unit (140) through the control unit (160), and the generated inspection guide is then provided to the user terminal (300) through the mobile inspection interface unit (150).

[0116] Additionally, the control unit (160) monitors the execution status, response delay, and error status of each component in real time, performs exception processing when an abnormal stop or abnormal response of a specific module is detected, and is linked with an administrator warning system. For example, if sensor data is not received for a certain period of time or if the judgment result repeatedly corresponds to a warning state, the control unit (160) sends a notification to the administrator terminal.

[0117] Furthermore, the control unit (160) can also perform a predictive maintenance-based policy application function that dynamically adjusts the priority of inspection items or automatically adjusts the inspection cycle by referring to maintenance history, recurring failure items, processing results, etc. accumulated in the elevator failure detection server (100).

[0118] Accordingly, the control unit (160) organically connects and controls the flow of all steps from data collection, anomaly determination, inspection item mapping, inspection guide generation, and user interface provision, thereby providing a technical effect that realizes intelligent operation of the elevator fault detection system, error minimization, and response time reduction.

[0120] FIG. 3 is a block diagram illustrating the internal structure of an elevator fault detection server according to another embodiment of the present invention.

[0121] Referring to FIG. 3, the elevator fault detection server (100) includes a multi-authentication unit (170) and an AI-based user behavior pattern analysis unit (180).

[0122] The multi-authentication unit (170) precisely verifies the user's authority by combining heterogeneous authentication means, such as RFID, smartphone authentication, and biometric recognition, for elevator access requests, and supports sequential execution or parallel verification among the authentication means. The multi-authentication unit (170) evaluates the success of each authentication means and determines whether to finally allow access according to the authentication combination policy.

[0123] First, when a user accesses an elevator call terminal or a floor selection panel, the multi-authentication unit (170) performs user authentication using an RFID card or an NFC-based digital key as a primary authentication. When the RFID card or the NFC tag of a smartphone is recognized by the reader, the multi-authentication unit compares the received user identification information with the registered user DB to determine whether the primary authentication was successful. If the primary authentication fails, subsequent procedures are not executed and access is blocked.

[0124] If primary authentication is successful, the multi-factor authentication unit performs biometric recognition as secondary authentication according to a predefined security policy. Biometric recognition is applied based on system settings using various methods such as fingerprint, facial, iris, and vein recognition, and receives biometric information in real time via sensors installed on user terminals or panels. The entered biometric information is compared with a pre-registered biometric profile, and if the similarity analysis result exceeds a threshold value, the biometric authentication is passed. This process is selectively activated only when additional authentication is required, such as in high-security layers or management areas.

[0125] As described above, the multi-authentication unit integrates the first authentication result (P1), the second biometric authentication result (P2), and the optional third AI analysis result (P3) to determine the final access status.

[0126] When final access is complete, the elevator control unit allows access to the requested floor, and all authentication logs are stored in the elevator fault detection server (100). The logs include user ID, authentication method, authentication time, whether authentication was successful or failed, whether anomalies were detected, etc. When authentication failure or abnormal access is determined, the information is transmitted to the administrator system in real time, and an emergency notification may be provided via push notification, email, or SMS.

[0127] Accordingly, the multi-authentication unit (170) of the present invention combines multiple authentication means, such as RFID / NFC, biometric recognition, and AI-based anomaly detection, and adaptively strengthens or adjusts the authentication means according to the situation, thereby overcoming the limitations of static authentication methods and providing a technical effect of simultaneously improving security and real-time responsiveness.

[0128] The AI-based user behavior pattern analysis unit (180) comprehensively analyzes data such as the user's past entry history, time zone, floor selection history, time of stay, and frequency of access, and determines whether the current request matches a normal access pattern. If the analysis result is classified as an abnormal access, the multi-authentication unit (170) may require administrator approval or a warning procedure, and automatically block access if necessary.

[0129] The AI-based user behavior pattern analysis unit (180) normalizes the elements to form a user-specific behavior characteristic vector and calculates a behavior similarity score by comparing it with a previously trained normal user behavior model. The similarity score can be calculated through a prediction function or an AI model, and if it is above a specified threshold, it is determined to be a normal approach.

[0130] If the AI-based user behavior pattern analysis unit (180) determines the access request as an abnormal access when the similarity score is below a threshold, it performs authentication reinforcement requests, access delay or blocking, alert transmission, log storage, etc.

[0131] The AI-based user behavior pattern analysis unit (180) can apply different standard models depending on the user grade, and for new users, automatically generates standard patterns through data collected over a certain period and determines whether there is an anomaly based on this. It may also include a function to collect data from multiple users and perform group-based anomaly detection.

[0132] The above AI-based user behavior pattern analysis unit (180) can perform more precise contextual recognition by linking with field sensors, access log systems, CCTV analysis servers, etc., and can continuously improve the accuracy of anomaly judgment.

[0133] Accordingly, the AI-based user behavior pattern analysis unit (180) of the present invention learns the normal access pattern of each individual user and preemptively blocks abnormal access through a consistency evaluation of real-time access requests, thereby providing a technical effect that can effectively prevent internal threats and bypass access that are difficult to detect with fixed security methods alone.

[0135] FIG. 4 is a flowchart illustrating an embodiment of an elevator fault detection method according to the present invention.

[0136] The elevator fault detection server (100) collects elevator operation data from a plurality of sensors (step S410). At this time, the plurality of sensors (200_1 to 200_N) detect signs of abnormalities in the elevator's operation status and components in real time and provide the corresponding data to the fault detection server (100) so that the fault can be determined early. The plurality of sensors (200_1 to 200_N) are classified according to their respective functions and installation locations and may include vibration sensors, current sensors, temperature sensors, door opening / closing sensors, acceleration and gyroscope sensors, inter-floor position sensors, etc.

[0137] The elevator fault detection server (100) determines whether there is an abnormality by comparing the operation data with a reference value (step S420).

[0138] In one embodiment of step S420, the elevator fault detection server (100) compares actual data with a preset reference value or dynamic threshold value, and if a fault possibility exists, generates a judgment result in the form of a binary flag or risk score so that it is utilized for matching inspection items and generating a maintenance guide.

[0139] First, the elevator fault detection server (100) loads a threshold value corresponding to a monitoring item (e.g., door opening / closing time, inter-floor movement speed, current value, etc.). At this time, the threshold value may be a fixed value defined in advance, or a dynamic threshold value that reflects the average, standard deviation, etc. of recent operation data. For example, the threshold value for the door opening / closing time may be set to 1.5 times the average opening / closing time, or the threshold value for the motor current may be set to ±10% of the design current value.

[0140] The elevator fault detection server (100) automatically matches inspection items corresponding to abnormal signs (step S430).

[0141] In one embodiment for step S430, the elevator failure detection server (100) refers to a Failure-Task Mapping Table or an ontology-based rule knowledge graph to automatically derive a corresponding maintenance task item for each type of failure.

[0142] At this time, the fault-inspection response mapping table is a reference information table for automatically matching maintenance inspection items (Tasks) to be performed based on predefined rules. The fault-inspection response mapping table defines a list of corresponding inspection items for each type of fault in units of fault codes. The above fault-inspection response mapping table is used to determine which inspection item is required and which part or sub-module needs to be inspected according to the relevant cause of the fault when an abnormal condition is detected in the inspection item matching unit (130).

[0143] The elevator fault detection server (100) automatically generates an inspection guide for each inspection item, provides the inspection guide to a user terminal, and stores input information (step S440).

[0144] In one embodiment for step S440, the elevator fault detection server (100) automatically generates a work inspection guide based on one or more sets of inspection items, which includes information such as the purpose of the work, the inspection sequence, precautions during the work, tools to use, and visual materials corresponding to each item.

[0146] Although the present invention has been described by the embodiments and drawings described above, the present invention is not limited to the above embodiments, and various modifications and variations are possible from this description by those skilled in the art to which the present invention pertains. Accordingly, the concept of the present invention should be understood only by the claims set forth below, and all equivalent or analogous variations thereof shall be considered to fall within the scope of the concept of the present invention. Explanation of the symbols

[0147] 110: Data Collection Unit 120 : Abnormal Sign Judgment Unit 130: Inspection Item Matching Section 140 : Inspection Guide Generation Section 150 : Mobile Inspection Interface Section

Claims

Claim 1 A data collection unit that collects elevator operation data from multiple sensors; an abnormality detection unit that determines whether there is an abnormality by comparing the operation data with a reference value; an inspection item matching unit that automatically matches inspection items corresponding to the abnormality; and an inspection guide generation unit that automatically generates an inspection guide for each inspection item. An elevator fault detection server comprising a mobile inspection interface unit that provides the inspection guide to a user terminal and stores input information, wherein the abnormal sign determination unit calculates an abnormality flag individually for each of a plurality of status items collected from the plurality of sensors by comparing them with a reference value, determines the overall probability of failure by logically combining the individually calculated abnormality flags by selecting either an AND condition or an OR condition according to a combination condition rule defined for each type of failure, and dynamically resets the AND condition or OR condition of the combination condition rule by considering at least one of the elevator's operation history, failure rate, and equipment usage period, wherein the inspection guide generation unit automatically inserts a warning phrase or a worker reference sentence into the inspection guide of the corresponding inspection item if there are matters requiring attention regarding a specific work step based on past work results, problem occurrence history, and worker feedback regarding the same item, and wherein the mobile inspection interface unit automatically activates and outputs conditional supplementary questions or additional inspection requests related to the input abnormal state on the screen of the user terminal when the user inputs the result of a specific inspection item as an "abnormal" state through the user terminal. Claim 2 An elevator fault detection server according to claim 1, wherein the abnormal sign determination unit calculates an abnormality determination index by comparing data including at least one of door opening / closing time, current value, and travel time with a reference value. Claim 3 An elevator fault detection server according to claim 1, wherein the inspection guide generation unit automatically generates a text document including a work sequence, tool list, precautions, and related drawings corresponding to each inspection item. Claim 4 An elevator fault detection server according to claim 1, further comprising a multiple authentication unit that performs multiple authentication means sequentially or in parallel for a user's elevator access request, wherein the multiple authentication unit performs user authentication using at least two of RFID, smartphone-based NFC authentication, and biometric recognition. Claim 5 An elevator fault detection server according to claim 1, further comprising an AI-based user behavior pattern analysis unit that determines whether there is an abnormal access based on at least one of the user's authentication history, time of entry and exit, destination floor number, and access frequency. Claim 6 A method for detecting elevator faults executed on an elevator fault detection server comprises: a step of collecting elevator operation data from a plurality of sensors; a step of determining whether there is an abnormality by comparing the operation data with a reference value; and a step of automatically matching inspection items corresponding to the abnormality. The method includes the step of automatically generating an inspection guide for each inspection item, providing the inspection guide to a user terminal, and storing input information. The step of determining whether there is an abnormality includes calculating an abnormality flag individually for each of the multiple status items collected from the multiple sensors by comparing them with a reference value, determining the overall probability of failure by logically combining the individually calculated abnormality flags by selecting either an AND condition or an OR condition according to a combination condition rule defined for each failure type, and dynamically resetting the AND condition or OR condition of the combination condition rule by considering at least one of the elevator's operation history, failure rate, and equipment usage period. The step of automatically generating the inspection guide includes automatically inserting a warning message or a worker reference sentence into the inspection guide of the relevant inspection item if there are matters requiring attention regarding a specific work step, based on past work results, problem history, and worker feedback regarding the same item. The step of storing input information includes, when a user inputs the result of a specific inspection item as an "abnormal" state through the user terminal, automatically activating and outputting a conditional supplementary question or additional inspection request related to the input abnormal state on the screen of the user terminal. Elevator malfunction detection method. Claim 7 A method for detecting elevator faults according to claim 6, wherein the step of determining whether there is an abnormality by comparing the above-mentioned operation data with a reference value includes the step of calculating an abnormality determination index by comparing data including at least one of door opening / closing time, current value, and travel time with a reference value. Claim 8 An elevator fault detection method according to claim 6, wherein the step of providing the inspection guide to a user terminal and storing input information includes the step of automatically generating a text document containing a work sequence, tool list, precautions, and related drawings corresponding to each inspection item. Claim 9 An elevator fault detection method according to claim 6, further comprising the step of performing at least two of the plurality of authentication means, such as RFID, smartphone-based NFC authentication, and biometric recognition, sequentially or in parallel in response to a user's elevator access request. Claim 10 An elevator fault detection method according to claim 6, further comprising a step of determining whether there is an abnormal access based on at least one of the user's authentication history, time of entry and exit, destination floor number, and access frequency.

Images