A counter mode adaptive detection switching method and system of a self-service check-in device

Abstract

The application discloses a self-service check-in equipment counter mode adaptive detection switching method and system, relates to the technical field of check-in equipment, and comprises the following steps: defining multiple counter service modes in the central processing unit of the self-service check-in equipment, wherein the counter service modes have mutually exclusive access conditions and exit conditions; collecting multi-dimensional state signals of an air information host, a programmable logic controller, an application execution agent and a passenger interactive terminal; performing hierarchical priority mode trigger judgment on the multi-dimensional state signals, sequentially responding to safety events, air information host instructions, equipment and business states and passenger interactive behaviors, and determining a target counter service mode; after safety verification and process cleaning steps before mode switching, performing mode switching of the target counter service mode; and after completing the mode switching, synchronously updating man-machine interactive interfaces and external system state information. The application improves the autonomy, safety and service continuity of the self-service check-in system.

Description

Technical Field

[0001] This application relates to the field of shipping equipment technology, and in particular to a method and system for adaptive detection and switching of counter mode for self-service shipping equipment. Background Technology

[0002] In the check-in areas of major hub airports, multiple self-service baggage drop-off kiosks are typically deployed to handle different time periods, passenger densities, and special passenger service needs. To optimize resource allocation and service efficiency, these kiosks need to dynamically switch between various counter service modes. For example, during peak flight periods, a "fully self-service mode" is activated to increase throughput; during off-peak periods or in areas with a high concentration of elderly or disabled passengers, a "semi-self-service assistance mode" is switched to, with staff providing remote or on-site assistance; when the kiosks experience hardware failure or are undergoing maintenance, they should enter a "maintenance shutdown mode" to prevent passenger misoperation; and in the event of a sudden security incident or instructions from the flight information system, a "safety pause mode" must also be supported.

[0003] However, the current mode switching mechanism of mainstream self-service baggage check-in equipment relies heavily on manual intervention—maintenance personnel need to manually select and issue the target mode through the back-end management system. This method has inherent drawbacks such as large response delays, cumbersome operation procedures, and susceptibility to human error.

[0004] More critically, existing self-service baggage check-in systems lack the ability to effectively fuse multi-source heterogeneous status signals: on the one hand, they cannot detect in real time global shutdown commands issued by the DCS (Distributed Control System) due to flight cancellations, check-in closures, etc.; on the other hand, they cannot respond promptly to critical hardware anomalies detected by the local programmable logic controller (PLC) (such as conveyor jamming, roller shutter malfunction, weighing module failure, etc.). In such scenarios, the equipment continues to operate in its original service mode, causing passengers to queue for extended periods in front of invalid terminals, which not only reduces the service experience but may also lead to disorder and operational risks.

[0005] Therefore, there is an urgent need for an intelligent control method that can achieve adaptive recognition, security verification, and seamless switching of counter service modes in order to improve the autonomy, security, and service continuity of self-service baggage check-in systems. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this application provides an intelligent control method capable of adaptive recognition, security verification, and seamless switching of counter service modes, thereby enhancing the autonomy, security, and service continuity of self-service baggage check-in systems. This application provides a method and system for adaptive detection and switching of counter modes for self-service baggage check-in equipment.

[0007] Firstly, the objective of this invention is achieved through the following technical solution: An adaptive detection and switching method for counter modes of a self-service baggage check-in device, comprising: Multiple counter service modes are predefined in the central processing unit of the self-service check-in equipment. Each counter service mode has mutually exclusive entry and exit conditions. Collect multi-dimensional status signals from the travel information host, programmable logic controller, application execution agent, and passenger interaction terminal, and attach timestamps; Based on preset priority rules, the multi-dimensional status signals are subjected to hierarchical priority mode triggering judgment, and security events, travel information host instructions, equipment and service status and passenger interaction behavior are responded to in sequence to determine the target counter service mode. The mode switch for the target counter service mode is performed only after the security verification and process cleanup steps before mode switching are completed. After the mode switch is completed, the human-computer interaction interface and external system status information are updated synchronously.

[0008] By adopting the above technical solution, all multi-dimensional status signals are timestamped to ensure temporal consistency, achieving comprehensive perception of the equipment operating environment of self-service baggage check-in equipment. Based on preset priority rules, the above signals are hierarchically judged, responding sequentially to security events, travel information commands, equipment and business status, and passenger interaction behavior, thereby dynamically and accurately determining the target counter service mode. A security verification and process cleanup mechanism is introduced before switching, effectively avoiding security risks caused by baggage handling interruptions or status conflicts during mode switching. After switching, the human-machine interface and external system status information are updated synchronously to ensure consistency between the passenger interaction terminal, on-site devices, and the back-end monitoring system. Compared with existing static configuration methods that rely on manual intervention, this invention significantly improves the real-time performance, accuracy, and security of switching between various counter service modes: on the one hand, it can automatically trigger the corresponding security mode when the travel information host issues a deactivation command or when local hardware malfunctions, preventing passengers from queuing ineffectively; on the other hand, through the multi-source signal fusion and priority decision mechanism of structured multi-dimensional status signals, the equipment can autonomously adapt to the optimal service strategy according to the actual operating situation, thereby enhancing system robustness, optimizing passenger service experience, and reducing the burden of airport operation and maintenance management.

[0009] In a preferred embodiment of this application: the counter service mode includes a fully self-service mode, a semi-self-service assistance mode, a manually operated mode, a maintenance shutdown mode, and an emergency pause mode; the step of triggering the hierarchical priority mode includes: If an emergency stop signal or an external safety linkage signal is detected reported by the programmable logic controller, the emergency stop mode is immediately triggered. If a counter service or mode switching instruction is received from the TravelSky host, the current mode will be overwritten and the service mode specified in the instruction will be switched. If the aforementioned high-priority event is not triggered, and the programmable logic controller or the application execution agent reports critical subsystem fault information, then the system will be downgraded to semi-self-service assistance mode or maintenance shutdown mode. In the absence of the aforementioned events, the current mode is adjusted to a mode with higher service granularity based on the operation timeout, help requests, or low interaction frequency reported by the passenger interactive terminal.

[0010] By adopting the above technical solutions, the counter service modes include fully self-service, semi-self-service assistance, manual operation, maintenance shutdown, and emergency suspension. Each mode is configured with different priority events and differentiated trigger logic (such as emergency stop signals immediately triggering emergency suspension, direct overriding of flight information commands, fault degradation, and user behavior-guided upgrades), enabling the system to accurately match the needs of actual operational scenarios. In the event of a sudden safety incident, the equipment can enter a safe state within milliseconds. It can also automatically select the optimal service granularity in different scenarios such as airline scheduling, equipment failure, or passenger assistance, avoiding resource waste caused by excessive reliance on manual intervention and preventing service failures caused by complete self-service.

[0011] In a preferred embodiment of this application, the security verification and process cleanup steps include: Send hardware enable or disable commands matching the target counter service mode to the programmable logic controller to control the operating status of the conveyor, roller shutter door, and weighing module; Send a mode change notification to the application execution agent to adjust subsequent business process logic; If any of the admission conditions for the target counter service mode are not met, the current mode will remain unchanged, and an exception log will be generated.

[0012] By adopting the above technical solution, hardware enable / disable commands are sent to the programmable logic controller before mode switching, and the application execution agent is notified of the process logic change, thus achieving atomic mode switching with hardware and software collaboration. Security verification and process cleanup steps ensure that the states of critical actuators such as conveyors, roller shutters, and luggage weighing sensors are strictly aligned with the target mode, preventing operational accidents caused by hardware insecurity (such as the conveyor unexpectedly starting in manual mode). Simultaneously, when the access conditions are not met, switching is rejected and an exception log is recorded, avoiding chaos caused by illegal state transitions.

[0013] In a preferred embodiment of this application: the passenger interactive terminal includes a camera, an ID card reader, a touchscreen operation log module, and an infrared proximity sensor; the application execution agent performs the following identification tasks based on the video stream and operation log uploaded by the passenger interactive terminal: Identify the current passenger's identity information, current operational stage, operational proficiency, and abnormal behavior; The abnormal behaviors include prolonged inactivity, repeated accidental touches, failed baggage drop-off, and multiple failed identity verifications. The operation proficiency is determined by comparing the current operation sequence with the historical normal user template. If the reconstruction error exceeds a preset error threshold, the user is marked as a low proficiency user. The recognition result serves as a status signal for passenger interaction behavior, which is then input into the hierarchical priority mode to trigger the judgment process.

[0014] By adopting the above technical solution, a camera, ID card reader, touch screen log and infrared sensor are integrated into the passenger interaction terminal. The application execution agent performs comprehensive identification of identity information, operation stage, operation proficiency and abnormal behavior on the video stream and operation log. The original passive user input is transformed into active user intent and ability perception. This allows for the early prediction of passenger operation difficulties (such as multiple verification failures or long periods of inactivity), which significantly reduces the interruption rate of self-service processes and passenger frustration.

[0015] In a preferred embodiment of this application, the specific process for determining the operational proficiency is as follows: The application execution agent obtains the passenger operation logs of successfully completed baggage check-in in the past. Each step in the passenger operation log includes the operation type, time consumption, and touch coordinates. A baseline model of normal behavior is trained using an LSTM autoencoder, and latent vectors are obtained based on the baseline model of normal behavior. Obtain the current passenger's operation sequence and encode the operation sequence into a current latent vector of the current passenger's operation sequence; calculate the reconstruction error based on the latent vector and the current latent vector; The reconstruction error is compared with a preset error threshold to identify low-skilled users, and a status signal suggesting enhanced guidance is generated and reported to the central processing unit.

[0016] By adopting the above technical solution, a baseline model of normal passenger operation behavior is constructed using an LSTM autoencoder, and the L2 reconstruction error between latent vectors is used as the basis for proficiency judgment, achieving a quantitative, objective, and data-driven evaluation of user operation patterns. By capturing more subtle behavioral anomalies (such as non-linear operation paths and dense invalid clicks), the accuracy and generalization ability of identifying less skilled users are improved; this helps to effectively bridge the digital divide and increase the success rate of baggage check-in for elderly or first-time users without increasing labor costs.

[0017] In a preferred embodiment of this application: the programmable logic controller (PLC) is connected to a baggage weighing sensor, a conveyor driver, a roller shutter door actuator, and an X-ray security imaging module; the PLC monitors equipment malfunctions and security risk events based on sensor data. The equipment malfunctions include luggage jamming, barcode printing failure, and payment channel interruption. The security risk events include unattended luggage left behind, suspicious item image features, and unauthorized personnel approaching; When a high-risk event is detected, the programmable logic controller generates a security event status signal and triggers an emergency pause mode or a manual monitoring mode with the highest priority. The method for detecting unattended luggage left behind is as follows: The luggage weighing sensor confirms that luggage has been placed; The infrared proximity sensor and video analysis are used to determine whether a passenger has left the area 1.5 meters in front of the device and to obtain the time of the passenger's departure. If the weight of the luggage remains and the passenger has been gone for more than 90 seconds, it is considered abandoned luggage. The suspicious item image feature detection uses a convolutional neural network model to classify the X-ray images acquired by the X-ray security inspection imaging module. Suspicious items include lithium batteries, liquid containers, and metal blocks. If the sum of the confidence scores of any suspicious item is greater than 0.8, it is marked as high-risk baggage.

[0018] By adopting the above technical solution, a programmable logic controller (PLC) is used to integrate multimodal sensor data such as baggage weighing, infrared proximity, and X-ray imaging to achieve real-time, closed-loop monitoring of equipment malfunctions (such as jamming or printing failures) and security risks (such as lost baggage, suspicious items, or unauthorized approach). In particular, the use of a triple condition of "weight + personnel location + time" to determine lost baggage significantly reduces the false alarm and false alarm rates. Once a high-risk event is detected, the emergency or manual mode is triggered with the highest priority, effectively preventing security loopholes and public safety incidents.

[0019] In a preferred embodiment of this application: the application execution agent constructs a passenger behavior prediction model based on historical passenger interaction data and device operation logs; Using real-time multidimensional state signals as input, the probability of failure for the current passenger to complete the next operation step is predicted; the passenger behavior prediction model adopts a temporal convolutional network, with the input tensor dimension being the multidimensional state signals within the past 30 seconds, and the output being the probability of operation failure within the next 10 seconds; When the failure probability exceeds a preset failure threshold, the application execution agent generates a high failure risk status signal, causing the central processing unit to adjust the current mode to a semi-self-service assistance mode with higher service granularity when there are no high-priority events, and optimize the layout of the human-computer interaction interface.

[0020] By adopting the above technical solutions, a passenger behavior prediction model based on Temporal Convolutional Network (TCN) is constructed. This model dynamically predicts the probability of operational failure within the next 10 seconds using multi-dimensional state signals from the past 30 seconds, and proactively adjusts service modes and interface layout during high-risk situations, achieving a paradigm shift from "passive response" to "proactive intervention." The self-service check-in equipment possesses proactive service capabilities: optimizing the interaction before passenger failure (e.g., highlighting key buttons, disabling back button), reducing operational confusion and process rollback; simultaneously, it only triggers mode fine-tuning when there are no higher-priority events, avoiding excessive intervention that interferes with normal users. Overall, while maintaining high throughput, it significantly improves the first-time operation success rate and passenger satisfaction.

[0021] Secondly, the objective of this invention is achieved through the following technical solution: A self-service check-in device counter mode adaptive detection and switching system is applied to the self-service check-in device counter mode adaptive detection and switching method described above. The system includes: The central processing unit is used to predefine multiple mutually exclusive counter service modes, each of which is configured with corresponding admission and exit conditions. The multi-source status signal acquisition module is communicatively connected to the travel information host, programmable logic controller, application execution agent and passenger interaction terminal, and is used to acquire multi-dimensional status signals from each module in real time and add a unified timestamp to the multi-dimensional status signals. The hierarchical priority decision engine, built into the central processing unit, is used to perform hierarchical priority mode triggering judgment on the multi-dimensional status signals based on preset priority rules, and respond sequentially to security events, travel information host commands, equipment and service status and passenger interaction behavior, and output the target counter service mode. The security verification and process cleanup unit is used to verify whether the access conditions of the target counter service mode are met before the execution mode switch, and to complete the interruption cleanup of the current business process. The mode execution and synchronization module is used to activate the target counter service mode after passing the security verification and process cleanup steps, and to synchronously update the content of the human-computer interaction interface and broadcast the current mode status information to external systems.

[0022] Thirdly, the objective of this invention is achieved through the following technical solution: A computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the above-described self-service check-in device counter mode adaptive detection and switching method.

[0023] Fourthly, the objective of this invention is achieved through the following technical solution: A computer program product includes a computer program / instructions that, when executed by a processor, implement the steps of a counter mode adaptive detection and switching method for a self-service check-in device as described above.

[0024] In summary, this application includes at least one of the following beneficial technical effects: 1. Mutually exclusive counter service modes are predefined in the central processing unit, and time-stamped collection and hierarchical priority judgment are performed based on multi-dimensional status signals from the travel information host, programmable logic controller, application execution agent and passenger interaction terminal, realizing structured, orderly and conditional control of mode switching; 2. When facing concurrent events from multiple sources, always respond in the order of priority: "security > command > device and business status > user behavior" to avoid mode conflicts or accidental switching. At the same time, through security verification and process cleanup before switching, as well as interface and status synchronization after switching, we can effectively prevent business interruption, data inconsistency or hardware malfunction. Attached Figure Description

[0025] Figure 1 This is a flowchart of an adaptive detection and switching method for the counter mode of a self-service check-in device according to an embodiment of this application; Figure 2 This is a flowchart of step S2 in a self-service check-in device counter mode adaptive detection and switching method according to an embodiment of this application. Detailed Implementation

[0026] The present application will be further described in detail below with reference to the accompanying drawings.

[0027] In one embodiment, such as Figure 1 As shown, this application discloses an adaptive detection and switching method for the counter mode of a self-service baggage check-in device, which specifically includes the following steps: S1: Multiple counter service modes are predefined in the central processing unit of the self-service check-in equipment. Each counter service mode has mutually exclusive entry and exit conditions.

[0028] In this embodiment, the counter service mode refers to a complete set of working state configurations adopted by the self-service check-in equipment under different operating scenarios, including human-computer interaction strategies, hardware enablement status, business process paths, and external communication behaviors. Mutually exclusive entry and exit conditions mean that the device can only be in one valid mode at any given time, and before switching from the current mode to the target mode, all entry conditions of the target mode (such as hardware readiness and no ongoing transactions) must be met, while all exit conditions of the current mode (such as process completion and resource release) must also be met. In this embodiment, each mode and its condition set are predefined in the configuration file of the central processing unit (e.g., an embedded industrial control host or an ARM server) in the form of structured data (such as JSON or XML).

[0029] Specifically, the central processing unit loads a mode configuration table when the equipment starts up. This table contains multiple mode entries, each defining a mode identifier (e.g., "MODE_AUTO"), a list of admission conditions (e.g., "Weighing module online = true AND conveyor idle = true"), a list of exit conditions (e.g., "No passenger session currently AND buffer queue empty"), and an associated set of hardware control instructions. For example, the admission condition for a certain mode might be set to "the safety door closing signal reported by the PLC is high and the flight information host has not issued a deactivation command," while its exit condition is "passengers have completed baggage drop-off and printed boarding passes." The admission and exit conditions are stored as Boolean expressions and evaluated in real-time by the central processing unit's rule engine.

[0030] S2: Collects multi-dimensional status signals from the travel information host, programmable logic controller, application execution agent and passenger interaction terminal, and adds timestamps.

[0031] In this embodiment, multidimensional status signals refer to various heterogeneous data reflecting equipment operating status, external commands, business processes, and user behavior, including but not limited to digital quantities (such as the status of the emergency stop button), analog quantities (such as weighing values), event streams (such as passenger click logs), and structured messages (such as flight information command messages). The flight information host refers to the core server of the airport check-in system, responsible for flight data synchronization and command issuance; the programmable logic controller (PLC) is responsible for the real-time control and status feedback of the underlying electromechanical equipment; the application execution agent (AEA) is a middleware process running on the device's local operating system, coordinating business logic and the user interface; and the passenger interaction terminal includes touchscreens, card readers, cameras, and other input / output devices directly facing passengers. In this embodiment, a high-precision system timestamp, such as a Unix millisecond timestamp, is uniformly added at the signal acquisition entry point.

[0032] Specifically, the central processing unit (CPU) accesses the four types of data sources in parallel through multi-protocol interfaces: receiving instruction messages based on HL7 or custom binary protocols via TCP / IP connection to the flight information host; exchanging I / O status with the PLC via Modbus TCP or EtherNet / IP; acquiring business events through local IPC (such as Unix Domain Socket or gRPC) communication with the application execution agent; and reading raw sensor data from the passenger interactive terminal via USB or GPIO. Before entering the message queue of the CPU, each signal is timestamped by the acquisition thread calling the system clock function (such as clock_gettime(CLOCK_MONOTONIC)). For example, when a passenger clicks the "Start Check-in" button on the touchscreen, the interactive terminal generates a "UI_EVENT_START" event, along with the current coordinates (320, 480) and a timestamp 1700000000123; at the same time, the PLC may report "CONVEYOR_IDLE=true", and the flight information host may send "FLIGHT_STATUS=ON_TIME". All signals are sorted by timestamp for subsequent judgment. Furthermore, S3: Based on preset priority rules, multi-dimensional status signals are triggered and judged in a hierarchical priority mode, and security events, travel information host instructions, equipment and service status and passenger interaction behavior are responded to in sequence to determine the target counter service mode.

[0033] In this embodiment, the hierarchical priority mode triggering judgment is a decision-making method based on event type classification. Specifically, the multi-dimensional status signals are divided into four levels: the first level is safety events (such as emergency stops, fire linkage), the second level is flight information host instructions (such as flight cancellations, counter scheduling), the third level is equipment and service status (such as printer malfunctions, payment timeouts), and the fourth level is passenger interaction behavior (such as assistance requests, operation delays). The central processing unit maintains a priority queue, and high-priority signals can interrupt the low-priority judgment process.

[0034] For example, the central processing unit (CPU) checks the status signals of each priority level sequentially in the main loop: First, it scans the digital input points reported by the PLC. If an emergency stop relay is detected to be disconnected (corresponding to a safety event), it immediately terminates all current services, locks the hardware, and prepares to switch to the highest security level mode. If there is no safety event, it parses the latest instruction from the flight information host. If it contains the "SET_MODE=MAINTENANCE" field, it ignores other signals and directly prepares to switch to the specified mode. If there is no instruction from the flight information host, it assesses the health status of the subsystems reported by the application execution agent. If a barcode printer is found to be offline for more than 30 seconds, it is marked as a device malfunction, and service degradation is considered. Finally, only when there are no high-priority events as described above is the behavioral data of the passenger interactive terminal analyzed. If two consecutive authentication failures are detected, it is considered a potential service upgrade requirement. The judgment result of each level generates a candidate target mode, and the candidate generated by the highest priority level is ultimately selected as the actual target mode.

[0035] Specifically, in step S3, the counter service modes include fully self-service mode, semi-self-service assistance mode, manually operated mode, maintenance shutdown mode, and emergency pause mode; the steps for triggering the hierarchical priority mode include: S31: If an emergency stop signal reported by the programmable logic controller or an external safety linkage signal is detected, the emergency stop mode will be triggered immediately.

[0036] In this embodiment, the emergency stop signal refers to a hardware-level safety signal generated by the self-service check-in equipment itself or the surrounding security system, used to forcibly interrupt all operations; the external safety linkage signal refers to a global safety command from the airport's fire protection, security, or building automation systems; the emergency pause mode is a service state with the highest security level. In the emergency pause mode, all passenger-facing interactive functions are frozen, and electromechanical actuators (such as conveyors and roller shutters) are immediately de-energized or enter a braking state, retaining only basic communication and alarm capabilities. In this embodiment, the emergency stop signal or external safety linkage signal is connected to the high-priority interrupt input port of the programmable logic controller (PLC).

[0037] The programmable logic controller (PLC) connects to baggage weighing sensors, a conveyor drive, a roller shutter actuator, and an X-ray security imaging module. The weighing sensors provide baggage weight information, the conveyor drive controls baggage transport, the roller shutter actuator manages the opening and closing of the baggage compartment, and the X-ray imaging module provides images of the internal structure. The PLC acquires sensor data in real time via high-speed I / O scanning (cycle ≤10ms) and determines whether there are any "equipment malfunctions" or "safety risk events" based on preset rules. Once confirmed, it immediately generates a high-priority status signal and reports it to the central processing unit. Specifically, the PLC adopts a modular architecture: an analog input module (AI) connects to the weighing sensor (0–10V output, range 0–32kg); a digital output module (DO) controls the start / stop of the conveyor inverter and the electromagnetic lock of the roller shutter; and an industrial camera interface (such as GigE Vision) connects to the X-ray imaging module to capture cross-sectional images of the baggage at 5fps. The PLC's internal operating status monitoring logic program (written based on the IEC 61131-3 standard) continuously compares sensor readings with normal thresholds. For example, if the weighing sensor shows no change within 3 seconds but the conveyor runs out of time, it is determined as "luggage jam"; if the X-ray image has no valid data for 3 consecutive frames, it is marked as "imaging interruption".

[0038] Specifically, the PLC's digital input module is equipped with a dedicated safety circuit: an emergency stop button is connected in series in the 24V safety relay control circuit. Once pressed, the relay is de-energized, and the PLC immediately reads the corresponding DI point transitioning from high to low level. Simultaneously, the PLC listens for Modbus broadcast messages from the airport's central security system via RS485 or Ethernet. If it receives the "FIRE_ALARM=ACTIVE" field, it considers it an external safety linkage signal. When any signal is triggered, the PLC sends a "SAFETY_EVENT=EMERGENCY_STOP" status packet to the central processing unit within 10 milliseconds and automatically cuts off the conveyor motor driver enable signal and locks the baggage compartment door solenoid valve. Upon receiving this, the central processing unit, ignoring any current business processes, immediately sets the target mode to "emergency stop mode" and marks the event type, time, and signal source in its local log.

[0039] In this embodiment, the programmable logic controller (PLC) monitors for equipment malfunctions and security risks based on sensor data. Equipment malfunctions include luggage jamming, barcode printing failure, and payment channel interruption. Security risks include unattended luggage being left behind, suspicious item image features, and unauthorized personnel approaching the device. When a high-risk event is detected, the PLC generates a security event status signal and triggers either an emergency pause mode or a manual monitoring mode with the highest priority. Both types of events are monitored independently by the PLC, but security risk events are given the highest priority and can directly trigger either an emergency pause or a manual monitoring mode.

[0040] Specifically, the PLC sets up composite judgment logic for each type of event: Luggage jam: The weighing sensor detected a weight >1kg and the conveyor running time >15 seconds, but the exit photoelectric switch was not triggered; Barcode printing failed: After the PLC sends the print command to the printer, it does not receive a "PRINT_SUCCESS" feedback signal within 5 seconds; Payment channel interruption: The application execution agent notifies the PLC of "PAYMENT_TIMEOUT" via shared memory, and the PLC records this software layer exception; Unauthorized personnel approaching: The infrared proximity sensor detects a distance of <0.5m for >10 seconds, while the camera's face recognition module (deployed in the PLC coprocessor) does not match the face of the passenger in the current session.

[0041] Once any security risk event is established (such as lost luggage or suspicious items), the PLC immediately sets the "SAFETY_ALERT=HIGH" flag and sends a pulse signal to the emergency stop circuit via hardwiring, while simultaneously sending a "SECURITY_EVENT" message to the central processing unit.

[0042] Unattended baggage refers to luggage left unclaimed by passengers after they have placed it in the equipment without completing the check-in process, posing a potential security risk. The detection method for unattended baggage is as follows: The system uses a baggage weighing sensor to confirm the presence of baggage. If the sensor reading remains ≥0.5kg for 5 seconds, it is considered valid baggage placement. An infrared proximity sensor and video analysis determine whether a passenger has left the area within 1.5 meters in front of the device, and the passenger's departure time is recorded. Passenger departure is determined when the infrared proximity sensor output distance is >1.5 meters, and the top wide-angle camera confirms no moving human silhouette within a 1.5m x 1.5m area in front of the device using background subtraction. If the baggage weight remains and the passenger has been gone for more than 90 seconds, it is considered abandoned baggage.

[0043] Suspicious item image feature detection uses a convolutional neural network (CNN) model to classify X-ray images acquired by the X-ray security imaging module. Suspicious items include lithium batteries, liquid containers, and metal blocks. If the sum of the confidence scores of any suspicious item is greater than 0.8, it is marked as high-risk luggage. Specifically, the X-ray imaging module outputs a single-channel image at 800×600 resolution, which is then fed into the CNN model after being triggered by a PLC. This model is a variant of MobileNetV2 and outputs four probabilities: {background, lithium battery, liquid, metal}. The system is set so that if the sum of the confidence scores of the three categories "lithium battery", "liquid container", or "metal block" is > 0.8, it is marked as high-risk luggage. For example, for a backpack containing a power bank, the model outputs: lithium battery = 0.75, liquid = 0.05, metal = 0.12, total = 0.92 > 0.8, thus classifying it as high-risk. To further improve reliability, this embodiment also incorporates low-level image features for initial screening: the PLC first calculates the grayscale gradient entropy H(I) and texture energy E(I) of the image. If H(I) < 2.1 and E(I) > 1800, it is initially determined to be a high-density metal object, and then sent to the CNN for secondary confirmation.

[0044] S32: If a counter disabling or mode switching instruction is received from the TravelSky host, the current mode will be overwritten and the system will switch to the counter service mode specified in the instruction.

[0045] In this embodiment, the flight information host refers to the core scheduling server of the airport check-in system; the counter deactivation command is used to temporarily shut down a certain device (such as flight cut-off or area adjustment); and the mode switching command explicitly specifies the target service mode.

[0046] Specifically, the central processing unit (CPU) listens to the instruction channel of the travel information host via a dedicated TCP long connection. Each instruction includes a device ID, a target mode identifier, operator credentials, and a digital signature. The device's built-in security chip verifies the signature; upon successful verification, the target mode is immediately parsed. For example, when an international flight is canceled due to weather, the travel information host broadcasts the instruction "SET_MODE=MAINTENANCE, REASON=CANCELLED_FLIGHT" to all associated self-service check-in devices. Upon receiving this instruction, the CPU first instructs the application execution agent to save the currently processed passenger data to an encrypted buffer, then sends the "DISABLE_ALL_ACTUATORS" command to the PLC, and finally activates the target mode.

[0047] S33: If the aforementioned high-priority event is not triggered, and the programmable logic controller or application execution agent reports critical subsystem fault information, then the system will be downgraded to semi-self-service assistance mode or maintenance shutdown mode.

[0048] In this embodiment, critical subsystems refer to hardware or software modules that directly affect the core baggage check-in process, including but not limited to baggage weighing units, barcode printers, payment gateways, ID card readers, and conveyor control systems. Degradation switching is a fault-tolerant strategy: when some functions fail but the main equipment can still operate, the system automatically switches to a more secure mode with limited service capacity, rather than a complete shutdown. The semi-self-service assistance mode retains basic self-service functions (such as identity verification and baggage drop-off), but key steps (such as payment confirmation and boarding pass printing) require remote or on-site authorization; the maintenance shutdown mode completely stops external services, allowing only technical personnel to log in for diagnostics.

[0049] For example, the PLC continuously monitors the heartbeat signals of each subsystem: if the barcode printer fails to respond to status queries five times consecutively, or if the weighing sensor output value exceeds the calibration range of ±2kg for 10 seconds, the PLC generates a "SUBSYSTEM_FAULT=PRINTER_OFFLINE" or "WEIGHING_ERROR" event; the application execution agent monitors software layer anomalies, such as the payment SDK returning "NETWORK_TIMEOUT" more than three times. These fault information are uploaded to the central processing unit via the internal message bus. If there are no step S31 / S32 events at present, the central processing unit decides the degradation path based on the fault type: for example, printer failure → switch to semi-self-service assistance mode, after the passenger completes baggage drop-off, the system prompts "Please scan the paper barcode provided by the staff"; while if the conveyor motor driver reports an error, due to physical safety concerns, it directly switches to maintenance shutdown mode. During the switching process, the interface displays "Some functions are temporarily unavailable, staff will assist you," and automatically pushes a work order to the operation and maintenance platform.

[0050] S34: In the absence of the above events, adjust the current mode to a mode with higher service granularity based on the operation timeout, help request, or low interaction frequency reported by the passenger interactive terminal.

[0051] In this embodiment, higher service granularity refers to the system proactively providing more guidance, simplifying processes, or introducing human assistance. Operation timeout refers to staying at a certain step for more than a reasonable time limit; a help request is when the passenger actively clicks the "Need Help" button; "low interaction frequency" reflects hesitation or confusion in operation.

[0052] Specifically, the touchscreen driver of the passenger interaction terminal records the time, coordinates, and type of each touch event in real time and calculates the interval between adjacent operations. If a passenger stays on the "Baggage Confirmation" page for more than 90 seconds (the default timeout threshold), clicks the "Back" button three times consecutively, or actively clicks the "?" icon on the interface, the terminal generates a "USER_ASSISTANCE_NEEDED" event, along with behavioral context (such as the current page ID and operation sequence). This event is encapsulated by the application execution agent and sent to the central processing unit. If there are no S31-S33 events at present, the central processing unit adjusts the target mode from "fully self-service mode" to a more granular service mode—for example, enabling voice guidance, enlarging key buttons, skipping unnecessary options (such as frequent flyer card binding), and displaying the remote agent video window entry in the corner of the screen.

[0053] S4: The target counter service mode switch is performed only after the security verification and process cleanup steps before mode switching are completed.

[0054] In this embodiment, security verification refers to verifying whether all admission conditions of the target mode are currently met; process cleanup refers to terminating or saving the currently running business context and releasing the occupied software and hardware resources.

[0055] Specifically, once the target mode is determined, the central processing unit (CPU) first calls the verification module to check the entry conditions of the mode item by item. For example, if the target mode requires "the conveyor to be stopped," it queries the PLC for the conveyor motor feedback signal. If the conveyor is still running, the verification fails. Simultaneously, the process cleanup module notifies the application execution agent to interrupt the current passenger session, saves incomplete baggage information to a temporary cache, and sends a "Service adjustment is coming soon, please wait" message to the passenger interactive terminal. If all verifications pass and cleanup is complete (e.g., the PLC confirms the conveyor is stopped and the application agent returns "SESSION_SAVED"), the CPU records the switching log (including the original mode, target mode, and timestamp) and broadcasts the "MODE_SWITCH_PENDING" signal to all subsystems. Only after this complete process is finished is the control logic of the target mode officially activated.

[0056] Specifically, the security verification and process cleanup steps include: S41: Sends a hardware enable or disable instruction to the programmable logic controller that matches the target counter service mode to control the operating status of the conveyor, roller shutter, and weighing module.

[0057] In this embodiment, hardware enable or disable instructions refer to a set of digital output (DO) control commands issued by the central processing unit to the programmable logic controller (PLC) to activate or deactivate the drive power or control signals of a specific electromechanical actuator. Matching means that each counter service mode is associated with a hardware state configuration table during the predefined phase, explicitly specifying the allowed states of each module in that mode.

[0058] Specifically, after determining the target mode, the central processing unit (CPU) queries the locally stored "mode-hardware mapping table." For example, if the target mode is a service state requiring manual intervention, the mapping table specifies: "conveyor = DISABLED, roller shutter = LOCKED, weighing module = READ_ONLY"; while if the target mode is a fully automated state, it specifies: "conveyor = ENABLED, roller shutter = AUTO_CONTROL, weighing module = FULL_ACCESS." Based on this, the CPU generates a structured control instruction packet (e.g., JSON format: {"conveyor": 0, "shutter": 1, "scale": 2}) and sends it to the PLC via the Modbus TCP protocol. Upon receiving the packet, the PLC immediately updates its output register: setting the enable terminal of the conveyor inverter to a low level to stop it, driving the electromagnetic lock to close the roller shutter, and simultaneously switching the weighing sensor to read-only mode (disabling zeroing or calibration).

[0059] S42: Send a schema change notification to the application execution agent to adjust subsequent business process logic.

[0060] In this embodiment, the application execution agent is the core business process running on the self-service check-in device operating system; the mode change notification information is a lightweight event message used to inform the application execution agent that the current service mode has changed and that the corresponding business rule engine, flowchart or permission policy needs to be dynamically loaded.

[0061] Specifically, the central processing unit (CPU) sends a "MODE_CHANGE_NOTIFY" message to the application execution agent via a local inter-process communication (IPC) mechanism (such as Unix Domain Socket or shared memory queue). This message contains the target mode identifier (e.g., "MODE_ASSISTED"), the effective timestamp, and the reason for the change (e.g., "USER_REQUEST" or "SUBSYSTEM_FAULT"). Upon receiving this message, the application execution agent immediately terminates the current process state machine and loads a new business process definition file (e.g., an XML-formatted state transition diagram) from a pre-loaded mode configuration library. For example, in fully self-service mode, the process requires passengers to complete four steps: "identity verification → baggage weighing → payment → printing." However, after switching to semi-self-service assistance mode, the process automatically adjusts to "identity verification → baggage weighing → [waiting for agent authorization] → printing," where the payment step is skipped or handled by a remote agent. Furthermore, the application execution agent updates internal permission flags, such as disabling passengers from accessing the "system settings" menu in manual mode.

[0062] S43: If any admission condition of the target counter service mode is not met, the current mode shall be maintained and an exception log shall be generated.

[0063] In this embodiment, the admission conditions are a set of Boolean expressions predefined for each mode in step S1, used to describe the system state prerequisites necessary to enter that mode (such as "no ongoing transactions" or "all sensors online"). This embodiment emphasizes that mode switching is not a mandatory command, but rather a result triggered by conditions. If any admission condition is found to be unmet before the execution of steps S41 / S42 (e.g., the target mode requires "conveyor idle," but there is currently luggage being transported), the entire switching process is aborted, the system remains unchanged, and detailed anomaly information is recorded for post-event analysis.

[0064] Specifically, before sending hardware and business notifications, the central processing unit calls the "admission condition verifier" module to evaluate all conditions of the target mode item by item. For example, the admission conditions of a target mode include: "(1) PLC.CONVEYOR_STATUS=IDLE, (2) APP.CURRENT_SESSION=NULL, (3) SAFETY_DOOR_CLOSED=TRUE". The verifier queries the PLC status register, the session manager of the application execution agent, and the magnetic switch input point of the safety door in sequence. If condition (1) is found to be false (the conveyor is running), the switching process is immediately terminated and a structured exception record is written to the log system, which includes: timestamp, original mode, target mode, failure condition ID, and current status snapshot of each subsystem. At the same time, the central processing unit can choose to push a gentle prompt to the passenger interactive terminal, such as "The system is preparing better service for you, please wait a moment".

[0065] S5: After completing the mode switch, the human-computer interaction interface and external system status information are updated synchronously.

[0066] In this embodiment, synchronous updates include two dimensions: first, internally updating the graphical user interface (GUI) and voice prompts of the passenger interaction terminal to align with the service strategy of the new model; second, externally informing external systems such as the travel information host and the airport operation and maintenance platform of the current equipment status to facilitate global scheduling and monitoring.

[0067] Specifically, after a successful mode switch, the central processing unit immediately sends an interface configuration command to the passenger interactive terminal. For example, if the switch is to a mode that requires manual intervention, the GUI hides the "Print your boarding pass yourself" button, displays the "Please wait for staff assistance" animation, and activates the microphone to prepare for voice communication. At the same time, it pushes a status message to the TravelSky host via MQTT or HTTP RESTful API, which includes the device ID, current mode identifier, availability status (AVAILABLE / UNAVAILABLE), and last operation time.

[0068] In one embodiment, such as Figure 2 As shown, in step S2, the passenger interactive terminal includes a camera, an ID card reader, a touchscreen operation log module, and an infrared proximity sensor; the application execution agent performs the following identification tasks based on the video stream and operation log uploaded by the passenger interactive terminal: S201: Identify the current passenger's identity information, current operation stage, operational proficiency, and abnormal behavior; abnormal behavior includes prolonged inactivity, repeated accidental touches, baggage drop failure, and multiple failed identity verifications.

[0069] In this embodiment, the touchscreen operation log module's touchscreen driver has a built-in log hook that captures InputEvents from the Android / Linux system in real time and writes them to a circular buffer. An infrared proximity sensor is mounted above the screen, with a detection range of 0.3–2.0 meters, and outputs an analog voltage value which is converted to digital distance via an ADC. Identity information refers to the unique passenger identifier obtained through cross-verification of legal documents and biometrics; the current operation stage refers to the logical node in the current business process, such as document scanning, baggage confirmation, and payment completion; operational proficiency is a comprehensive assessment of the passenger's operational fluency and accuracy; abnormal behavior refers to observable events that deviate from the normal interaction paradigm.

[0070] Specifically, the application execution agent maintains a "passenger session state machine" whose state transitions are driven by sensor events. For example, when the ID card reader successfully reads the ID card and the camera's face liveness detection passes (using simple rules based on blinking and slight head movements), it marks "identity information confirmed"; when the system detects a signal that the luggage has been placed in place (from a sudden change in the weighing sensor) and the passenger clicks the "confirm" button, it enters the "luggage confirmation" stage.

[0071] Operational proficiency is assessed using statistical indicators: if the average time for a certain step exceeds twice the historical median, or the frequency of erroneous operations (such as clicking on invalid areas) exceeds 5 times per minute, it is marked as "low proficiency". Abnormal behavior is determined using preset thresholds: Long period of inactivity: Staying on any interactive page for more than 120 seconds; Repeated accidental clicks: The same invalid area is clicked ≥3 times consecutively; Baggage drop-off failed: The weighing sensor detected weight but the conveyor did not start, and the timeout period exceeded 60 seconds; Identity verification failed multiple times: the similarity of three consecutive face comparisons was <0.6.

[0072] Once any exception is triggered, the application execution agent immediately generates a "USER_BEHAVIOR_ALERT" event, along with the behavior type and context.

[0073] S202: Operation proficiency is determined by comparing the current operation sequence with the historical normal user template. If the reconstruction error exceeds the preset error threshold, the user is marked as a low-proficiency user.

[0074] In this embodiment, the historical normal user template refers to the standard behavioral patterns extracted from a large number of passenger operation logs that have successfully completed baggage check-in, and is stored in the form of a statistical feature vector of operation sequences. The reconstruction error can be calculated using Euclidean distance or weighted difference scores.

[0075] Specifically, before deployment, a "normal template library" was built by analyzing 10,000 successful sessions offline. Each template includes the expected time for each step (e.g., document scanning: 5±2 seconds), the distribution of effective click areas (e.g., 90% of users click on the central 30% area of ​​the screen), and the compliance rate of the operation order (e.g., 98% of users follow the order of "document → luggage → payment").

[0076] During online execution, the application execution agent extracts the current passenger's operation feature vector in real time. ,in The time taken for step i The click coordinates are normalized values. Then, the operation feature vector and the nearest neighbor template vector are calculated. Weighted Euclidean distance: The weight The weighting is determined by the importance of each step, such as giving higher weight to the payment step. Let E be the feature vector, and k be the variable index. If E > 0.4 (0.4 is the preset error threshold), then the user is judged as having low proficiency.

[0077] S203: The recognition result is used as a status signal for passenger interaction behavior and is input into the hierarchical priority mode to trigger the judgment process.

[0078] In this embodiment, the passenger interaction behavior status signal is a standardized message object that encapsulates the aforementioned identification results (identity, stage, proficiency, anomaly), including signal type, confidence level, timestamp, and associated session ID. The passenger interaction behavior status signal is injected into the fourth priority judgment layer of the central processing unit, i.e., the lowest priority, and only takes effect when there are no security, instruction, or fault events.

[0079] Specifically, the application execution agent packages the recognition results into a JSON format message, which is then published via the local message bus.

[0080] In one embodiment, the specific process for determining operational proficiency in step S2 is as follows: S211: The application execution agent obtains the passenger operation logs of successfully completed historical check-in. Each step in the passenger operation log includes the operation type, time consumption, and touch coordinates.

[0081] In this embodiment, the passenger operation log of successfully completed baggage check-in refers to the complete user interaction sequence that is automatically recorded and quality-screened during normal operation of the equipment, and which ultimately resulted in the successful completion of baggage check-in.

[0082] Specifically, after each passenger successfully prints their boarding pass and drops off their baggage, the application execution agent writes the complete operation flow of this session to a local encrypted database and marks it as "SUCCESS". The system background periodically runs a log cleaning script to remove sessions containing timeouts, error codes, or manual intervention, retaining approximately 10,000 high-quality samples for model training. Each sample is split into a time-ordered sequence of operations. The training data is normalized to form a uniform-dimensional input tensor.

[0083] S212: Train a baseline model of normal behavior using an LSTM autoencoder, and obtain the latent vectors based on the baseline model of normal behavior.

[0084] In this embodiment, the LSTM autoencoder is an unsupervised deep learning architecture composed of a Long Short-Term Memory (LSTM) network, consisting of an encoder and a decoder. Normal operation sequences possess inherent temporal regularity and can be effectively compressed using low-dimensional latent vectors; while abnormal sequences, deviating from this regularity, are difficult to reconstruct accurately. This embodiment only uses the encoder portion to extract features, and the training objective is to minimize the mean squared error (MSE) between the input sequence and the reconstructed sequence. After training, the fixed-length vector output by the encoder is the "latent vector," representing the abstract representation of the operation sequence in the latent space.

[0085] Specifically, the application execution agent calls the built-in lightweight deep learning inference engine (such as TensorFlow Lite for Embedded Linux) to load the pre-trained LSTM autoencoder model. The structure of the LSTM autoencoder model is as follows: the encoder contains two LSTM layers (64 units per layer), outputting a 16-dimensional latent vector z; the decoder symmetrically reconstructs the original sequence in reverse. The training phase is completed on an offline server: 10,000 cleaned successful logs are divided into a training set (80%) and a validation set (20%), using the Adam optimizer with a learning rate of 0.001, and trained for 200 epochs until the validation loss converges (MSE < 0.05). After training, only the encoder is deployed to the device. For example, after a standard operation sequence is processed by the encoder, the output latent vector z = [0.23, -0.41, 0.87, ..., 0.05] ∈ R 16 This vector is stored in the local cache as a digital fingerprint of "normal behavior".

[0086] S213: Obtain the current passenger's operation sequence and encode the operation sequence into the current latent vector of the current passenger's operation sequence; calculate the reconstruction error based on the latent vector and the current latent vector.

[0087] In this embodiment, the current passenger's operation sequence refers to all valid operation events of the passenger using the device from the start of the session to the current moment, formed as an input sequence after undergoing the same preprocessing process (normalization, alignment) as the historical logs. The current latent vector is the feature vector generated by the current passenger's operation sequence through a trained LSTM encoder. The reconstruction error here actually refers to the minimum distance between the current latent vector and the normal latent vector set (since decoding and reconstruction are not required, the representation space distance is used instead of the traditional reconstruction error), and this distance is used to quantify the degree of deviation between the current behavior and the normal pattern.

[0088] Specifically, during online inference, the application execution agent collects the current passenger's operation events in real time. When the session reaches a critical decision point (such as before baggage confirmation), the accumulated operation sequence is encapsulated in S211 format and input into the LSTM encoder to obtain the current latent vector. ∈R 16 Subsequently, the system calculates... With the normal latent vector set cached locally The Euclidean distance is calculated, and the minimum value is taken as the final error. , where N is the total number of reference latent vectors used to characterize "normal behavior".

[0089] For example, a skilled passenger quickly completes the operation. With a certain high-efficiency template The distance is 0.12; while a hesitant passenger's operating path is chaotic, its With all normal The distances are all greater than 0.5, indicating significant deviation.

[0090] S214: Compare the reconstruction error with the preset error threshold to identify low-skilled users and generate a status signal for suggested enhanced guidance, which is then reported to the central processing unit.

[0091] In this embodiment, the preset error threshold is an empirical constant derived from the statistical analysis of error distribution on an offline validation set, used to delineate the decision boundary between normal and abnormal behavior. This embodiment sets this threshold to 0.35; that is, when E > 0.35, the current user is determined to be either "low proficiency" or "abnormal operation." This determination does not directly interrupt the process but is instead transformed into a high-priority service optimization suggestion signal.

[0092] Specifically, when E=0.42>0.35, the application execution agent immediately constructs an "ENHANCED_GUIDANCE_REQUEST" status signal, which includes: session ID, error value, and suggested actions. This status signal recommending enhanced booting is sent to the central processing unit via the local IPC channel. Upon receiving this signal in stage S34, if there are no higher priority events, the central processing unit updates the target mode candidate to a service mode that supports enhanced booting. For example, the system then overlays a semi-transparent AR arrow on the touchscreen pointing to the "Next" button and activates the TTS voice module to play: "Please click the green confirmation button to continue."

[0093] In one embodiment, in step S2, the application execution agent constructs a passenger behavior prediction model based on historical passenger interaction data and equipment operation logs, specifically including: S221: Using real-time multidimensional state signals as input, predict the probability of failure for the current passenger to complete the next operation step; the passenger behavior prediction model uses a temporal convolutional network, with the input tensor dimension being the multidimensional state signals within the past 30 seconds, and the output being the probability of operation failure within the next 10 seconds.

[0094] In this embodiment, the real-time multidimensional status signals include operation event type, page dwell time, number of errors, touch frequency, number of authentication attempts, and baggage weight change rate. The "passenger behavior prediction model" is a deep learning architecture based on Temporal Convolutional Network (TCN). The failure probability is defined as the likelihood that a passenger will not be able to successfully complete the current operation step (such as not clicking confirm, not completing payment, or abandoning the process) within the next 10 seconds, with a value range of [0, 1].

[0095] Specifically, the application execution agent maintains a sliding time window of 30 seconds, sampling 12-dimensional feature vectors at a frequency of 1Hz to form the input tensor X∈R. 30×12 For example, the eigenvector at second t is: The feature vectors are input into a pre-trained TCN model. The TCN model consists of four dilated convolution layers with dilation coefficients of 1, 2, 4, and 8, each with 64 channels, followed by global average pooling and a sigmoid output layer. The TCN model is trained offline using 100,000 historical session data points, with labels y=1 indicating "current step not completed within 10 seconds" and y=0 indicating successful completion. Training uses binary cross-entropy loss, and the optimizer is AdamW. During deployment, the model runs on a device GPU using the TensorRT engine, with a single inference time of <150ms.

[0096] S222: When the failure probability is greater than the preset failure threshold, the application execution agent generates a high failure risk status signal, so that the central processing unit adjusts the current mode to a semi-self-service assistance mode with higher service granularity when there are no high priority events, and optimizes the layout of the human-computer interaction interface.

[0097] In this embodiment, the preset failure threshold is an empirical constant determined through validation set ROC curve analysis, set to 0.7 in this embodiment. That is, when the predicted failure probability p > 0.7, the system determines that the current passenger is highly likely to abandon the operation, requiring proactive intervention. The high failure risk status signal is a structured message containing metadata such as the predicted probability, current step ID, and suggested measures. The high failure risk status signal is injected into the fourth priority judgment layer (i.e., S34 level) of the central processing unit, and only takes effect when there are no safety events, flight information instructions, or equipment failures, ensuring the rationality and appropriate timing of intervention.

[0098] Specifically, when the central processing unit receives a high failure risk status signal, if there are no S31–S33 events at present, the central processing unit parses the high failure risk status signal in stage S34 and performs the following actions: (1) Update the target mode candidate to "semi-self-service assistance mode"; (2) Issue interface optimization instructions to passenger interactive terminals: highlight the "Manual Assistance" button, disable the "Back" function (to prevent process rollback from causing confusion), and pop up a remote agent video window preview on the right side of the screen; (3) Simultaneously send an assistance request to the airport's intelligent dispatch platform to reserve the nearest available check-in staff.

[0099] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0100] In one embodiment, a self-service check-in device counter mode adaptive detection and switching system is provided, which corresponds to the self-service check-in device counter mode adaptive detection and switching method in the above embodiment.

[0101] A self-service baggage check-in device counter mode adaptive detection and switching system includes a central processing unit, a multi-source status signal acquisition module, a hierarchical priority decision engine, a security verification and process cleanup unit, and a mode execution and synchronization module. Detailed descriptions of each functional module are as follows: The central processing unit is used to predefine multiple mutually exclusive counter service modes, each of which is configured with corresponding admission and exit conditions. The multi-source status signal acquisition module is communicatively connected to the travel information host, programmable logic controller, application execution agent and passenger interaction terminal. It is used to collect multi-dimensional status signals from various modules in real time and add a unified timestamp to the multi-dimensional status signals. The hierarchical priority decision engine, built into the central processing unit, is used to perform hierarchical priority mode triggering judgment on multi-dimensional status signals based on preset priority rules, and respond sequentially to security events, travel information host instructions, equipment and service status and passenger interaction behavior, and output the target counter service mode. The security verification and process cleanup unit is used to verify whether the access conditions of the target counter service mode are met before the execution mode switch, and to complete the interruption cleanup of the current business process. The mode execution and synchronization module is used to activate the target counter service mode after passing the security verification and process cleanup steps, and to synchronously update the content of the human-computer interaction interface and broadcast the current mode status information to external systems.

[0102] For specific limitations regarding the self-service baggage check-in equipment counter mode adaptive detection and switching system, please refer to the limitations of the self-service baggage check-in equipment counter mode adaptive detection and switching method mentioned above, which will not be repeated here; each module in the above self-service baggage check-in equipment counter mode adaptive detection and switching system can be implemented in whole or in part by software, hardware and their combination; each module can be embedded in the processor of the computer device in hardware form or independent of the processor, or it can be stored in the memory of the computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0103] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps of the self-service check-in device counter mode adaptive detection switching method as described above.

[0104] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0105] In one embodiment, particularly according to an embodiment of the invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, embodiments of the invention include a computer program product comprising a computer program / instructions that, when executed by a processor, implement the steps of a counter mode adaptive detection and switching method for a self-service check-in device as described above. In such embodiments, the computer program can be downloaded and installed from a network via a communication module, and / or installed from a removable medium. When the computer program is executed by a central processing unit (CPU), it performs the various functions defined in this invention.

[0106] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0107] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method for adaptive detection and switching of counter mode for self-service baggage check-in equipment, characterized in that, include: Multiple counter service modes are predefined in the central processing unit of the self-service check-in equipment. Each counter service mode has mutually exclusive entry and exit conditions. Collect multi-dimensional status signals from the travel information host, programmable logic controller, application execution agent, and passenger interaction terminal, and attach timestamps; Based on preset priority rules, the multi-dimensional status signals are subjected to hierarchical priority mode triggering judgment, and security events, travel information host instructions, equipment and service status and passenger interaction behavior are responded to in sequence to determine the target counter service mode. The mode switch for the target counter service mode is performed only after the security verification and process cleanup steps before mode switching are completed. After the mode switch is completed, the human-computer interaction interface and external system status information are updated synchronously.

2. The self-service baggage check-in device counter mode adaptive detection and switching method according to claim 1, characterized in that, The counter service modes include fully self-service mode, semi-self-service assistance mode, manual duty mode, maintenance shutdown mode, and emergency suspension mode; The steps for triggering the hierarchical priority mode judgment include: If an emergency stop signal or an external safety linkage signal is detected reported by the programmable logic controller, the emergency stop mode is immediately triggered. If a counter service or mode switching instruction is received from the TravelSky host, the current mode will be overwritten and the service mode specified in the instruction will be switched. If the aforementioned high-priority event is not triggered, and the programmable logic controller or the application execution agent reports critical subsystem fault information, then the system will be downgraded to semi-self-service assistance mode or maintenance shutdown mode. In the absence of the aforementioned events, the current mode is adjusted to a mode with higher service granularity based on the operation timeout, help requests, or low interaction frequency reported by the passenger interactive terminal.

3. The self-service baggage check-in device counter mode adaptive detection and switching method according to claim 1, characterized in that, The security verification and process cleanup steps include: Send hardware enable or disable commands matching the target counter service mode to the programmable logic controller to control the operating status of the conveyor, roller shutter door, and weighing module; Send a mode change notification to the application execution agent to adjust subsequent business process logic; If any of the admission conditions for the target counter service mode are not met, the current mode will remain unchanged, and an exception log will be generated.

4. The self-service baggage check-in device counter mode adaptive detection and switching method according to claim 1, characterized in that, The passenger interaction terminal includes a camera, an ID card reader, a touchscreen operation log module, and an infrared proximity sensor; the application execution agent performs the following identification tasks based on the video stream and operation log uploaded by the passenger interaction terminal: Identify the current passenger's identity information, current operational stage, operational proficiency, and abnormal behavior; The abnormal behaviors include prolonged inactivity, repeated accidental touches, failed baggage drop-off, and multiple failed identity verifications. The operation proficiency is determined by comparing the current operation sequence with the historical normal user template. If the reconstruction error exceeds a preset error threshold, the user is marked as a low proficiency user. The recognition result serves as a status signal for passenger interaction behavior, which is then input into the hierarchical priority mode to trigger the judgment process.

5. The self-service baggage check-in device counter mode adaptive detection and switching method according to claim 4, characterized in that, The specific process for determining the operational proficiency is as follows: The application execution agent obtains the passenger operation logs of successfully completed baggage check-in in the past. Each step in the passenger operation log includes the operation type, time consumption, and touch coordinates. A baseline model of normal behavior is trained using an LSTM autoencoder, and latent vectors are obtained based on the baseline model of normal behavior. Obtain the current passenger's operation sequence and encode the operation sequence into a current latent vector of the current passenger's operation sequence; calculate the reconstruction error based on the latent vector and the current latent vector; The reconstruction error is compared with a preset error threshold to identify low-skilled users, and a status signal suggesting enhanced guidance is generated and reported to the central processing unit.

6. The self-service baggage check-in device counter mode adaptive detection and switching method according to claim 4, characterized in that, The programmable logic controller (PLC) is connected to the baggage weighing sensor, conveyor driver, roller shutter door actuator, and X-ray security imaging module; the PLC monitors equipment malfunctions and security risk events based on sensor data. The equipment malfunctions include luggage jamming, barcode printing failure, and payment channel interruption. The security risk events include unattended luggage left behind, suspicious item image features, and unauthorized personnel approaching; When a high-risk event is detected, the programmable logic controller generates a security event status signal and triggers an emergency pause mode or a manual monitoring mode with the highest priority. The method for detecting unattended luggage left behind is as follows: The luggage weighing sensor confirms that luggage has been placed; The infrared proximity sensor and video analysis are used to determine whether a passenger has left the area 1.5 meters in front of the device and to obtain the time of the passenger's departure. If the weight of the luggage remains and the passenger has been gone for more than 90 seconds, it is considered abandoned luggage. The suspicious item image feature detection uses a convolutional neural network model to classify the X-ray images acquired by the X-ray security inspection imaging module. Suspicious items include lithium batteries, liquid containers, and metal blocks. If the sum of the confidence scores of any suspicious item is greater than 0.8, it is marked as high-risk baggage.

7. The self-service baggage check-in device counter mode adaptive detection and switching method according to claim 6, characterized in that, The application execution agent builds a passenger behavior prediction model based on historical passenger interaction data and equipment operation logs; Using real-time multidimensional state signals as input, the probability of failure for the current passenger to complete the next operation step is predicted; the passenger behavior prediction model adopts a temporal convolutional network, with the input tensor dimension being the multidimensional state signals within the past 30 seconds, and the output being the probability of operation failure within the next 10 seconds; When the failure probability exceeds a preset failure threshold, the application execution agent generates a high failure risk status signal, causing the central processing unit to adjust the current mode to a semi-self-service assistance mode with higher service granularity when there are no high-priority events, and optimize the layout of the human-computer interaction interface.

8. A self-service baggage check-in device counter mode adaptive detection and switching system, characterized in that, An adaptive detection and switching method for counter mode of a self-service baggage check-in device as described in any one of claims 1 to 7, the system comprising: The central processing unit is used to predefine multiple mutually exclusive counter service modes, each of which is configured with corresponding admission and exit conditions. The multi-source status signal acquisition module is communicatively connected to the travel information host, programmable logic controller, application execution agent and passenger interaction terminal, and is used to acquire multi-dimensional status signals from each module in real time and add a unified timestamp to the multi-dimensional status signals. The hierarchical priority decision engine, built into the central processing unit, is used to perform hierarchical priority mode triggering judgment on the multi-dimensional status signals based on preset priority rules, and respond sequentially to security events, travel information host commands, equipment and service status and passenger interaction behavior, and output the target counter service mode. The security verification and process cleanup unit is used to verify whether the access conditions of the target counter service mode are met before the execution mode switch, and to complete the interruption cleanup of the current business process. The mode execution and synchronization module is used to activate the target counter service mode after passing the security verification and process cleanup steps, and to synchronously update the content of the human-computer interaction interface and broadcast the current mode status information to external systems.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the self-service check-in device counter mode adaptive detection and switching method as described in any one of claims 1 to 7.

10. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instruction is executed by the processor, it implements the steps of the counter mode adaptive detection and switching method for a self-service check-in device as described in any one of claims 1 to 7.

Images