Aerospace final assembly worker adaptive AR training method and system based on cognitive load real-time monitoring

Abstract

The invention discloses a cognitive load real-time monitoring-based self-adaptive AR (augmented reality) training method and a cognitive load real-time monitoring-based self-adaptive AR training system for spaceflight final assembly workers, and relates to the crossing field of augmented reality technologies and intelligent manufacturing. Comprising electroencephalogram signals, eye movement tracks, heart rate variability, electromyographic signals, voice data, hand actions and electrodermal response; the physiological signals are analyzed in real time through a processing unit, and a cognitive load index and an emotional state index are calculated; when the cognitive load index exceeds a first threshold value, automatically adjusting a display mode of the AR training content, including reducing information density, increasing fonts or switching a demonstration mode; when the cognitive load index continuously exceeds a second threshold value, a forced pause mechanism is triggered, the station is locked, and an alarm is given out; collected data is anonymously uploaded to a central system after being locally processed, and is used for optimizing a training strategy. The system comprises a sensor module, a processing module, an AR display module, an intervention module and a communication module.

Description

Technical Field

[0001] This invention relates to the intersection of augmented reality technology and intelligent manufacturing, specifically to an adaptive AR training method and system for aerospace assembly workers based on real-time monitoring of cognitive load. Background Technology

[0002] Aerospace assembly is a typical knowledge-intensive and skills-intensive field, where assembly precision directly determines the success or failure of hundreds of millions or even billions of dollars in national assets. Currently, domestic and international aerospace manufacturing companies have begun to experiment with AR technology to assist in assembly training, such as the static AR prompting system used by Boeing. These existing technologies primarily use AR glasses to overlay 3D models, assembly steps, and other information onto the actual workpiece, reducing the burden of blueprint reading for workers.

[0003] However, existing AR training systems have five fatal flaws that fail to meet the aerospace assembly's extreme requirement of "zero errors": 1. Inability to perceive cognitive load and emotional state in real time: The existing system can only output information unidirectionally and cannot perceive whether workers experience negative states such as cognitive overload, tension, or frustration when facing complex assembly tasks. The core risk point of workers "making mistakes when nervous" has not been effectively monitored and intervened.

[0004] The guidance strategy is rigid and unable to adapt: ​​the existing system presents information in a fixed way, and cannot dynamically adjust the information density, complexity, or guidance method according to the worker's real-time cognitive load level. When workers are under high load, the fixed information flow will actually increase their burden and lead to errors.

[0005] Lack of effective mandatory intervention mechanisms: When workers are on the verge of or already on the verge of cognitive collapse, the existing system has no physical "brake" mechanism to prevent erroneous operations from occurring. It can only trace back after the fact, and cannot prevent it in advance, resulting in huge economic losses.

[0006] The system has a low level of intelligence and cannot continuously evolve: the training strategy of the existing system is based on preset rules and cannot utilize the large amount of data generated in the real production environment (especially cognitive load data with physiological labels) for continuous learning and optimization. It is difficult to cope with increasingly complex assembly tasks and the individual differences of new workers.

[0007] Data security faces a national-level risk: workers' raw physiological signals, such as electroencephalograms (EEGs), are extremely sensitive biological data. Existing systems lack adequate security design, and if this data is leaked during transmission or storage, it will pose a significant national security risk.

[0008] Therefore, there is an urgent need in this field for an intelligent AR training solution that can deeply integrate cognitive science, real-time intervention, continuous evolution, and ensure absolute data security in order to overcome the long-standing problem of "human error" in the aerospace assembly field. Summary of the Invention

[0009] The purpose of this invention is to provide an adaptive AR training method and system for aerospace assembly workers based on real-time monitoring of cognitive load. By sensing the physiological state of workers in real time, the information presentation method and intervention strategy are dynamically adjusted, thereby improving the adaptability and effectiveness of the training process while ensuring operational safety.

[0010] To achieve the above objectives, the present invention is implemented through the following technical solution: an adaptive AR training method and system for aerospace assembly workers based on real-time monitoring of cognitive load. The method and system are characterized by adaptive adjustment through a series of steps. In step (1), the AR glasses integrate multiple sensors to continuously collect multimodal physiological signals of workers in a non-invasive manner: electroencephalogram (EEG) signals monitor brainwave activity through scalp electrodes, eye movement trajectories track eye movements through an infrared camera, heart rate variability is detected by photoelectric sensors, electromyography (EMG) signals measure muscle electrical activity through surface electrodes, speech data records speech features through a microphone, hand movements are captured by an inertial measurement unit (IMU), and skin conductance is monitored by electrodes to monitor changes in skin conductivity. In step (2), the processing unit (such as an embedded microprocessor) performs real-time filtering, feature extraction, and fusion analysis on the raw physiological signals, and uses predefined mathematical models (such as weighted average or threshold comparison) to calculate the cognitive load index (reflecting the degree of psychological burden) and emotional state indicators (such as stress or focus). In step (3), when the cognitive load index exceeds the first threshold (based on pre-calibration settings), the system automatically adjusts the display method of the AR training content: for example, by reducing the information density by reducing text or graphic elements in the interface, increasing the font size to improve readability, or switching from 3D animation mode to static image demonstration mode. In step (4), if the cognitive load index continues to exceed a higher second threshold within a specified time, the system triggers a forced pause mechanism, locks the workstation equipment through software instructions (such as disabling tool power), and activates a local sound and light alarm to alert workers and monitoring personnel. In step (5), the collected physiological data is anonymized locally on the AR glasses (such as deleting personal identifiers), and then uploaded to the central system through an encrypted channel for statistical analysis and training strategy optimization (such as adjusting the content difficulty).

[0011] Furthermore, the adjustments to the AR training content specifically include simplifying complex 3D assembly drawings. When cognitive load is high, the system automatically converts the 3D assembly model into a 2D key step view, extracting the core sequences in the assembly process (such as component installation order) and presenting them in a simplified schematic diagram. Simultaneously, the system uses highlight colors or flashing effects to emphasize the optimal operation path (such as tool movement trajectory or part positioning points), reducing visual interference and helping workers focus on critical tasks.

[0012] Furthermore, the assessment of the emotional state indicators incorporates speech tone analysis. The system analyzes the acoustic features of the speech data (such as pitch, speech rate, and intensity) to help determine the worker's emotional state (such as anxiety or calmness). Based on the assessment results, the system selects personalized voice prompts from a pre-recorded voice library: encouraging voices include positive feedback statements (such as "operating well"), operational instructions include step-by-step instructions (such as "please check the connectors"), and the prompt style is dynamically selected based on the worker's historical responses to enhance training adaptability.

[0013] Furthermore, the data acquisition and processing emphasize localization and security. All physiological signals are analyzed in real time on the processor built into the AR glasses, and the raw data is always stored in local memory and is not transmitted to external networks. The system is equipped with security monitoring circuitry that activates a hardware self-destruct mechanism (such as melting the data bus or erasing the storage chip) if an unauthorized transmission attempt is detected (e.g., unauthorized port access) to ensure physical data isolation.

[0014] Furthermore, the analysis of the processing module is based on an updatable rule base. The rule base contains a series of conditional rules (such as "if the eye movement frequency exceeds X, then increase the cognitive load index"), which are periodically revised based on historical operational data (such as past training records) and updated manually or automatically to optimize calculation accuracy.

[0015] Furthermore, the system supports special scenarios in aerospace assembly, such as zero-gravity environments. In this environment, the system detects changes in gravity through environmental sensors (such as accelerometers) and automatically adjusts the cognitive load threshold (e.g., raising the threshold to compensate for the physiological adaptation period) to ensure that the training content matches the environmental conditions.

[0016] Furthermore, the system is integrated with the worker performance management system. Cognitive load events (such as pause trigger records) are tagged and linked to worker profiles for performance evaluation (such as efficiency scores) and training improvement (such as identifying weaknesses), forming a closed-loop feedback loop.

[0017] Furthermore, the system possesses offline operation capabilities. All key functions (such as signal processing, content adjustment, and pause mechanisms) are implemented locally by hardware and software, without relying on external network connections. The system can still operate fully without a network, providing training content and support through a local repository.

[0018] This invention provides an adaptive AR training method and system for aerospace assembly workers based on real-time cognitive load monitoring, which has the following beneficial effects: 1. This system collects various physiological signals from workers in real time, such as EEG and eye-tracking data, and dynamically calculates the cognitive load index. This allows it to automatically adjust the display of AR training content based on the worker's current state. For example, when it detects worker distraction or increased stress, the system simplifies 3D assembly drawings into 2D views, increases font size, or switches presentation modes to make the training content easier to understand. This adaptive adjustment effectively reduces the cognitive burden on workers during complex operations, making the training process more tailored to individual needs and significantly improving learning efficiency and operational accuracy. Workers can master key skills more quickly, reducing errors caused by information overload, thereby increasing the first-time success rate in high-precision operations such as aerospace assembly and shortening the training cycle.

[0019] When the system detects that a worker's cognitive load is persistently high, it triggers a forced pause mechanism. This mechanism uses hardware relays to lock the workstation and issue an alarm, ensuring the worker can rest immediately. This effectively prevents workers from performing high-risk operations under fatigue or high pressure, avoiding potential major errors and product scrap. Simultaneously, the forced pause is linked to the central monitoring system, recording the event for subsequent analysis. This not only protects the physical and mental health of workers but also improves workplace safety. Through timely intervention, the system minimizes potential quality risks, reduces waste generation, saves the company significant costs, and maintains the high quality standards required for aerospace assembly.

[0020] All physiological data collection and processing are completed locally on the AR glasses. Raw signals are not transmitted externally; only anonymously uploaded summary data is uploaded for strategy optimization. If an unauthorized transmission attempt is detected, the system activates a hardware self-destruct mechanism to ensure data security. This design greatly protects worker privacy, prevents the leakage of sensitive information, and meets the confidentiality requirements of the aerospace industry. Simultaneously, the system has offline operation capabilities, does not rely on a network connection, and ensures stable operation in environments without a network (such as remote factories or special scenarios). This reliability and security enhance the company's data management capabilities and provide a solid foundation for long-term training.

[0021] The system supports the unique environments of aerospace assembly, such as zero-gravity scenarios, and can automatically adjust cognitive load thresholds to adapt to different conditions. Furthermore, it is deeply integrated with worker performance management, using cognitive load event records for evaluation and training improvement, making management more scientific. For example, the system can optimize training content based on worker performance, promoting individual growth. This adaptability ensures the system's effectiveness in various real-world scenarios, helping companies respond quickly to changes and improve overall training quality. By integrating performance data, the system can also identify outstanding workers as instructor candidates, accelerating talent pipeline development and enhancing corporate competitiveness.

[0022] By analyzing voice tone and generating personalized voice prompts, such as encouraging voices or operational instructions, the system provides emotional support to workers, enhancing the interactivity and motivation of training. This human-centered design reduces the psychological stress on workers under high-pressure environments, making training more acceptable. Combined with a modular architecture, the system is easy to maintain and upgrade, enabling it to serve the company's needs long-term. Overall, the system not only improves the effectiveness of skills training but also enhances workers' sense of belonging through its care mechanisms, helping to reduce employee turnover and providing strong support for the sustainable development of the aerospace assembly industry. Attached Figure Description

[0023] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings in the following description are merely exemplary, and those skilled in the art can derive other embodiments based on the provided drawings without creative effort.

[0024] Figure 1 This is the main workflow diagram of the system of the present invention; Figure 2 This invention provides a flowchart for the adaptive adjustment of AR training content. Figure 3 This is a flowchart illustrating the forced pause and safety warning process of this invention. Figure 4 This is a flowchart illustrating the data security and processing procedures of this invention. Figure 5 This is a flowchart of the system module linkage and optimization process of the present invention. Detailed Implementation

[0025] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numerals in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this disclosure. Rather, they are merely examples of apparatuses consistent with some aspects of this disclosure as detailed in the appended claims.

[0026] How to use: Phase 1: System Preparation and Startup 1. Device Wearing and Inspection: Operators should correctly wear the AR glasses integrated with multimodal physiological signal sensors. Check that the sensor contacts of the AR glasses are in good contact with the skin, and that the device's power supply and self-test status are normal.

[0027] Workstation Environment Verification: Confirm that the operator's workstation is correctly connected to the system's intervention module (hardware relay) and ensure that the workstation locking function is available.

[0028] System startup and identity association: The AR glasses and local processing unit are started. The system initializes, the operator's identity information is bound to the training task, and the system loads the corresponding training content rule base.

[0029] Phase Two: Training Process Implementation and Adaptive Interaction 4. Signal Acquisition and Real-time Calculation: The operator begins to follow the instructions displayed on the AR glasses. The system continuously acquires EEG signals, eye movement tracking, heart rate variability, electromyography (EMG) signals, speech data, hand movements, and skin conductance. The integrated local processing module processes these physiological signals in real time, calculating real-time cognitive load and emotional state indicators.

[0030] Initial adaptive content adjustment: The processing module compares the calculated cognitive load index with a preset first threshold. When the index exceeds the first threshold, the system automatically instructs the AR display module to adjust the display method of the training content. Adjustment measures include: reducing information density, increasing the font size of key information, or switching complex 3D assembly animations to simplified 2D key step views and highlighting the optimal operation path.

[0031] Voice prompts and emotional support: The system simultaneously analyzes the operator's voice data to help assess their emotional state. Based on the emotional state analysis results, the system generates personalized voice prompts, which may include preset encouraging statements or specific guidance for the current operation step.

[0032] Secondary Intervention and Forced Suspension: If the cognitive load index remains at a high level and exceeds a higher second threshold, it indicates that the operator may be at risk of overload. At this time, the system's intervention module is triggered, forcibly locking the current workstation via a hardware relay, suspending all operations, and simultaneously issuing visual and auditory alarms through the AR glasses. This forced suspension event is pushed to the central monitoring system for recording via the communication module.

[0033] Phase Three: Training Conclusion and Data Processing 8. Training Completion and Data Upload: Upon completion of a single training session, the system ceases data collection. All physiological data processed locally, along with its analysis results, after removing the operator's personal identification information, is anonymously uploaded to the central system via the communication module for optimizing the overall training strategy and rule base.

[0034] Performance correlation and analysis: Key events recorded by the system, such as the number of times the cognitive load threshold is triggered, content adjustment records, and forced pause events, can be correlated with the performance management of operators, providing data support for training effectiveness evaluation and improvement.

[0035] Example: Example 1: Training on Precision Cable Assembly for Satellite Payloads During a training session on cable laying and connection for a new type of communication satellite payload bay, senior technician Zhang practiced using the system. The task involved hundreds of cables of different specifications, with strict process requirements for their routing, fixing points, and interface sequence. At the start of the training, AR glasses overlaid a complete 3D model of the cable bundle in his field of vision, marking the connection points for each step. Zhang then had to perform precise operations within the simulation bay according to the prompts.

[0036] Approximately fifteen minutes into the operation, the system's built-in sensors detected a specific frequency increase in Mr. Zhang's EEG signal. Simultaneously, the frequency of AR prompts from his eye-tracking scans significantly increased, and heart rate variability data indicated a rise in his psychological stress level. Based on these multimodal signals, the local processing unit determined that his cognitive load index had exceeded a preset first threshold. The system then triggered an adaptive adjustment mechanism: the previously complex three-dimensional cable bundle model was replaced with a series of simplified two-dimensional key-step views. Each step highlighted only the two to three cables requiring processing and their optimal laying path, while other non-critical information was temporarily hidden. Simultaneously, the font size of the prompts automatically increased. Furthermore, by analyzing the tone of Mr. Zhang's subconscious, brief self-talk during the operation (such as "This clamp position..."), the system determined that he was slightly confused and played a preset, calm voice instruction through his bone conduction headphones: "Please first confirm the orientation of the blue-marked cable connector." These adjustments reduced information complexity, allowing Mr. Zhang to refocus his attention. Approximately five minutes later, the physiological signals gradually returned to normal, and the system subsequently resumed its standard 3D view display, allowing the training to proceed smoothly until completion. All stress events and system intervention records from this training were anonymously uploaded to the central system for analysis of the challenges of this training module, providing a basis for subsequent optimization.

[0037] Example 2: Training on Pre-tightening Bolts for Launch Vehicle Section Docking New employee Li is undergoing training on the pre-tightening of high-strength bolts at the docking surface between the rocket's first-stage oxidizer compartment and fuel tank. This operation requires the use of a specialized torque wrench to tighten dozens of bolts in three progressively increasing torque steps in a specific diagonal sequence, demanding extremely high precision. During the training, an AR system tracks his wrench movements and sequence through gesture recognition. Initially, Li was able to complete the task methodically. However, as the task progressed, his movements became hurried due to concerns about making mistakes. Electromyography (EMG) signals in his hand showed continuous muscle tension, and his skin conductance also increased sharply.

[0038] The system processing unit calculated that the cognitive load index was rapidly increasing and quickly exceeded the set second threshold. The system immediately activated the forced pause mechanism: first, the training content in the AR field of view was covered by a prominent red warning icon, accompanied by a continuous buzzing alarm. Simultaneously, the processing module sent an electrical signal to the hardware relay connected to the simulated module docking platform, triggering a physical locking mechanism that locked the training torque wrench in Li's hand, preventing operation, and also secured the simulated module to prevent any potential misoperation. The system interface displayed: "Excessive load detected, operation paused. Please remain in place and take deep breaths." At the same time, this event was pushed to the training center's central monitoring system in real time, and an alarm message popped up on the duty instructor's monitoring screen, displaying the workstation number and event type. The instructor immediately went to the scene to provide face-to-face guidance and psychological support to Li. After confirming that Li's condition had stabilized, the instructor unlocked the workstation with the appropriate permissions, and the training resumed from the steps before the pause. This forced pause record will serve as important data to assess Li's stress resistance and operational stability, and will be included in his periodic performance evaluation to help instructors develop more targeted follow-up training plans.

[0039] Example 3: Zero-gravity environment adaptation training of space station experimental module At an underwater training base simulating a weightless environment, Engineer Wang wore specially designed waterproof AR glasses to undergo adaptive training on installing scientific instruments inside the space station's experimental module. Because the underwater environment simulates microgravity, the operator's body posture control and tool usage differ from those on land, resulting in a higher baseline level of cognitive load. The system had been pre-set to automatically increase the cognitive load threshold based on "zero-gravity environment" parameters to adapt to this special working condition.

[0040] The training task required installing a precision spectrometer into a designated cabinet within the bulkhead. While attempting to secure the equipment, Engineer Wang needed to constantly adjust his body posture to counteract the buoyancy of the water while simultaneously performing precise alignment. The AR system monitored his eye movements in real time, which frequently switched rapidly between the equipment interface, the securing clips, and his own hand and foot anchor points. EEG signals indicated a highly strained allocation of his attentional resources. Despite the high workload, it did not exceed the adjusted second threshold. The system primarily adopted a content simplification strategy: replacing the original dynamic 3D animation demonstrating the installation process with a series of static, high-contrast 2D diagrams, clearly marking the docking sequence and required tools for each anchor point. Simultaneously, to avoid increasing the burden on the auditory channel, the system only played a short, affirmative sound effect after each successful sub-step, rather than lengthy voice instructions. These adjustments, adapted to the characteristics of the microgravity environment, effectively helped Engineer Wang maintain operational accuracy in the complex environment. Throughout the training process, due to the unreliable network connection in the underwater environment, all data acquisition, processing, and interaction were completed locally on the AR device, ensuring the continuity and reliability of the training. Only after the training was completed were the data uploaded to the central ground system for analysis.

[0041] Example 4: Training on Cleaning and Inspection Procedures for Precision Components of Inertial Navigation Systems Technician Zhao was responsible for training on the pre-assembly cleaning and inspection of key gyroscope components in the rocket's inertial navigation system. This process was extremely meticulous, involving various cleaning solvents, the order and techniques for using lint-free cloths, and inspection standards under high magnification. During the training, an AR system tracked his hand movements via camera to ensure compliance with cleanroom operating procedures.

[0042] During the training, at the crucial mirror cleaning step, technician Zhao, worried about residual fibers, repeatedly wiped the same area. The system detected abnormal repetitive movements in his operation, and combined with fluctuations in his heart rate variability data, calculated a consistently high cognitive load index. The system determined that he might be overly focused on a localized area, neglecting the overall process. Therefore, the system first adjusted the AR display, highlighting and enlarging the text instructions for the current operation step. Next, based on the rule base's response strategies for "hesitation and repetition," the system did not directly play the instructional voice, but instead triggered a short demonstration video, showing the correct, one-way wiping technique at standard speed. This video interrupted technician Zhao's repetitive actions, providing a visual reference. Simultaneously, by analyzing the slight tremor in his voice when he silently recited the check items earlier, the system determined that he was experiencing anxiety. After he completed this step, it played a pre-set encouraging voice: "Operation is correct, please continue to the next step." This intervention method, combining visual demonstration and positive feedback, effectively alleviated technician Zhao's tension and guided him back to the correct procedure. All physiological data generated in this incident were processed within the AR glasses. The raw data never left the device. Only the anonymized result of "increased load and system intervention during the cleaning step" was uploaded to assess whether the difficulty setting of this training session was reasonable.

[0043] Example 5: Comprehensive Evaluation and Exercise of Riveting Process for Large Cabin Structures During the half-day comprehensive assessment exercise, senior technician Team Leader Sun was tasked with completing multiple riveting tasks simulating a spacecraft cabin. The system not only provided guidance during the exercise but also handled some of the assessment recording functions. The tasks covered riveting in different locations and with different models, requiring frequent tool changes and adjustments to process parameters. After an hour of continuous high-intensity operation, Team Leader Sun moved on to the final and most complex riveting section on the curved section.

[0044] The system monitored multiple physiological indicators, particularly eye-tracking data (such as blink rate and fixation drift) and electromyography signals reflecting fatigue levels. The data showed that while the cognitive load did not trigger a forced pause, it remained consistently high. Based on this, the system dynamically adjusted the level of detail in the prompts. For example, for standard riveting steps that Team Leader Sun had already mastered, only brief icon prompts were displayed; while for the complex curve riveting, detailed information such as the riveting gun's angle and pressure parameters was shown. After the exercise, the system generated a detailed report, including not only the task completion time and quality but also the cognitive load data curve, marking the operational nodes with higher loads. This report was synchronized to the performance management system. In subsequent evaluations, the training supervisor could combine task completion and load data for a comprehensive assessment: for example, Team Leader Sun maintained high-precision operation even under high load conditions, demonstrating excellent skill stability and psychological resilience. The report also suggested that the training content for riveting curve sections might need further breakdown and optimization to reduce the overall load level. Throughout the exercise, the system operated independently in a secure workshop without network connectivity. All data was encrypted and stored locally, and after the exercise, it was unidirectionally imported into the central system through a secure interface, ensuring the security of core process data.

[0045] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An adaptive AR training method and system for aerospace assembly workers based on real-time cognitive load monitoring, characterized in that, The package includes the following steps: (1) Real-time collection of workers' multimodal physiological signals through AR glasses, including EEG signals, eye movement trajectory, heart rate variability, electromyography signals, speech data, hand movements and skin conductance response; (2) The processing unit performs real-time analysis on the physiological signals and calculates the cognitive load index and emotional state index; (3) When the cognitive load index exceeds the first threshold, the display method of AR training content is automatically adjusted, including reducing information density, increasing font size, or switching the presentation mode; (4) When the cognitive load index continues to exceed the second threshold, a forced pause mechanism is triggered, the workstation is locked and an alarm is issued; (5) The collected data is processed locally and then anonymously uploaded to the central system for use in optimizing training strategies.

2. The adaptive AR training method and system for aerospace assembly workers based on real-time cognitive load monitoring as described in claim 1, characterized in that: The adjustments to the AR training content include simplifying complex 3D assembly diagrams into 2D key step views and highlighting the optimal operation path.

3. The adaptive AR training method and system for aerospace assembly workers based on real-time cognitive load monitoring according to claim 1, characterized in that: The emotional state index is assessed using voice tone analysis and used to generate personalized voice prompts, with prompt styles including preset encouraging voices or operation instructions.

4. The adaptive AR training method and system for aerospace assembly workers based on real-time cognitive load monitoring as described in claim 1, characterized in that: The forced pause mechanism uses a hardware relay to lock the workstation and sends the data to the central monitoring system for recording.

5. The adaptive AR training method and system for aerospace assembly workers based on real-time cognitive load monitoring according to claim 1, characterized in that: The data acquisition and processing are all completed locally on the AR glasses, and the original physiological signals do not leave the country. If an illegal transmission attempt is detected, the hardware self-destruct mechanism is activated.

6. The adaptive AR training method and system for aerospace assembly workers based on real-time cognitive load monitoring according to claim 1, characterized in that, include: The sensor module, integrated into the AR glasses, is used to collect physiological signals; The processing module is used to calculate the cognitive load index in real time; AR display module, used for dynamically rendering training content; The intervention module is used to execute a forced pause; Communication module, used for secure data transmission.

7. The adaptive AR training method and system for aerospace assembly workers based on real-time cognitive load monitoring according to claim 6, characterized in that: The processing module uses a rule base for data analysis, and the rule base is updated regularly based on historical operation data.

8. The adaptive AR training method and system for aerospace assembly workers based on real-time cognitive load monitoring according to claim 6, characterized in that: The system supports special scenarios in aerospace assembly, including zero-gravity environments, where the cognitive load threshold is automatically adjusted.

9. The adaptive AR training method and system for aerospace assembly workers based on real-time cognitive load monitoring according to claim 6, characterized in that: The system is linked to worker performance management, and cognitive load event records are used for performance evaluation and training improvement.

10. The adaptive AR training method and system for aerospace assembly workers based on real-time cognitive load monitoring according to claim 6, characterized in that: The system has offline operation capabilities, and all key functions do not rely on a network connection.

Images