"Exploring data literacy and human-AI collaboration skills in Predictive Maintenance Training."

International Journal of Modern Research in Engineering and Technology (IJMRET) www.ijmret.org Volume 10 Issue 10 ǁ October 2025. w w w . i j m r e t . o r g I S S N : 2 4 5 6 - 5 6 2 8 Page 71 “Exploring data literacy and human-AI collaboration skills in Predictive Maintenance Training.” Manyanga David Victor Vanessa1, Wu Honglan2 1,2(Dept. of Civil Aviation Nanjing University of Aeronautics and Astronautics, Nanjing 210016, China) ABSTRACT: The aviation industry’s shift toward Predictive Maintenance (PdM), powered by Artificial Intelligence (AI) and Big Data, is transforming the role of maintenance technicians from manual troubleshooters to data-driven decision-makers. While PdM systems can forecast component degradation and optimize maintenance schedules, their success ultimately depends on the technician’s ability to interpret and act upon probabilistic information. Current research has concentrated on the technical development of PdM algorithms and traditional human error, leaving a significant gap in understanding the cognitive and collaborative skills required in this emerging digital environment. This paper explores the competencies of data literacy and human AI collaboration as critical enablers for effective PdM implementation. It argues that existing Maintenance Resource Management (MRM) training does not adequately prepare technicians for decision-making under uncertainty, thereby increasing the risks of automation bias and data misinterpretation. Using a conceptual qualitative synthesis, the study develops a structured competency framework that integrates two key domains: (1) data literacy emphasizing the evaluation of data quality and probabilistic Remaining Useful Life (RUL) outputs and (2) human AI collaboration focusing on calibrated trust, interpretability, and feedback mechanisms in AI-assisted maintenance environments. By shifting the emphasis from manual proficiency to cognitive readiness, this framework supports a safer, human centered integration of AI into aircraft maintenance practice and establishes the foundation for future curriculum design and regulatory guidance in predictive maintenance training. KEYWORDS - Predictive Maintenance (PdM); Data Literacy; Human–AI Collaboration; Aviation Maintenance Training; Cognitive Readiness; Automation Bias. I. INTRODUCTION 1.The Paradigm Shift: From Reactive to Predictive Maintenance High-reliability industries particularly commercial aviation are undergoing a profound transformation in maintenance philosophy, driven by the convergence of the Internet of Things (IoT), bigdata analytics, and artificial intelligence (AI) (ICAO, 2023). [1]Traditional approaches relied primarily on reactive maintenance (repairing components after failure) or time-based preventive maintenance (servicing components at fixed intervals). International Journal of Modern Research in Engineering and Technology (IJMRET) www.ijmret.org Volume 10 Issue 10 ǁ October 2025. w w w . i j m r e t . o r g I S S N : 2 4 5 6 - 5 6 2 8 Page 72 Modern aircraft now generate vast quantities of operational data, enabling a shift toward predictive maintenance (PdM) and prognostics and health management (PHM) (Al-Jumaili et al., 2012). [2] These data-driven methods use machine-learning algorithms to estimate the remaining useful life (RUL) of critical components, allowing maintenance to be scheduled precisely when needed reducing costs, minimizing unscheduled downtime, and improving aircraft availability (Patibandla, 2023).[3] 1. Problem Statement: The Human–AI Interpretation Gap While AI-based PdM systems have demonstrated strong predictive capability, their success ultimately depends on the human interpreter: the maintenance technician. The technician’s role is evolving from that of a hands-on fault-remedier to a data-centric decisionmaker (Jasper, 2023). [6]This cognitive shift introduces a new class of human-factors hazards that existing Maintenance Resource Management (MRM) training programs fail to address (Reason & Hobbs, 2003).[4] In the PdM environment, failures are no longer purely mechanical they increasingly result from cognitive errors and digital-literacy gaps. Two recurring issues illustrate this challenge: Data Misinterpretation: The inability to evaluate the quality, context, and probabilistic nature of an RUL prediction can lead to either missed safety events or unnecessary component replacements (TU Delft Repository, 2022).[5] Uncalibrated Reliance: Technicians may exhibit automation bias over-trusting AI outputs and neglecting verification or, conversely, algorithmic distrust, rejecting valid predictions (ICAO, 2023).[1] Consequently, the enormous investment in PdM infrastructure is constrained by a shortage of technicians trained to make sound, data-informed judgments. 2.Aims of the Paper To address these emerging safety and operational challenges, this paper pursues two primary aims: To define core competencies: conceptually identify and categorize the data-literacy and human–AI collaboration skills essential for maintenance professionals operating in PdM environments. To propose a training framework: outline a structured model for integrating these competencies into existing maintenance-training curricula (e.g., EASA Part 147 or FAA-approved programs) to support effective, safe, and efficient human AI teamwork. 3. Research Gap and Contribution Most existing research on PdM in aviation emphasizes the technical performance of AI and machine-learning models (Patibandla, 2023)[3] or explores regulatory and ethical aspects of automation (Henneberry et al., 2023)[14]. Limited attention has been paid to the human competencies required to interpret and apply predictive outputs in real maintenance contexts. Recent reviews also show that current aviationtraining curricula underemphasize AI, data analytics, and machine-learning concepts, creating a misalignment between workforce preparation and industry needs (Transport and Telecommunication Institute, 2023). This study contributes by moving beyond technology to address the human dimension of PdM adoption. It proposes a structured, data-centric competency model grounded in human-factors and cognitive-science theory that links the demands of prognostic data to measurable technician skills. The resulting framework provides aviationmaintenance organizations and training institutions with a foundation for developing the digitally proficient maintenance workforce required for the next generation of predictive maintenance systems. International Journal of Modern Research in Engineering and Technology (IJMRET) www.ijmret.org Volume 10 Issue 10 ǁ October 2025. w w w . i j m r e t . o r g I S S N : 2 4 5 6 - 5 6 2 8 Page 73 II. CONCEPTUAL FRAMEWORK DEVELOPMENT Core Competency Framework for Predictive Maintenance (PdM) Technicians 1. Overview The transition to Predictive Maintenance (PdM) requires more than the adoption of new technologies it demands a fundamental transformation of technician competencies. Traditional maintenance training emphasizes procedural compliance and mechanical skill; however, PdM environments require technicians to engage in data interpretation, probabilistic reasoning, and collaborative decisionmaking with AI systems. To address this shift, a three-pillar PdM Competency Model is proposed. The framework defines the cognitive and behavioral skills necessary for technicians to operate effectively in data-driven maintenance environments. The three core competency domains are: Data Literacy – interpreting, validating, and applying probabilistic data. Human–AI Collaboration (HAIC) and Trust – calibrating reliance on AI systems and engaging in explainable decision-making. Organizational and Ethical Awareness – understanding accountability, feedback, and safety implications in AI-supported operations. Each domain integrates targeted training interventions grounded in established human-factors and cognitive-learning principles. 2.Core Competency Framework for Predictive Maintenance (PdM) Technicians Building on the aims outlined in Section 1, this paper develops a conceptual framework that identifies and organizes the essential human competencies required for effective Predictive Maintenance (PdM) in aviation. While existing literature has primarily emphasized algorithmic accuracy and system reliability, there is limited attention to the human cognitive and organizational capabilities that determine the success of PdM implementation. To address this gap, the following framework defines three integrated domains of technician competence: Data Literacy, Human–AI Collaboration and Trust, and Organizational and Ethical Awareness. Each domain is supported by a structured set of skills, cognitive challenges, and training interventions designed to enhance safety, interpretability, and human oversight in AI-driven maintenance environments. A: Data Literacy Definition: Data Literacy in the context of Predictive Maintenance (PdM) refers to the technician’s ability to understand, interpret, and critically evaluate maintenance data generated by AI systems, particularly probabilistic indicators such as Remaining Useful Life (RUL). It extends beyond technical data handling it represents a cognitive competency that integrates analytical reasoning, contextual understanding, and safety-centered judgment. In traditional maintenance, technicians relied on deterministic indicators such as fixed inspection intervals or binary fault codes (“OK”/“Not OK”). PdM, however, introduces probabilistic information a component may have a 70% chance of failure within 100 flight hours, or a predicted RUL of 50 hours ±10%. These outputs demand interpretive reasoning, where the technician must assess data reliability, consider operational context, and decide whether to act immediately or continue operation safely. Developing data literacy is therefore fundamental to avoiding two new hazards in predictive environments: False confidence in misleading data (e.g., acting on noise or faulty sensors), and Complacency toward uncertain predictions (e.g., ignoring early failure warnings). International Journal of Modern Research in Engineering and Technology (IJMRET) www.ijmret.org Volume 10 Issue 10 ǁ October 2025. w w w . i j m r e t . o r g I S S N : 2 4 5 6 - 5 6 2 8 Page 74 A data-literate technician becomes a critical thinker rather than a passive receiver of AI information, ensuring maintenance decisions remain grounded in both analytical evidence and operational context. Table I. Data Literacy Competency Framework Core Competency Domain A: Data Literacy skills, cognitive challenges, and training interventions for effective interpretation of predictive maintenance data. Through these activities, data literacy evolves from a purely technical skill to a cognitive competency, supporting safer and more cost-effective maintenance planning. Data Quality and Context Awareness In PdM, the accuracy of predictions depends on the quality of sensor inputs and contextual factors (e.g., temperature, pressure, aircraft utilization rate). Poor data quality can mislead even the most advanced algorithms. A technician must therefore develop an analytical habit of cross-verification checking sensor health, comparing multiple parameters, and understanding the operational history of each component. This competency relates to the situation awareness model proposed by Endsley (1995)[15], where technicians must perceive data accurately, comprehend its meaning in context, and project its implications for maintenance outcomes. A lack of contextual evaluation can cause “automationinduced complacency,” where technicians accept AI outputs without question, undermining safety. Uncertainty Interpretation and Probabilistic Thinking Unlike deterministic maintenance systems, PdM outputs rarely provide absolute answers. Instead, technicians receive confidence intervals or probabilistic predictions. For instance, an AI tool may suggest that “the hydraulic pump is 80% likely to fail within 60 flight hours.” Understanding this information requires technicians to develop statistical literacy the ability to interpret probability as a decision-support tool rather than a prediction of certainty. This demands higher-order cognitive processing, as technicians must integrate uncertainty with operational judgment, risk tolerance, and safety margins. Inadequate probabilistic understanding can lead to automation bias, where technicians defer decisions to AI without assessing underlying reliability, or to algorithmic distrust, where valid warnings are ignored. Cognitive Training Outcomes Through structured, simulation-based interventions, data literacy training transforms technicians from procedural operators into analytical evaluators. It equips them to: Critically assess data reliability and AI output consistency. International Journal of Modern Research in Engineering and Technology (IJMRET) www.ijmret.org Volume 10 Issue 10 ǁ October 2025. w w w . i j m r e t . o r g I S S N : 2 4 5 6 - 5 6 2 8 Page 75 Balance risk decisions under uncertainty with confidence. Communicate data-driven reasoning clearly during team decision-making. Ultimately, data literacy evolves from being a technical proficiency (e.g., “reading system outputs”) to a core cognitive competency one that underpins every decision in AI-assisted maintenance environments. This transformation supports both safety assurance and cost efficiency, ensuring technicians remain the intelligent interpreters of predictive systems, not their passive executors. 2.Core Competency Domain B: Human–AI Collaboration (HAIC) and Trust Definition: Human–AI Collaboration (HAIC) in Predictive Maintenance (PdM) refers to a technician’s ability to work effectively and safely alongside AI systems, treating them as analytical partners rather than infallible authorities. The goal is to maintain calibrated trust a balance between reliance and skepticism so that technicians neither over-trust AI outputs (automation bias) nor dismiss accurate predictions due to distrust or misunderstanding (algorithmic distrust). In PdM contexts, AI algorithms generate probabilistic predictions about equipment health. For example, an AI system may forecast that a bleed air valve has an 85% chance of failure within 40 flight hours. A technician with calibrated trust uses this information to inform, not replace, their judgment cross-checking system parameters, consulting maintenance logs, and validating sensor consistency before acting. Developing effective HAIC skills ensures that AI serves as a decision-support tool, not a decisionmaker, preserving human accountability and situational awareness. Table II. Human–AI Collaboration (HAIC) Competency Framework Core Competency Domain B: Human–AI Collaboration and Trust skills, challenges, and training interventions supporting calibrated trust and explainable-AI utilization. Developing these skills ensures technicians become informed supervisors of AI, capable of interpreting model reasoning rather than deferring blindly to its outputs. Understanding Calibrated Trust The relationship between humans and intelligent systems has long been described through the “appropriate reliance” framework proposed by Lee and See (2004), which emphasizes that optimal performance occurs when human trust aligns with system reliability. In PdM, this means technicians must learn to adjust their level of reliance International Journal of Modern Research in Engineering and Technology (IJMRET) www.ijmret.org Volume 10 Issue 10 ǁ October 2025. w w w . i j m r e t . o r g I S S N : 2 4 5 6 - 5 6 2 8 Page 76 dynamically trusting AI outputs when data is consistent and questioning them when anomalies appear. Uncalibrated trust can manifest as: Automation Bias: unquestioning acceptance of AI outputs, leading to overlooked faults or deferred maintenance actions. Algorithmic Distrust: premature rejection of valid AI alerts, often due to lack of understanding of how predictions are generated. Trust calibration training develops metacognition awareness of one’s thought process enabling technicians to monitor and adjust their confidence in AI over time. By practicing in simulated environments with intentionally varying AI reliability, trainees build the habit of continuous evaluation rather than blind acceptance. Explainable AI (XAI) and Cognitive Transparency Explainable AI (XAI) is critical for fostering human interpretability and accountability in PdM. Traditional AI systems operate as “black boxes,” offering predictions without showing how those conclusions were reached. This opacity undermines trust and limits human learning. By exposing internal reasoning such as showing that a specific temperature anomaly contributed 45% to a predicted failure XAI allows technicians to understand why the AI reached its decision. This fosters cognitive transparency, aligning algorithmic reasoning with physical intuition. Training technicians to use XAI interfaces also develops cross-domain literacy: they learn to bridge data-science concepts (e.g., feature weighting) with engineering knowledge (e.g., thermodynamic relationships). Over time, this interdisciplinary awareness leads to better fault interpretation, faster root-cause identification, and improved maintenance efficiency. Cognitive and Behavioral Transformation Human–AI collaboration training moves technicians from being system users to system supervisors. After targeted HAIC training, technicians: Learn to challenge AI predictions constructively, not emotionally. Understand confidence thresholds and model limitations. Maintain accountability by documenting the rationale for AI acceptance or override. Communicate AI findings effectively within maintenance teams and across departments (engineering, data analytics, operations). These behaviors foster a culture of shared situational awareness, where human insight and machine intelligence complement each other rather than compete. Ultimately, HAIC training creates resilient decisionmakers who can adapt to both system uncertainty and AI evolution ensuring safety remains humancentered in increasingly automated environments. 3.Core Competency Domain C: Organizational and Ethical Awareness Definition: Organizational and Ethical Awareness in Predictive Maintenance (PdM) refers to the technician’s understanding of their role within a larger socio-technical system how individual actions, decisions, and data inputs influence organizational safety, accountability, and learning. In AI-assisted maintenance environments, technicians are no longer just executors of scheduled tasks; they are data contributors, interpreters, and decision influencers whose choices directly impact predictive model reliability and overall airworthiness. This domain emphasizes two interconnected dimensions: Organizational Accountability – recognizing one’s responsibility within safety management structures and communication channels. International Journal of Modern Research in Engineering and Technology (IJMRET) www.ijmret.org Volume 10 Issue 10 ǁ October 2025. w w w . i j m r e t . o r g I S S N : 2 4 5 6 - 5 6 2 8 Page 77 Ethical Responsibility – understanding the moral and professional duty to ensure AI outputs are used safely, transparently, and fairly. Developing awareness in these areas ensures that predictive maintenance does not become a technically efficient but ethically fragile system, but rather a resilient, learning-oriented ecosystem grounded in human judgment and organizational trust Table III. Proposed Three-Stage Validation Framework equential stages for validating the PdM Competency Model, detailing objectives, methods, and expected outcomes across expert evaluation, simulation, and curriculum-integration pilots. These competencies reinforce the human’s role as the final authority on airworthiness decisions while fostering a culture of continuous organizational learning. Accountability and Ethical Decision-Making In traditional maintenance, accountability was clearcut: technicians followed procedures, and errors were traced to human performance. In PdM, accountability becomes more complex decisions are shared between human and AI systems. A technician may decide to act or not act based on an AI-generated probability, blurring the boundary between mechanical reliability and cognitive judgment. This raises key ethical questions: Who is responsible when an AI recommendation leads to a maintenance oversight? How should technicians balance organizational pressure for operational efficiency against the duty to prioritize safety? Training must therefore emphasize ethical discernment the ability to make and justify decisions that align with both technical evidence and moral responsibility. The “Just Culture” framework (Reason, 1997) provides the ideal foundation, promoting accountability without punishment and encouraging honest reporting of AI-related errors. When technicians feel psychologically safe to report near-misses or data misinterpretations, organizations gain invaluable learning opportunities that strengthen the predictive system. Feedback Loop Engagement and Continuous Learning Predictive maintenance systems are inherently datadependent their predictive accuracy relies on feedback from real-world outcomes. If a technician replaces a component early but fails to record the actual failure status, the AI model cannot learn whether its prediction was correct. Over time, this lack of feedback erodes model reliability. Technicians must therefore see themselves not only as maintenance executors but also as co-creators of data integrity. By systematically entering maintenance outcomes, they participate in a continuous improvement loop that enhances both the AI system and the organization’s operational intelligence. This feedback engagement aligns with the principles of Safety Management Systems (SMS) and Organizational Learning Theory, where each International Journal of Modern Research in Engineering and Technology (IJMRET) www.ijmret.org Volume 10 Issue 10 ǁ October 2025. w w w . i j m r e t . o r g I S S N : 2 4 5 6 - 5 6 2 8 Page 78 maintenance action contributes to collective knowledge. Proper training reinforces that accurate data entry and reflection are not administrative burdens but safety-critical tasks that uphold the long-term effectiveness of PdM programs. Organizational Culture and Ethical Climate The development of organizational and ethical awareness also depends on a supportive culture. Organizations that prioritize transparency, fairness, and communication are more likely to succeed in implementing PdM safely. When leaders frame AI not as a tool for surveillance but as an enabler of learning, technicians feel empowered to question predictions and share observations without fear of blame. Training programs should therefore include organizational-level interventions such as: Leadership workshops on managing AI accountability. Cross-departmental discussions between data scientists, engineers, and technicians. Reflective exercises where teams analyze how their collective decisions influence safety metrics. By cultivating an ethical climate rooted in trust, transparency, and learning, organizations can ensure that AI-driven maintenance complements not replaces human judgment and moral responsibility. Cognitive and Behavioral Outcomes Training in this domain transforms technicians into ethically aware system participants who: Take ownership of their maintenance decisions and document justifications clearly. Actively contribute feedback to AI systems, improving predictive accuracy. Engage in transparent discussions about AI performance and model reliability. Demonstrate ethical integrity by prioritizing safety even under operational pressure. Such outcomes contribute to a culture of informed accountability, where human and AI collaboration is reinforced by ethical consistency and organizational trust. Organizational and Ethical Awareness completes the PdM competency model by linking technical decision-making with organizational learning and moral responsibility. It ensures that as AI transforms aviation maintenance, humans remain the ethical anchors and quality guardians of predictive systems. III. CURRICULUM INTEGRATION AND EDUCATIONAL IMPLICATIONS 1. Integration and Curriculum Implications The proposed Predictive Maintenance Competency Model can be effectively embedded into existing EASA Part 147 and FAA-approved Maintenance Training Organization (MTO) curricula by expanding beyond procedural skill instruction to include cognitive, analytical, and ethical competencies. Integration can occur through the following three instructional modifications: Digital-Simulation Environments The use of high-fidelity digital twin environments and virtual maintenance simulators provides a controlled yet realistic platform for experiential learning. These systems replicate actual aircraft systems and PdM dashboards, allowing trainees to interact with synthetic or historical maintenance data safely and repeatedly. Through simulated fault conditions and AIgenerated Remaining Useful Life (RUL) predictions, technicians can practice data validation, uncertainty interpretation, and AI collaboration without operational risk. This approach supports experiential cognition, enabling learners to visualize system behaviors, understand probabilistic patterns, and develop decision confidence in predictive contexts. International Journal of Modern Research in Engineering and Technology (IJMRET) www.ijmret.org Volume 10 Issue 10 ǁ October 2025. w w w . i j m r e t . o r g I S S N : 2 4 5 6 - 5 6 2 8 Page 79 According to Boeing (2023)[12], digital twin training environments not only improve knowledge retention but also shorten skill acquisition time by enabling immediate feedback and iterative learning. Incorporating digital simulation into Part 147 programs thus redefines the “hands-on” component from purely mechanical manipulation to data-driven scenario management, aligning maintenance education with the realities of intelligent maintenance systems. Scenario-Based and Experiential Learning Traditional maintenance training often emphasizes procedural compliance—performing checklists and routine tasks by rote. In a PdM context, however, technicians must navigate uncertainty and probabilistic risk, requiring flexible and adaptive thinking. Scenario-Based Training (SBT) immerses learners in dynamic, data-rich situations where no single correct answer exists. For example, trainees may face conflicting AI outputs and physical observations, forcing them to balance risk, reliability, and safety priorities. This form of cognitive simulation encourages decision-making under ambiguity, reflective reasoning, and the development of critical thinking under pressure. According to TU Delft Repository (2022)[5], scenario-based learning promotes transfer of knowledge to real operational settings by engaging higher-order cognitive processes rather than procedural memory. Embedding SBT in aviation maintenance curricula ensures that trainees can interpret AI information, justify decisions transparently, and respond confidently to uncertain or conflicting system feedback. Interdisciplinary Collaboration Modules Predictive Maintenance is inherently crossdisciplinary requiring interaction among maintenance technicians, data scientists, reliability engineers, and AI specialists. Current maintenance programs rarely address this collaborative interface, leading to communication barriers that hinder the practical use of PdM systems. Structured interdisciplinary collaboration modules can bridge this gap by fostering mutual understanding between human and algorithmic reasoning. For example, trainees could participate in joint workshops where data scientists explain how AI models interpret sensor data, while technicians provide operational insights on physical system behavior. These exchanges cultivate shared mental models, ensuring both groups interpret PdM outputs consistently and communicate effectively across technical boundaries. ICAO (2023)[16] highlights that such collaboration improves not only technical integration but also organizational resilience, as cross-functional literacy enables teams to identify potential system errors earlier and implement corrective strategies proactively. Educational Impact Collectively, these curricular adaptations reposition aviation maintenance training from a complianceoriented paradigm toward a cognition-oriented paradigm. Trainees transition from “following procedures” to analyzing systems, from “reacting to faults” to anticipating failures, and from “executing commands” to collaborating with intelligent systems. This reorientation aligns with the broader aviation goal of developing human-AI synergy, where cognitive readiness, ethical accountability, and technical skill operate in tandem to sustain safety and efficiency. Summary of the Model’s Value International Journal of Modern Research in Engineering and Technology (IJMRET) www.ijmret.org Volume 10 Issue 10 ǁ October 2025. w w w . i j m r e t . o r g I S S N : 2 4 5 6 - 5 6 2 8 Page 86 Predictive Maintenance,” Delft University of Technology, Delft, The Netherlands, 2022. [6] T. Jasper, “Human Factors and Data-Centric Decision Making in Predictive Maintenance,” International Journal of Aviation Technology and Management, vol. 15, no. 2, pp. 85–97, 2023. [7] J. D. Lee and K. A. See, “Trust in Automation: Designing for Appropriate Reliance,” Human Factors, vol. 46, no. 1, pp. 50–80, 2004. [8] J. Reason, Managing the Risks of Organizational Accidents, Aldershot, U.K.: Ashgate Publishing, 1997. [9] J. Parasuraman and V. Riley, “Humans and Automation: Use, Misuse, Disuse, Abuse,” Human Factors, vol. 39, no. 2, pp. 230–253, 1997. [10] S. Gregor and A. Hevner, “Positioning and Presenting Design Science Research for Maximum Impact,” MIS Quarterly, vol. 37, no. 2, pp. 337–355, 2013. [11] V. Braun and V. Clarke, “Using Thematic Analysis in Psychology,” Qualitative Research in Psychology, vol. 3, no. 2, pp. 77–101, 2006. [12] Boeing Global Services, Maintenance Synthetic Trainer Documentation, Seattle, WA, USA: Boeing, 2023. [13] Transport and Telecommunication Institute, “Integration of Artificial Intelligence Competencies in Aviation Education,” Riga, Latvia, 2023. [14] A. Henneberry, D. Zhao, and M. Lin, “Regulatory Oversight of Artificial Intelligence in Predictive Maintenance Systems,” Aerospace Policy Review, vol. 19, no. 3, pp. 211–225, 2023. [15] M. R. Endsley, “Toward a Theory of Situation Awareness in Dynamic Systems,” Human Factors, vol. 37, no. 1, pp. 32–64, 1995, doi: 10.1518/001872095779049543. [16] International Civil Aviation Organization, Next Generation of Aviation Professionals (NGAP) Strategy, Montréal, QC, Canada, 2021 (rev. 2023). Available: www.icao.in