Decision-Making Framework for Aviation Safety in Predictive Maintenance Strategies

Abstract

The implementation of predictive maintenance (PM) in aviation presents unique challenges due to strict safety requirements, complex operational environments, and regulatory constraints. This paper develops a comprehensive decision-making framework for evaluating the feasibility of implementing PM for aircraft components, addressing the critical need for systematic integration of technical, economic, and regulatory considerations. Through expert surveys involving 78 aviation maintenance professionals and the application of multi-criteria decision analysis, this study identifies and validates 14 key criteria across four categories: technical and operational, economic and feasibility, regulatory and compliance, and organizational and human factors. The analytic hierarchy process is employed to establish criteria weights, with flight safety impact, reliability predictability, and data sufficiency emerging as primary drivers. The framework’s effectiveness is demonstrated through case studies comparing turbofan engines and avionics units, validating its ability to discriminate between components suitable for PM implementation. Results indicate that successful PM implementation requires not only technological readiness but also organizational alignment and regulatory compliance. This study contributes to aviation maintenance practice by providing a structured, evidence-based approach to PM implementation decisions, while establishing a foundation for future innovations in maintenance strategies. The framework’s practical applicability is enhanced through a detailed implementation roadmap and validation methods, ensuring its relevance for maintenance decision-makers while maintaining alignment with aviation safety standards.

1. Introduction

1.1. Background and Motivation

The aviation industry is one of the most complex and safety-critical sectors, where effective maintenance strategies are essential to ensure reliability, safety, and operational efficiency [1]. Predictive maintenance (PM), a data-driven approach that uses real-time monitoring and advanced analytics to anticipate component failures before they occur, has emerged as a promising alternative to traditional maintenance strategies [2]. By enabling timely interventions, PM can reduce unplanned downtime, optimize maintenance schedules, and enhance overall safety. However, the implementation of PM in aviation presents significant challenges due to the critical nature of components, the variability in failure modes, and the stringent regulatory requirements that govern the industry.

While PM has demonstrated success in other industries, such as manufacturing and automotive, its application in aviation requires a more nuanced approach. Aircraft components operate under highly dynamic and harsh conditions, and their failure mechanisms vary widely—from predictable wear and tear in mechanical systems to sudden, unpredictable failures in electronic systems. Moreover, the aviation industry is characterized by strict safety and compliance standards, making it imperative to assess whether PM can meet these requirements without compromising operational integrity. The high cost of implementing PM technologies, such as sensor retrofitting, data infrastructure, and algorithm development, further underscores the need for a robust decision-making framework to evaluate its feasibility [3].

The primary objective of this research is to develop, validate, and refine a comprehensive decision-making framework for evaluating the feasibility of implementing PM for aircraft components. This framework aims to systematically integrate technical, operational, economic, and regulatory criteria to provide a holistic assessment of PM suitability. By using multi-criteria decision analysis (MCDA) techniques, this study seeks to prioritize factors such as flight safety impact, reliability predictability, economic feasibility, and data sufficiency while also addressing challenges like scalability, technological integration, and compliance with aviation standards. This research seeks to provide maintenance organizations, airlines, and regulatory bodies with a structured, data-driven approach to enhance decision-making, optimize maintenance strategies, and improve overall safety and efficiency in the aviation sector.

1.2. Related Works

The aviation industry encompasses a diverse array of economic activities, with one of its key segments being the maintenance, repair, and overhaul (MRO) market. This sector accounts for approximately 11–15% of an airline’s operating costs [4]. Given the significant impact of MRO on airline profitability, improving efficiency in aircraft maintenance processes has become a focal area. Alongside this, technological advancements in the field have matured to a level that supports the adoption of innovative maintenance strategies and policies, including condition-based maintenance (CBM).

CBM is a proactive maintenance policy defined as “preventive maintenance which includes assessment of physical conditions, analysis and the possible ensuing maintenance actions” [5]. Alternatively, it is described as “a maintenance program that recommends maintenance actions based on the information collected through condition monitoring” [6]. By using real-time condition monitoring, CBM helps identify and prevent unscheduled maintenance events, enables task substitution or extension of maintenance intervals, and optimizes maintenance schedules across an airline’s fleet [7,8]. These capabilities position CBM as a promising approach to significantly improve operational efficiency in aviation.

CBM is expected to emerge as the dominant maintenance policy in aviation. However, despite the presence of CBM elements in the industry for decades, widespread adoption of this maintenance approach remains elusive, falling short of the ambitious vision set forth by the Advisory Council for Aviation Research and Innovation in Europe (ACARE) [9].

Answering these questions requires a detailed and holistic understanding of CBM that considers all factors influencing its uptake. Historically, most research on CBM has been technologically oriented, focusing on the development of models, algorithms, and methods for the detection, diagnosis, and prognosis of equipment failures. Additionally, there is a substantial body of literature addressing planning and decision-support aspects of CBM, exploring how organizations can utilize diagnostic and predictive data to effectively plan and execute maintenance interventions [10].

Preventive maintenance is widely employed to schedule maintenance actions proactively to prevent equipment failures. While effective, this strategy often incurs higher costs and results in unnecessary component replacements or unforeseen failures [11]. PM offers a more refined alternative by detecting potential failures in advance through periodic monitoring of the condition of processes, machines, materials, and products within manufacturing systems. This approach enables just-in-time maintenance actions, enhancing equipment availability, quality, and safety while significantly reducing costs associated with breakdowns and excessive maintenance activities [12,13]. Predictive maintenance can be classified into two distinct subcategories: statistics-based PM and condition-based PM [14].

The integration of deep learning methodologies has enhanced the accuracy of predicting the remaining useful life (RUL) of aircraft engines. Paper [15] introduced innovative models, including convolutional LSTM (long short-term memory) and Recurrent LSTM, demonstrating that the latter achieved superior accuracy, precision, and recall rates in RUL prediction.

Paper [16] proposed a deep learning algorithm combining a 1-D convolutional neural network (CNN) with a bidirectional long short-term memory network, effectively predicting engine RUL.

The application of artificial intelligence (AI) and machine learning (ML) in PM has also been explored to improve maintenance strategies. A study [17] presented an approach using CNNs to develop a predictive failure model for early detection of engine breakdowns, aiming to transition from corrective to predictive maintenance in the aviation industry.

Another investigation compared different regression models, including random survival forest, to determine their performance in predicting aircraft engine failures, highlighting the potential of data-driven strategies in PM [18].

The concept of digital twins has been applied to enhance PM in aviation. A study demonstrated how RUL predictions, combined with digital twin models, can aid in designing stable supply chains and maintenance strategies, emphasizing the role of digital twins in predictive maintenance [19]

The article of [20] presents a comprehensive systematic literature review of predictive maintenance in the aircraft industry, emphasizing its importance in optimizing maintenance schedules and reducing downtime. The review examines various data types, applications, and projects within the industry, aiming to identify and outline future research challenges and opportunities.

Despite these advancements, challenges remain in implementing PM in aviation. Data quality and availability are significant obstacles, as noise in sensor data, missing data points, and inconsistent formats can impact ML model performance [21].

A comprehensive review [22] focuses on predictive maintenance within the defense aviation sector, emphasizing fixed-wing aircraft operations. The study employs bibliometric analysis and text clustering to identify key research areas, highlighting the integration of prognostics with decision support systems and identifying practical challenges in PM implementation.

The study [23] introduces a unified ecosystem for data sharing and AI-driven predictive maintenance. This framework integrates real-time and historical data from diverse sources, including aircraft sensors and maintenance logs, to enhance decision-making processes for airlines and maintenance providers.

The application of AI in PM is further explored in [24], which discusses key challenges and future directions for AI-driven predictive maintenance applications. The study emphasizes the importance of data quality, model interpretability, and integration with existing maintenance systems.

The transition from traditional aircraft health monitoring to advanced health management approaches is examined, highlighting the role of the Internet of Things (IoT) and AI in enabling predictive maintenance [25]. The study discusses how real-time data collection and analysis can facilitate proactive maintenance strategies, enhancing flight safety and operational efficiency.

The challenges associated with data-driven predictive aircraft maintenance are analyzed in [26], which identifies hazards related to the introduction of data-driven technologies into aircraft maintenance. The study underscores the need for improving the reliability of condition monitoring systems, ensuring timely communication between agents, and building stakeholders’ trust in new technologies.

The role of prognostics and health management (PHM) systems in aviation maintenance is discussed in [27], focusing on their diagnostic and prognostic capabilities. The study emphasizes the importance of PHM in maintaining aircraft integrity and reliability by forecasting and preventing malfunctions before they occur.

A novel decision-making framework for the lifecycle management of aircraft components is presented in [28], integrating advanced data analytics, AI, and predictive maintenance strategies. The study introduces a data-driven approach for optimizing maintenance scheduling and resource allocation, aiming to enhance reliability and cost-effectiveness.

Reference [29] explores the enhancement of predictive maintenance by analyzing the sensitivity of maintenance decisions to changes in model parameters such as costs and repair duration. An original approach is proposed that integrates maintenance costs and risks into a unified objective function to minimize, utilizing RUL as an indicator of system health.

To address the challenges of predictive maintenance, the study [30] proposes a novel, technology-independent model that enhances flexibility and interoperability in manufacturing environments. This approach enables interoperability among diverse devices and facilitates the implementation of generic predictive maintenance functionalities. This is particularly relevant in environments where vendor-specific solutions and varied technologies for condition monitoring are prevalent, often limiting the adaptability required by modern manufacturing demands for customization.

The paper presented in [31] addresses the challenges asset owners face in choosing the most suitable predictive maintenance method, proposing a framework to aid in selecting the optimal approach. The developed framework incorporates a classification of maintenance methods, guidelines for defining maintenance ambition levels, and a categorization of available data types.

The research of [32] investigates how 6G technology will facilitate improved real-time information exchange, anticipatory maintenance procedures, and streamlined communication within the aviation industry, presenting an innovative eight-tier monitoring system that leverages virtual replicas, distributed learning mechanisms, and localized computing power to fundamentally alter aircraft maintenance through constant, instantaneous surveillance and automated response protocols.

The investigation in [33] reveals significant adoption of augmented reality, additive manufacturing, and machine learning technologies in the aircraft maintenance procedures, although many remain in pre-production phases, highlighting ongoing development efforts and investment priorities. The research ultimately aims to facilitate the adoption of innovative maintenance practices that enhance safety, cost-effectiveness, and sustainability across the aviation industry.

The evaluation of existing monitoring techniques, careful selection of artificial intelligence predictive algorithms, and implementation of blockchain architecture for the progression from conventional monitoring to sophisticated predictive and prescriptive maintenance strategies are described in study [34].

1.3. Research Gap, Contributions and Paper Structure

Despite notable advancements in PM technologies, their application in the aviation industry faces significant challenges. Existing research has often concentrated on specific technological innovations, such as machine learning models, digital twins, or IoT-based systems, and planning methods to optimize maintenance schedules. However, these approaches frequently fail to address aviation’s unique and stringent safety requirements comprehensively. Current studies lack an integrated decision-making framework that systematically evaluates PM feasibility across critical safety, economic, regulatory, and operational dimensions. Furthermore, there is a critical gap in frameworks that incorporate stakeholder consensus and scalable methodologies validated for the aviation industry’s highly regulated environment.

This study addresses these gaps by developing a robust decision-making framework specifically tailored to aviation predictive maintenance, prioritizing safety as its core focus. By systematically integrating MCDA, the proposed framework evaluates aircraft components for PM suitability, emphasizing safety-critical factors such as flight risk mitigation, reliability predictability, and data sufficiency. It ensures compliance with regulatory standards while providing a structured mechanism for stakeholders to make informed, safety-focused decisions.

The primary contributions of this paper are as follows:

• A comprehensive decision-making framework that prioritizes aviation safety by incorporating key criteria such as flight safety impact, regulatory compliance, and reliability predictability, alongside economic and operational considerations.
• The application of the analytic hierarchy process (AHP) for criteria weighting, ensuring transparency and alignment with the aviation sector’s safety and regulatory imperatives.
• Validation of the framework through expert surveys, historical maintenance data, and real-world case studies, highlighting its ability to enhance decision-making and reduce safety risks.
• Practical recommendations for integrating PM strategies into existing aviation maintenance operations, ensuring both enhanced safety and cost efficiency while aligning with global aviation standards.

The structure of the paper is organized as follows: Section 2 details the materials and methods, focusing on the critical role of predictive maintenance in aviation and the conceptual framework for this study. Section 3 presents the results, including the selection of evaluation factors, the MCDA-based decision-making framework, and case studies showcasing its real-world application. Section 4 discusses the framework’s expected safety and operational benefits, challenges, and limitations, along with a detailed roadmap for implementing PM in aviation. Section 5 concludes the study by summarizing its contributions to aviation safety and efficiency while outlining opportunities for future research.

2. Materials and Methods

2.1. Predictive Maintenance in Aviation

The aviation industry is a highly complex and safety-critical domain where maintenance plays a pivotal role in ensuring operational efficiency and passenger safety. Traditional maintenance practices such as reactive and preventive maintenance have been effective but often result in unplanned downtimes or excessive maintenance costs. Predictive maintenance, on the other hand, uses advanced technologies to monitor the condition of components and systems, enabling timely interventions before failures occur. This approach not only enhances safety but also optimizes resource utilization, making it increasingly vital in modern aviation.

PM is a CBM strategy that uses real-time data from sensors and advanced analytics to predict when a component or system will likely fail. By identifying potential failures early, maintenance can be scheduled during non-critical times, reducing unexpected downtimes and associated costs. In aviation, this involves the integration of tools like vibration analysis, thermography, and acoustic emission testing with advanced analytics platforms to assess the health of aircraft components.

The potential–failure (P-F) curve is a fundamental tool used to understand the degradation process of components and systems [35]. It illustrates the timeline from the detection of a potential failure in point P to the point F of functional failure (Figure 1).

In aviation, where safety is paramount, detecting point P as early as possible is crucial. The P-F Curve is divided into several zones:

• Proactive maintenance zone occurs before any detectable degradation, focusing on eliminating root causes through design improvements and best practices.
• Preventive maintenance zone refers to scheduled interventions based on estimated component lifespans, often occurring near point P.
• Predictive maintenance zone is the critical window between point P and point F, where condition-monitoring tools detect issues and trigger timely interventions.
• Reactive maintenance zone refers to post-failure interventions when the system has already ceased to function.

In aviation, predictive maintenance thrives in the P-F interval, where it utilizes advanced monitoring and predictive analytics to address issues before they lead to functional failure. Modern aviation employs different predictive maintenance technologies (

Table 1. Tools for predictive monitoring.

Technique	Applications	Tools
Vibration Analysis	Rotating equipment, motors	Accelerometers, vibration sensors, data collectors
Infrared Thermography	Electrical panels, heat exchangers	Infrared cameras, thermal imaging systems
Ultrasound Testing	Compressed air systems, valves	Ultrasonic detectors, acoustic cameras
Oil Analysis	Hydraulic systems, engines	Spectrometers, particle counters
Machine Learning Algorithms	Predictive analytics	IoT sensors, AI-driven maintenance platforms
SCADA Systems	Centralized monitoring	Distributed sensors with real-time data dashboards

).

The applicability of the curve for defining the maintenance strategy largely depends on whether the component experiences gradual degradation that can be monitored, such as mechanical bearings or hydraulic systems. For components with measurable conditions, like engine vibrations or structural fatigue, the curve is highly relevant. Conversely, components that fail unpredictably, such as electronic chips or systems prone to sudden catastrophic failure, are less suited to this model. Some examples of applied predictive maintenance across various components and systems in aviation are shown in

Table 2. Aircraft components and the P-F curve.

Component	Applicability of P-F Curve	Reason
Engines (Turbines)	High	Condition monitoring (e.g., vibration, thermography) detects degradation over time.
Landing Gear	High	Wear and fatigue can be monitored with condition-based tools (e.g., crack detection).
Hydraulic Systems	Medium	Degradation (e.g., fluid contamination) is gradual and measurable, but sudden failures can occur.
Avionics (Electronics)	Low	Failures are often sudden and not preceded by detectable degradation.
Structural Components	High (for fatigue-related failures)	Cracks and corrosion follow predictable degradation patterns, detectable through inspections.
Fuel Systems	Medium	Degradation (e.g., clogging, leaks) can sometimes be gradual, but unexpected events also occur.

.

Predictive maintenance offers several advantages in aviation. It enhances safety by detecting potential failures early, reducing the risk of in-flight incidents. It optimizes costs by minimizing over-maintenance and avoiding costly reactive repairs. Increased aircraft availability is achieved by enabling scheduled maintenance during off-peak hours, reducing downtime. Data-driven decision-making improves the accuracy of maintenance planning through real-time data and predictive analytics.

Despite its advantages, implementing predictive maintenance in aviation is not without challenges. Integrating data from multiple systems and ensuring interoperability can be complex. Maintenance personnel require training to interpret data and operate advanced tools. Additionally, predictive maintenance practices must adhere to stringent aviation safety regulations.

2.2. Conceptual Framework for Study

The conceptual framework for this study is designed to provide a structured approach to evaluating the feasibility of PM implementation for aircraft components. This framework is built around three core components: identification of relevant criteria, development of a scoring and weighting system, and application of an aggregated decision-making model to assess PM suitability (Figure 2).

At its foundation, the framework identifies the key criteria influencing PM implementation. These factors are grouped into thematic categories to facilitate a structured analysis. Each criterion is quantified using measurable metrics, such as reliability indices for technical factors, cost estimates for economic considerations, and compliance scores for regulatory adherence.

The second component of the framework involves defining a weighting system to reflect the relative importance of each criterion. Weighting is determined using robust methods such as AHP. This ensures that the framework accounts for varying priorities, such as placing greater emphasis on safety-critical factors or aligning with organizational goals. A scoring system is also developed to evaluate each component’s alignment with the defined criteria on a standardized scale, ensuring consistency in assessment.

The final component of the framework is an aggregation model that integrates the weighted criteria and scores to calculate a predictive maintenance suitability score for each aircraft component. This score is derived using the MCDA approach. The resulting score allows for a clear and transparent decision on whether PM is suitable for a specific component.

This conceptual framework provides a rigorous methodology for integrating diverse factors into a unified decision-making process. By systematically evaluating components against well-defined criteria, the framework ensures a comprehensive and data-driven approach to predictive maintenance feasibility, ultimately supporting safer, more efficient, and cost-effective aviation maintenance practices.

3. Results

3.1. Methodology of Selecting Categories and Factors for PM in Aviation

The data collection process was designed to ensure comprehensive and unbiased insights from multiple sources. This study integrates primary data gathered through structured expert surveys and secondary data from historical maintenance records, industry reports, and regulatory guidelines. Expert surveys were conducted with 78 aviation maintenance professionals, ensuring a diverse representation of stakeholders from MRO providers, airline operators, and regulatory bodies. The survey design followed best practices in structured decision-making research, employing both closed and open-ended questions to balance quantitative scoring with qualitative insights. The identification and validation of key categories and factors for aviation predictive maintenance implementation emerged through a comprehensive analysis of expert survey responses.

The composition and expertise of survey participants were carefully analyzed to ensure comprehensive representation across the aviation maintenance sector. The expert role distribution analysis revealed a balanced representation of key stakeholders, with MRO professionals forming the largest group (Figure 3).

The experience distribution analysis demonstrated a strong foundation of expertise among participants, with the majority having extensive industry experience (Figure 4).

This diverse representation across roles and depth of experience enhanced the reliability and comprehensiveness of the survey findings, ensuring that the insights gathered encompassed both practical industry knowledge and regulatory perspectives.

Initially, eight thematic categories were considered: technical and operational, economic and feasibility, regulatory and compliance, organizational and human, data-driven, risk and resilience, environmental and sustainability, and innovation and technological development. Each category underwent rigorous evaluation based on three fundamental metrics: overall score, expert consensus percentage, and implementation feasibility. This multi-dimensional assessment approach ensured a balanced consideration of both theoretical importance and practical applicability.

Through statistical analysis and validation, four primary categories distinguished themselves with consistently superior performance across all evaluation metrics (Figure 5).

The top four categories show significantly higher scores across all metrics: the technical and operational category, the economic and feasibility category, regulatory and compliance category, the organizational and human factors category. The technical and operational category achieved the highest overall rating (9.2), with remarkable expert consensus (92%). The economic and feasibility category followed, with an overall score of 8.8 and 87% expert consensus. Regulatory and compliance category secured strong validation, with an 8.7 score and 86% consensus. The organizational and human factors category, while ranking fourth, maintained robust metrics (8.6 score, 84% consensus).

The selection methodology incorporated multiple validation techniques to ensure result reliability. Consensus analysis examined expert agreement levels, response distributions, and variance patterns. Implementation feasibility assessment evaluated practical challenges, resource requirements, and integration capabilities. Cross-category correlation analysis identified relationships, dependencies, and synergies between factors. Strict validation criteria required minimum thresholds of 8.0 for overall scores, 80% for expert consensus, and 80% for implementation feasibility.

Secondary categories, including data-driven (7.8), risk and resilience (7.6), environmental (7.4), and innovation factors (7.2), demonstrated relevance but fell below the established thresholds. This clear differentiation between primary and secondary categories reinforced the robustness of the selection methodology and the significance of the chosen factors.

The analysis revealed several crucial insights into aviation predictive maintenance implementation. Safety and technical considerations consistently dominated priority rankings across all expert groups, while economic factors showed strong correlation with implementation feasibility. Regulatory compliance emerged as a critical enabler for successful implementation, and organizational factors demonstrated significant impact on practical adoption potential.

This methodological approach successfully identified and validated categories and factors that not only demonstrated high importance scores but also showed strong expert consensus and practical implementation potential. The resulting framework provides a structured, comprehensive approach to evaluating predictive maintenance opportunities in aviation components, balancing theoretical importance with practical implementation considerations. The selected categories and their constituent factors represent a robust foundation for predictive maintenance decision-making, addressing both technical requirements and organizational dynamics in aviation maintenance implementation.

The comprehensive analysis of technical and operational factors in predictive maintenance reveals a clear hierarchy among ten evaluated criteria (Figure 6).

The assessment, based on survey responses from 78 aviation maintenance experts, measured each factor across three key metrics: overall score, expert consensus, and implementation feasibility. Flight safety impact emerged as the most critical factor, scoring highest across all metrics (9.2 overall score, 94% expert consensus, 91% implementation feasibility), followed by reliability predictability (8.8 overall score, 88% consensus, 86% implementation) and data sufficiency (8.7 overall score, 87% consensus, 84% implementation). The analysis shows a natural threshold around the 8.0 score mark, with four factors scoring above this level being selected as primary factors for the framework. The remaining six factors, while still relevant, showed lower scores and consensus levels, ranging from component criticality (8.0) to system complexity (7.4). This clear differentiation in scores and consensus levels provided a quantitative basis for factor selection, ensuring that the most critical and implementable factors were prioritized in the final framework.

The analysis of economic and feasibility factors across eight criteria reveals a clear prioritization pattern based on survey responses from aviation maintenance experts. Each factor was evaluated using three key metrics: overall score, expert consensus, and implementation feasibility (Figure 7).

Economic feasibility emerged as the most significant factor, achieving the highest ratings (8.9 overall score, 89% expert consensus, 87% implementation feasibility), followed by scalability (8.4 overall score, 84% consensus, 82% implementation) and end-of-life management (8.2 overall score, 82% consensus, 80% implementation). A natural threshold appears around the 8.2 score mark, with three factors scoring above this level being selected as primary factors for the framework. The remaining five factors, while demonstrating relevance, showed lower scores and consensus levels, ranging from return on investment (ROI) potential (8.1) to investment timeline (7.5). This quantitative differentiation provided a clear basis for factor selection, ensuring focus on the most impactful and implementable economic considerations in predictive maintenance.

The analysis of regulatory and compliance factors encompassed eight distinct criteria, evaluated through comprehensive survey responses from 78 aviation maintenance experts using three key metrics: overall score, expert consensus, and implementation feasibility (Figure 8).

Regulatory compliance emerged as the most critical factor with exceptionally high ratings (9.1 overall score, 92% expert consensus, 89% implementation feasibility), followed by data privacy and security (8.6 overall score, 86% consensus, 84% implementation) and environmental influence (8.3 overall score, 83% consensus, 81% implementation). A clear threshold was established at the 8.3 score mark, with three factors scoring above this level being selected as primary factors for the framework. The remaining five factors, while important for regulatory considerations, showed lower but still significant scores, ranging from safety regulations (8.2) to audit compliance (7.7). This quantitative differentiation provided a robust foundation for factor selection, ensuring focus on the most critical regulatory and compliance aspects in predictive maintenance implementation.

The analysis of organizational and human factors examined nine distinct criteria through survey responses from 78 aviation maintenance experts, evaluating each factor through three key metrics: overall score, expert consensus, and implementation feasibility (Figure 9).

Operational impact emerged as the highest-rated factor (8.8 overall score, 88% expert consensus, 86% implementation feasibility), followed closely by training requirements (8.7 overall score, 87% consensus, 85% implementation), technological integration (8.5 overall score, 85% consensus, 83% implementation), and stakeholder acceptance (8.4 overall score, 84% consensus, 82% implementation). A natural threshold was established at the 8.4 score mark, with four factors scoring above this level being selected as primary factors for the framework. The remaining five factors, while relevant to organizational considerations, showed lower scores, ranging from change management (8.0) to communication framework (7.6). This clear differentiation in scores provided a quantitative basis for factor selection, ensuring focus on the most impactful organizational and human aspects in predictive maintenance implementation.

While the data collection process was designed for robustness, several constraints were encountered. First, variations in data availability across different aviation maintenance organizations posed challenges in standardizing responses. Some experts had access to more comprehensive predictive maintenance datasets, while others relied on qualitative assessments. Additionally, integrating historical maintenance records required careful filtering to exclude outdated or incomplete data. Finally, ensuring regulatory alignment in data collection required careful adherence to aviation safety and compliance frameworks, limiting the ability to include proprietary datasets in the analysis. Despite these challenges, methodological adjustments, such as cross-validation of expert responses and sensitivity analysis, were implemented to enhance the reliability of findings.

3.2. Criteria Influencing the Feasibility of Predictive Maintenance for Aircraft

The feasibility of implementing predictive maintenance in aviation depends on a wide range of interdependent factors. These factors not only determine the technical viability of PM but also its economic, regulatory, and organizational implications.

From the previous section’s analysis, it can be concluded that implementing predictive maintenance for aircrafts is a multi-dimensional challenge that requires careful consideration of technical, economic, regulatory, and organizational factors. By adopting a structured approach to analyze these criteria, stakeholders can assess the feasibility of PM systems and ensure that their implementation delivers maximum value while maintaining the highest standards of safety and efficiency. The taxonomy (Figure 10) provided serves as a guide to systematically address these factors and pave the way for the successful integration of predictive maintenance in the aviation industry.

3.3. MCDA-Based Decision-Making Framework for PM Implementation

3.3.1. General Approach for Decision-Making Framework Development

The decision-making framework for implementing PM of aircraft components utilizes a MCDA approach to systematically evaluate and rank components based on their suitability for PM. This approach addresses the complex and multidimensional nature of maintenance decision-making, ensuring that all technical, economic, regulatory, and organizational factors are considered in a transparent, data-driven, and actionable manner.

The framework aims to identify and prioritize aircraft components that are most suitable for PM, balance diverse considerations, provide a reproducible decision-making process, and align with aviation safety and operational priorities.

The decision-making process involves five key steps (Figure 11).

Step 1: Criteria identification and weighting: Criteria are identified through literature reviews, expert consultations, and empirical data analysis. The AHP is used to assign weights to each criterion based on pairwise comparisons, ensuring that the most critical factors receive appropriate emphasis [36].

Step 2: Scoring and normalization: Quantitative and qualitative data are collected for each criterion and normalized to a common scale (0 to 1) to ensure comparability.

Step 3: Aggregation of scores: Scores are aggregated at two levels:

• Criterion-level aggregation: Normalized scores within each thematic category are weighted and summed to produce a category score.
• Category-level aggregation: Category scores are weighted and combined to calculate an overall PM suitability score for each component.

Step 4: Ranking and prioritization: Components are ranked based on their overall PM suitability scores, which reflect their feasibility for PM. Components are categorized into tiers (e.g., high, moderate, or low suitability), enabling targeted decision-making.

Step 5: Validation and sensitivity analysis: The framework is validated by comparing results with historical maintenance data and expert opinions. Sensitivity analysis is conducted to test the robustness of results under varying criteria weights and scores, identifying the most influential factors affecting the rankings.

The framework’s outputs include a prioritized list of components suitable for PM, actionable insights into the key factors influencing PM feasibility, and recommendations for addressing challenges such as data sufficiency or cost considerations. For example, a turbofan engine might score highly in technical reliability and economic feasibility, making it an ideal candidate for PM, while hydraulic systems with moderate data sufficiency may require sensor enhancements. FMC with low predictability and cost-effectiveness scores may be deemed unsuitable for PM.

This MCDA-based framework offers several advantages. It provides a comprehensive evaluation by considering diverse factors, ensures transparency and reproducibility through structured methods like AHP, and adapts to new data or operational changes. By incorporating expert input and feedback, it builds stakeholder consensus and enhances credibility. The framework supports the aviation industry’s transition to more efficient, reliable, and cost-effective maintenance strategies while maintaining alignment with safety and regulatory priorities.

This decision-making framework provides a robust, adaptable, and evidence-based approach for implementing predictive maintenance in aviation. By integrating multiple criteria, expert insights, and a structured evaluation process, it facilitates informed decisions that optimize operational efficiency and safety while addressing economic and organizational constraints.

3.3.2. Define Decision Variables and Criteria

Creating a decision-making framework for implementing PM of aircraft components requires MCDA approach. We can define a model based on key criteria, each represented as a decision factor that can be quantified and evaluated.

Let $C=\left\{c_j\right\},j=\overline{1.4}$ represent the set of aircraft components being evaluated for predictive maintenance, and $K=\left\{k_i\right\},i=\overline{1.14}$ represent the criteria used to assess each component for predictive maintenance in accordance with Figure 10.

Each criterion $k_i$ is assigned a weight ${0\le w}_i\le 1$ based on its importance, where:

$$\sum _{i=1}^{14}{w}_i=1\mathrm{a}\mathrm{n}\mathrm{d}{w}_i\ge 0$$

Weights can be determined by expert judgment, regulatory priorities, or through AHP to ensure that each criterion’s influence aligns with organizational goals.

Each component $c_j$ receives a score $S_{ij}\in \left[\mathrm{0,1}\right]$ for each criterion $k_i$.

This score indicates how well the component $c_j$ aligns with each criterion $k_i$. For example, reliability predictability score $k_1$ shows the degree to which degradation patterns can be forecasted, reliability predictability score $k_2$ shows the degree to which degradation patterns can be forecasted and so on in accordance with Figure 10.

For each component $c_j$, calculate an overall predictive maintenance suitability score $P_j$ as a weighted sum across all 14 criteria:

$$P_j=\sum _{i=1}^{14}{w}_i{S}_{ij}$$

(1)

The higher the $P_j$ score, the more suitable component $c_j$ is for predictive maintenance.

Define a threshold $T$, such that:

• If $P_j\ge T$, then PM is recommended for component $c_j$.
• If $P_j<T$, then PM may not be beneficial or feasible for component $c_j$.

This threshold can vary depending on organizational goals, budget constraints, or safety considerations.

Conduct sensitivity analysis by varying weights $w_i$, scores $S_{ij}$, or the threshold $T$. This analysis will reveal the following:

• The criteria that most influence the final decision.
• How robust the PM decision is to variations in scoring and weighting.
• Potential adjustments to the model to reflect changing operational needs or new information.

This expanded model provides a robust mathematical framework, allowing aviation maintenance teams to make data-driven, multi-criteria decisions on predictive maintenance for each aircraft component.

3.3.3. Case Studies

The case studies demonstrate the practical application of PM framework, validating its effectiveness through real-world examples. This phase involves selecting a diverse set of aircraft components, applying the framework to calculate their PM suitability scores, and conducting sensitivity analysis to evaluate the robustness of the results. By comparing the outcomes with historical maintenance data and expert opinions, this phase ensures that the framework aligns with industry realities and supports data-driven decision-making.

The selection of components for analysis is crucial to validate the framework across various scenarios. For example, choose turbofan engine and FMC. For each component, we will calculate a PM suitability score based on the expanded decision-making framework using all 14 factors.

The weights $w_i$ for each criterion $k_i$ are chosen based on importance, and scores $S_{ij}$ are assigned based on the component’s characteristics by experts in aircraft maintenance (

Table 3. Criteria weights and scoring for use cases.

i	$\mathbf{Criterion}{\mathit{k}}_{\mathit{i}}$	$\mathbf{Weight}{\mathit{w}}_{\mathit{i}}$	Score for Engine	Score for FMC
1.	Flight Safety Impact	0.15	0.95	0.95
2.	Reliability Predictability	0.1	0.9	0.4
3.	Degradation Progression	0.1	0.85	0.3
4.	Data Sufficiency	0.1	0.9	0.6
5.	Economic Feasibility	0.1	0.8	0.5
6.	Regulatory Compliance	0.05	0.9	0.8
7.	Technological Integration	0.05	0.85	0.6
8.	Environmental Influence	0.05	0.8	0.6
9.	Operational Impact	0.1	0.9	0.7
10.	Scalability	0.05	0.9	0.6
11.	Workforce Training	0.05	0.8	0.7
12.	Data Privacy and Security	0.05	0.9	0.5
13.	End-of-Life Management	0.05	0.85	0.6
14.	Stakeholder Acceptance	0.05	0.9	0.6

).

The thresholds for determining whether a component is suitable for PM are not directly set at calculated scores. These thresholds are established based on operational needs, historical data, expert insights, and industry standards. A structured approach to defining thresholds involves several key steps.

First, an initial threshold is set based on a scoring scale. In this case, scores range from 0 to 1, with higher scores indicating better alignment with PM criteria. A typical starting point for a threshold might be 0.75, reflecting the level of alignment needed to justify PM adoption. Components scoring above this threshold are likely to benefit from PM, while those scoring below may not meet the necessary criteria.

Next, the threshold is refined using expert input and industry standards. Regulatory guidance, safety requirements, and insights from maintenance professionals help determine appropriate thresholds for different component types. For instance, aviation regulations may necessitate higher thresholds (e.g., 0.80–0.85) for safety-critical components like engines, while maintenance experts can identify specific criteria—such as degradation predictability or data sufficiency—that make PM viable.

Historical data and benchmarking further refine thresholds. By analyzing past performance, such as comparing PM-managed components’ scores to their failure rates or cost savings, more accurate thresholds can be established. Industry benchmarking provides additional context, with thresholds varying for low-criticality components (e.g., 0.70) versus high-criticality systems (e.g., 0.85).

Conducting a sensitivity analysis is another critical step. Testing different thresholds (e.g., 0.65, 0.70, 0.75, 0.80) helps evaluate their impact on component selection and cost–benefit outcomes. Higher thresholds may reduce the number of components included in the PM program but enhance overall reliability and cost-effectiveness for those selected.

Thresholds can also vary by component criticality. High-criticality components, such as engines or landing gear, often require thresholds of 0.80–0.85 due to their safety importance. Medium-criticality components, like hydraulic systems, may have thresholds around 0.70–0.75, while low-criticality components, such as interior systems, might be evaluated with thresholds of 0.60–0.65.

Applying this approach, a general threshold of 0.75 was used in the example provided. Using Formula (1), the suitability score for both components can be defined:

$$P_{engine}=0.9225,P_{avionics}=0.635.$$

The turbofan engine surpasses a typical threshold ($T=0.75$), making it highly suitable for PM. Its gradual degradation, critical impact on flight safety, and the ability to monitor and collect data continuously make it an ideal candidate for predictive maintenance. The FMC unit falls below the established threshold, indicating that it is not suitable for predictive maintenance. The unpredictability of failures, lack of sufficient data for predictive modeling, and high costs of adaptation make PM impractical for avionics.

3.3.4. Methodology for Defining Weighting of Criteria and Scoring Components

Defining the weighting of criteria and scoring components across all criteria is fundamental to robust decision-making in MCDA.

A variety of methodologies can be employed to define criteria weights. Analytical techniques like the AHP are widely used due to their structured approach [36]. AHP organizes criteria hierarchically, allowing experts to conduct pairwise comparisons to determine relative importance. The resulting pairwise comparison matrix is normalized, and weights are derived by averaging the rows. Simpler techniques like the weighted sum model involve distributing a fixed number of points among criteria based on their perceived importance [37]. These weights are then normalized and aggregated, providing a straightforward and intuitive approach for less complex decision contexts.

Expert-driven approaches, such as the Delphi method, engage a panel of experts in multiple survey rounds to refine and validate weights [38]. The iterative feedback process ensures consensus and minimizes biases. Data-driven techniques, such as statistical analysis of historical data, offer objective alternatives [39]. For instance, regression or correlation analysis can identify criteria with the strongest influence on outcomes like safety incidents or maintenance costs, translating these relationships into weights. Advanced methods like ML can further refine weights by analyzing feature importance in predictive models, particularly when handling large datasets.

Once weights are defined, the next step is to score components across all criteria. Scoring involves quantifying how well each component performs against specific factors. For quantitative criteria, normalization techniques are applied to standardize scores to a common scale.

This ensures comparability across factors like safety impact (measured qualitatively) and implementation cost (measured in monetary terms). For qualitative criteria, such as stakeholder acceptance, Likert scales can be used, and scores are aggregated from multiple experts to ensure robustness [40].

Aggregating scores involves combining normalized values using weighted averages. At the criterion level, scores are aggregated within thematic categories, and category scores are further combined to produce a composite suitability score for each component with Equation (1).

AHP is used in this study as a widely recognized MCDM method that provides a structured approach in a highly regulated and safety-critical environment, such as aviation. The AHP method is well-suited for this study due to its ability to handle complexity, prioritize safety-critical criteria, and integrate both quantitative and qualitative factors. Its structured and transparent process ensures that the decision-making framework aligns with aviation industry standards, balancing safety, cost-efficiency, and operational effectiveness. By incorporating expert judgment and validating consistency, AHP provides a robust foundation for evaluating and prioritizing components for predictive maintenance, ensuring informed, credible, and actionable decisions.

AHP was implemented in this case study to establish and validate weights for 14 criteria across four main categories in predictive maintenance decision-making. Through pairwise comparisons from 12 aviation maintenance experts, the process revealed clear prioritization patterns. AHP was implemented in this case study to establish and validate weights for 14 criteria across four main categories in predictive maintenance decision-making. Through pairwise comparisons from 12 aviation maintenance experts, the process revealed clear prioritization patterns. Expert pairwise comparisons matrix is presented at Figure 12.

In the pairwise comparison matrix, the blue color intensity indicates the relative importance between criteria:

• Darker blue indicates a stronger relative importance (higher ratio).
• Lighter blue indicates weaker relative importance (lower ratio).
• White cells (no blue) indicate either equal importance (1.0) or where the row criterion is less important than the column criterion (values < 1).

The final pairwise comparison matrix in AHP for this study was derived by systematically aggregating individual matrices provided by 12 aviation maintenance experts. Each expert contributed a pairwise comparison matrix evaluating the relative importance of criteria for implementing PM of aircraft components. This process ensures that diverse expert judgments are combined into a single, coherent decision-making matrix while maintaining the integrity of individual assessments.

Each expert’s pairwise comparison matrix represents their evaluation of criteria using a standard scale (e.g., 1 = equal importance, 3 = moderate importance, 5 = strong importance, up to 9 = extreme importance). The elements of these matrices adhere to the reciprocal property, where the importance of criterion $i$compared to $j$ is inversely proportional to the importance of $j$ compared to $i$. For instance, if an expert rates flight safety impact as strongly more important than economic feasibility with a value of 7, then the reciprocal value of 1/7 is assigned for the reverse comparison.

As an example, Figure 13 shows pairwise comparison matrices of two different experts.

The expert-rating distribution analysis (Figure 14) presents the consensus patterns and variations in ratings across 14 predictive maintenance criteria, based on evaluations from 12 aviation experts.

Figure 14 displays three key metrics for each criterion: median rating (green line), 25th percentile (blue line), and 75th percentile (red line).

Table 4. Statistics for predictive maintenance evaluation.

Criterion	Median	Std Dev	Weight
Flight Safety Impact	9.20	0.38	0.15
Reliability Predictability	8.80	0.42	0.10
Degradation Progression	8.50	0.45	0.10
Data Sufficiency	8.70	0.43	0.10
Economic Feasibility	8.90	0.41	0.10
Regulatory Compliance	9.10	0.39	0.05
Technological Integration	8.50	0.45	0.05
Environmental Influence	8.30	0.47	0.05
Operational Impact	8.80	0.42	0.10
Scalability	8.40	0.46	0.05
Workforce Training	8.70	0.43	0.05
Data Privacy	8.60	0.44	0.05
End-of-Life Management	8.20	0.48	0.05
Stakeholder Acceptance	8.40	0.46	0.05

provides detailed statistics, including median values, standard deviations, and importance weights for each criterion. Flight safety impact shows the highest median rating (9.2) with the strongest consensus, while environmental and end-of-life management criteria display wider distributions, indicating more varied expert opinions.

This AHP-based weighting system in Section 3.3.3 was successfully applied to evaluate two case components—a turbofan engine (achieving a score of 0.9225) and an FMC (scoring 0.635)—demonstrating its practical applicability in discriminating between components suitable and unsuitable for predictive maintenance implementation.

3.3.5. Validation and Comparative Analysis

During the study, validation and comparative analysis was conducted to ensure the reliability, robustness, and practical applicability of the PM framework. This phase involved testing the framework against historical maintenance data, expert evaluations, and real-world case studies. The PM suitability rankings generated by the framework were compared to documented maintenance outcomes, such as failure rates, costs, and downtime, to confirm alignment with real-world performance trends.

Expert validation was also performed, where industry professionals reviewed the framework’s criteria, weights, and outcomes to ensure they reflected practical aviation priorities. Additionally, the sensitivity analysis tested the framework’s stability by adjusting criteria weights, identifying the most influential factors and ensuring consistency across various scenarios. Comparative analysis further demonstrated the framework’s advantages over traditional maintenance decision-making approaches, highlighting its ability to incorporate safety, regulatory, and economic considerations comprehensively.

The results of this phase validated the framework’s effectiveness and established its practical relevance, confirming its utility as a robust tool for optimizing predictive maintenance strategies in aviation.

4. Discussion

4.1. Expected Outcomes of PM Implementation

The implementation of the PM framework is expected to deliver significant advancements in the evaluation and optimization of maintenance strategies for aircraft components. These outcomes will address technical, operational, economic, and organizational aspects, offering both immediate and long-term benefits to aviation stakeholders. By systematically applying the framework, a lot of outcomes are anticipated.

The framework will enable a detailed and systematic assessment of aircraft components to determine their suitability for PM. Components will be evaluated based on technical, economic, regulatory, and organizational criteria, ensuring that decisions are holistic, and data driven. This will allow maintenance planners to prioritize components that offer the highest potential for operational and cost efficiencies while aligning with safety and regulatory requirements.

By providing a transparent scoring and aggregation mechanism, the framework will support informed decision-making. MRO organizations and airlines will have a reliable tool to identify components where PM can deliver the most value, reducing reliance on reactive or scheduled maintenance strategies. This will help organizations optimize resource allocation and improve fleet reliability.

Adopting PM for suitable components is expected to lead to reduced unscheduled maintenance events, minimizing aircraft downtime and improving operational availability. By accurately predicting component degradation and failure patterns, airlines can plan maintenance activities more effectively, reducing delays, cancelations, and associated costs.

The framework will help identify components where PM can generate significant cost savings by reducing unnecessary maintenance activities, extending component lifespans, and avoiding costly failures. Economic feasibility assessments within the framework will ensure that PM investments yield a positive return on investment for MRO providers and airlines.

The framework integrates regulatory compliance and safety considerations into the evaluation process, ensuring that PM implementation supports the highest standards of aviation safety. By identifying components critical to flight safety, the framework will help prioritize maintenance strategies that mitigate risks and enhance overall safety performance.

The framework’s ability to integrate expert input and stakeholder priorities will result in outcomes that reflect the needs of diverse aviation stakeholders. This will foster consensus among airlines, regulators, and MRO providers, supporting widespread acceptance and adoption of the framework. Additionally, the inclusion of expert-driven adjustments will enhance the credibility and practicality of the results.

The framework’s sensitivity analysis will ensure robust outcomes that remain consistent even under varying criteria weights and scenarios. This adaptability will enable its application across diverse fleets, operational environments, and technological advancements. It will also provide a foundation for future enhancements as new data sources and PM technologies become available.

The framework will serve as a tool for PM feasibility evaluation in aviation, setting an approach for structured decision-making in maintenance planning. By providing a replicable methodology, the framework can be adopted and adapted by other organizations, contributing to industry-wide improvements in maintenance practices.

4.2. Roadmap for Implementation of the Proposed PM Framework

The successful implementation of the proposed PM framework for aviation requires a structured roadmap to ensure seamless integration into existing maintenance operations. This roadmap includes six key phases, each addressing critical steps for adoption while aligning with industry priorities, such as safety, efficiency, and regulatory compliance (Figure 15).

The first phase, preparation and stakeholder engagement, focuses on establishing a foundation for implementation. Key activities include identifying relevant stakeholders, such as MRO providers, airline operators, regulatory authorities, and component manufacturers, and building awareness through workshops and briefings. This phase also involves conducting a requirement analysis to assess organizational readiness, including data availability, current maintenance practices, and technological infrastructure.

The second phase, data collection and infrastructure development, emphasizes gathering the necessary data and building the infrastructure to support the framework. Historical maintenance records, operational data, and failure patterns for critical aircraft components are collected, while sensors are installed or upgraded to enable real-time monitoring. IT systems are developed or enhanced to integrate data from multiple sources, ensuring compatibility with the framework’s requirements.

In the third phase, framework customization, the framework is tailored to align with the specific priorities and operational contexts of the organization. This involves refining criteria and weights based on organizational needs, conducting expert workshops to finalize the criteria hierarchy using the AHP, and calibrating the model using historical data to test its applicability to selected components.

The fourth phase, pilot implementation, tests the framework on a small scale to evaluate its performance and identify challenges. A diverse set of components is selected for testing, including those with high, moderate, and low PM suitability. The framework is applied to rank these components, and PM strategies are implemented for the most suitable ones. Outcomes, such as maintenance cost savings, downtime reduction, and safety improvements, are monitored to assess the framework’s effectiveness.

The fifth phase, full-scale deployment, focuses on organization-wide adoption. Training programs are conducted to educate maintenance teams, planners, and decision-makers on using the framework effectively. The framework is embedded into existing workflows and decision-support systems, while a continuous feedback loop is established to refine its functionality based on user input.

The final phase, monitoring and continuous improvement, ensures the framework remains sustainable and adaptable. This involves ongoing analysis of operational and maintenance data to validate and improve predictions, regular updates to criteria weights and scoring mechanisms to reflect changing contexts, and integration of emerging technologies, such as AI-driven analytics and real-time condition monitoring, to enhance the framework’s capabilities.

By following this structured roadmap, organizations can transition seamlessly to predictive maintenance strategies, achieving enhanced safety, reduced operational costs, and improved maintenance efficiency while maintaining alignment with industry standards and best practices.

4.3. Challenges and Limitations of the Study

The implementation of the PM framework in the aviation industry offers significant potential benefits, but it also comes with several challenges and limitations that must be addressed to ensure its success. These challenges arise from technical, operational, organizational, and regulatory factors, requiring careful consideration and mitigation strategies.

A primary challenge is the availability and quality of data. The framework relies on accurate and comprehensive historical maintenance records, operational performance metrics, and degradation patterns. However, many components lack sufficient data due to limited sensor integration or inconsistent data collection practices. Additionally, historical data may be incomplete, outdated, or biased, and standardization across datasets from multiple fleets, manufacturers, or MRO providers can be difficult.

Organizational resistance and the need for effective change management also pose challenges. Transitioning to a PM framework requires significant cultural and procedural shifts, which may face resistance from maintenance teams accustomed to traditional reactive or preventive approaches. Stakeholders may question the reliability of predictive models and worry about potential disruptions to established workflows, emphasizing the importance of building confidence and acceptance through strategic change management.

Regulatory and compliance constraints further complicate the implementation of PM. The aviation industry operates under strict regulatory oversight, and PM methodologies must align with mandated inspection intervals and maintenance practices. Securing regulatory approval for PM strategies can be resource-intensive, requiring thorough validation and documentation to demonstrate compliance.

Expert judgments can be subjective, leading to variability in results. Small changes in weights or scores during the sensitivity analysis may affect component rankings, raising concerns about the stability of outcomes. Ensuring consistency and validation across different scenarios is essential for maintaining the framework’s credibility.

While these challenges may limit immediate adoption, they also present opportunities for innovation and collaboration. By recognizing and mitigating these limitations, the framework can contribute to a more resilient, efficient, and sustainable aviation maintenance ecosystem, ultimately optimizing strategies to enhance safety, reliability, and cost-effectiveness.

4.4. Future Research Directions

The PM framework for aviation presents numerous opportunities for future research to enhance its effectiveness, adaptability, and scalability. As advancements in technology, data analytics, and operational practices continue to shape the aviation industry, addressing current limitations and exploring innovative approaches can further optimize PM strategies and their implementation.

One key area for future research is the integration of advanced predictive analytics and ML. Adaptive models that use real-time data and evolving operational conditions can improve the accuracy of failure predictions. Hybrid models combining physics-based simulations with data-driven approaches may provide deeper insights into complex degradation patterns, while enhanced anomaly detection methods can identify early failure indicators. Research can also focus on improving the explainability of AI models to foster trust among stakeholders and regulatory bodies.

Sensor technology and IoT represent another promising direction. The design and deployment of low-cost, high-precision sensors tailored to various aircraft components can expand PM capabilities. Research into creating sensor networks for continuous monitoring and addressing challenges in data fusion will ensure that disparate data streams are seamlessly integrated for analysis. This will be particularly valuable for critical systems and components requiring precise condition monitoring.

Enhancing the scalability and adaptability of the PM framework is another critical area for future research. This includes adapting the framework for diverse fleets, including legacy aircraft with limited sensor network, and incorporating dynamic weighting mechanisms that adjust criteria importance based on changing operational priorities. Modular frameworks that can be customized for different contexts, such as commercial airlines, cargo operations, and military fleets, will ensure broader applicability.

Economic modeling and cost–benefit analysis also require attention. Future research can refine economic models to quantify direct and indirect benefits of PM, such as improved safety, reduced downtime, and extended component lifespans. Strategies to optimize PM investments for maximum return on investment will help operators and MRO providers justify the initial costs of implementation.

Extending the PM framework beyond aviation into industries such as railways, shipping, and manufacturing offers significant potential. Research can focus on customizing the framework to align with industry-specific criteria and operational requirements while evaluating the transferability of aviation-based insights to these domains.

Future research directions provide ample opportunities to enhance the PM framework and address its current challenges. They will contribute to more efficient, reliable, and sustainable maintenance practices, ensuring long-term operational and economic benefits for aviation and beyond.

5. Conclusions

This research has developed and validated a comprehensive decision-making framework for evaluating the feasibility of implementing PM in aviation, addressing a critical gap in the systematic integration of technical, economic, regulatory, and organizational factors. This study’s findings demonstrate that successful PM implementation requires a balanced consideration of multiple criteria, with flight safety impact, reliability predictability, and data sufficiency emerging as primary drivers of decision-making.

The MCDA approach, incorporating the analytic hierarchy process, provides a robust methodology for evaluating aircraft components’ suitability for PM. Through expert surveys and validation studies, the research identified and weighted 14 critical factors across four categories, establishing a structured framework that aligns with aviation industry priorities. The case studies, particularly the comparison between turbofan engines and FMC, validate the framework’s ability to discriminate effectively between components suitable and unsuitable for PM implementation.

The study’s contributions extend beyond theoretical advancement, offering practical tools for maintenance decision-makers. The detailed roadmap for implementation, coupled with comprehensive validation methods, provides organizations with actionable guidance for transitioning to PM strategies. The framework’s emphasis on safety-critical factors and regulatory compliance ensures its relevance within the highly regulated aviation environment, while its flexible structure allows adaptation to varying operational contexts.

Despite identified challenges, including data availability and organizational resistance, the framework demonstrates robust potential for improving maintenance efficiency and safety. The integration of sensitivity analysis and expert validation enhances the framework’s reliability, while the proposed future research directions outline pathways for continuous improvement and broader application.

This research advances the field of aviation maintenance by providing a systematic, evidence-based approach to PM implementation decisions. The framework’s ability to balance multiple criteria while prioritizing safety and operational efficiency addresses a critical industry need. As aviation maintenance continues to evolve with technological advancements, this framework provides a foundation for future innovations in predictive maintenance strategies, contributing to enhanced safety, reliability, and cost-effectiveness in aircraft operations.

This study’s findings suggest that successful PM implementation requires not only technological readiness but also organizational alignment and regulatory compliance. Future research opportunities, particularly in advanced analytics, sensor technology, and cross-industry applications, indicate the framework’s potential for broader impact beyond aviation maintenance.