Reducing human error in aircraft maintenance: a dual-framework analysis of smart technologies using SWOT and HFACS in Philippine MROs

Human error remains a critical concern in aviation safety, particularly in aircraft maintenance, where failures often originate from latent organizational conditions, incomplete inspections, documentation lapses, and supervisory breakdowns rather than real-time cockpit decision aids (ICAO, 2021; Hobbs and Williamson, 2003). Maintenance work is frequently performed under time pressure, fatigue, and incomplete information, conditions that intensify vulnerability to slips, lapses, and procedural deviations—especially in developing contexts such as the Philippines (Li and Harris, 2006; CAAP, 2022). Maintenance-related errors contribute to an estimated 12%–15% of aviation incidents, with potentially severe operational consequences (Reason, 2020; ICAO, 2023). In the Philippine setting, Civil Aviation Authority of the Philippines (CAAP)audit and oversight findings repeatedly flag improper installations, inadequate inspections, and weak documentation practices across MRO environments, indicating persistent gaps in systemic human-error management (CAAP, 2022).

A key practical challenge is that much of the available evidence for maintenance human factors is text-based (e.g. audit findings, occurrence narratives, corrective action reports, and SMS documentation). Such texts are heterogeneous in detail and terminology, shaped by reporting culture, and may omit critical operational context, complicating consistent interpretation and the extraction of actionable causal chains. While established frameworks support structured coding of such narratives, document-driven research must explicitly address ambiguity, missingness, and interpretive subjectivity to avoid overconfident causal claims.

Smart technologies—augmented reality (AR), artificial intelligence (AI), radio frequency identification (RFID) tool control, and digital twins—have demonstrated potential to reduce cognitive load, improve decision accuracy, and strengthen procedural compliance in maintenance settings (Ceruti et al., 2018; Airbus, 2021; Hrúz et al., 2021). However, implementation is not purely technical: adoption depends on workforce adaptability, infrastructure readiness, change management capacity, and regulatory enablement, which remain uneven across Philippine MROs (Ma et al., 2022; Salas and Bernal, 2021). This scenario creates a dual problem: organizations must diagnose where and why human error emerges and determine which interventions are feasible and scalable under real constraints.

The Human Factors Analysis and Classification System (HFACS) is widely used to trace contributing factors across unsafe acts, preconditions, supervision, and organizational influences (Wiegmann and Shappell, 2003). However, in document-based corpora, HFACS can be diagnostic but not decisional: it classifies error mechanisms yet provides limited guidance for prioritizing interventions, assessing organizational readiness, or incorporating external constraints (e.g., regulatory clarity, vendor ecosystem, or investment capacity) that shape technology uptake. To address this limitation, this study adopts a dual-framework design that integrates HFACS + SWOT (strengths, weaknesses, opportunities, threats). HFACS structures the causal diagnosis of maintenance error signals from text evidence. SWOT also looks at both internal and external factors that affect whether smart technologies can be used, kept up, and turned into safety performance gains (Benzaghta et al., 2021). This integration provides a bridging logic from error diagnosis → feasibility assessment → strategy-aligned intervention mapping, which single-framework approaches often leave implicit.

Accordingly, this study integrates HFACS and SWOT to (1) classify common maintenance-related human errors in Philippine MRO contexts and (2) evaluate the readiness and strategic potential of smart technologies to mitigate these errors. Using document analysis, thematic coding, and framework synthesis, the study supports regulators, MRO leaders, and training institutions in designing human-centered modernization pathways aligned with CAAP’s shift toward more predictive and performance-based oversight (ICAO, 2023; CAAP, 2022).

This study offers three contributions that extend prior HFACS-only or SWOT-only work. First, it introduces a decision-oriented dual-framework method (SWOT + HFACS) that links HFACS error-mechanism diagnosis to adoption feasibility and prioritization under organizational and regulatory constraints. Second, it operationalizes the technology–error linkage in an implementation-ready form through a technology–HFACS mapping explicitly tied to PSFs and structured documentary evidence, thereby strengthening interpretability in text-driven human factors research. Third, it provides a context-specific application to Philippine MRO environments, producing an actionable basis for capability building and technology selection in an emerging-market maintenance setting where evidence is predominantly documentary and adoption constraints are non-trivial.

Human error in aircraft maintenance continues to pose a critical threat to aviation safety, often resulting from complex interactions among organizational, environmental, and individual factors (ICAO, 2021). Aircraft maintenance technicians (AMTs) may commit slips, lapses, mistakes, or violations—as categorized by Reason (1990)—due to conditions such as time pressure, unclear procedures, fatigue, and poor supervision (Yazgan and Yılmaz, 2019; Nkosi et al., 2020).

Globally, maintenance errors cause approximately 12%–15% of aviation incidents, and many of these incidents involve latent failures that are only discovered after flight operations (ICAO, 2023). In the Philippines, CAAP (2022) identified recurring issues in local MRO reports, including improper installation, incomplete inspections, and documentation lapses. These errors can lead to serious outcomes, including air turnbacks, runway excursions, or even in-flight decompression, as illustrated in the 1990 British Airways B747 case (Flight Safety Foundation, 2000). Empirical studies affirm the operational impact of these errors. Hobbs and Williamson (2003) noted that even minor oversights—like missing fasteners—can escalate into catastrophic failures. Li and Harris (2006) further emphasized that AMTs often face adverse working conditions that heighten error risk, particularly in regions such as the Philippines, where workforce shortages and tight schedules prevail.

While advancements in aviation technology have enhanced maintenance workflows, human error remains a leading safety challenge, often more consequential than errors made by flight crews (Reason, 2020). To address this, the industry is shifting toward predictive, data-driven approaches, integrating tools such as AI and proactive human factors assessments to minimize risks (Mendes et al., 2022).

The integration of smart technologies into aircraft maintenance is transforming MRO operations by enhancing safety, improving efficiency, and reducing errors. Among the most impactful tools is AR, which overlays digital instructions onto components, reducing reliance on manuals and minimizing procedural errors. Trials conducted by Boeing and Lufthansa Technik have demonstrated AR’s effectiveness in reducing cognitive load and improving task accuracy, particularly in complex installations (Ceruti et al., 2018; Lufthansa Technik, 2020).

AI and predictive analytics are also reshaping maintenance by enabling early fault detection and data-driven decision-making. Platforms like Airbus Skywise and Honeywell Forge use real-time aircraft sensor data to predict failures and optimize maintenance schedules, thereby reducing unscheduled events and technician errors (Airbus, 2021; Kabashkin and Perekrestov, 2024).

In tool management, RFID systems, such as the Snap-on Level 5 Tool Control System, support foreign object damage prevention by enabling real-time tracking and accountability, ensuring tools are neither misplaced nor left onboard (Hrúz et al., 2021). Other emerging technologies, including digital twins, internet of things-enabled sensors, and wearable or voice-guided maintenance systems, offer new layers of support. Digital twins allow diagnostics without physical disassembly, while wearables and voice systems enhance technician coordination and hands-free operation (Gao et al., 2020; Mohammed et al., 2024).

Additionally, virtual reality (VR) has proven to be an effective training tool in aircraft maintenance education. A recent study conducted in the Philippines demonstrated that VR significantly enhances student engagement and competency development, with notable improvements in knowledge retention, task accuracy, and problem-solving (Dela Peña, 2025). These innovations reflect the broader transition toward Aviation 4.0, where data integration, automation, and AI converge to streamline processes. However, challenges persist, particularly in regulatory alignment, system maturity, and implementation scalability (Ma, H.L. et al., 2022). Continued collaboration among original equipment manufacturers (OEMs), regulators, and MROs is essential to harness the potential of smart technologies in maintenance fully.

Aviation professionals widely use the SWOT framework to evaluate strategic readiness and guide the adoption of emerging technologies. It provides a structured method for assessing internal capabilities and external challenges, particularly in maintenance, repair and overhaul (MRO) operations (Benzaghta et al., 2021). In MRO settings, the SWOT framework has been used to align technology upgrades with workforce skills, regulatory demands, and infrastructure constraints (Chang et al., 2019). Recent studies have applied the SWOT analysis to assess the implementation of smart technologies, including AI, AR and predictive maintenance. Salas and Bernal (2021) examined the adoption of predictive maintenance in Asian airlines, identifying internal resistance and regulatory gaps as key barriers while highlighting the operational benefits. Similarly, Šváb et al. (2024) conducted a SWOT analysis of AI integration in aviation, emphasizing the necessity of improved risk management strategies.

SWOT has also been instrumental in evaluating specific technologies, such as drone-based inspections, VR training and 3D printing. Taneja and Ghosh (2020) identified reduced inspection time as a key strength of drone usage, while concerns included data security and regulatory readiness. In VR-based AMT training, the SWOT analysis identified both accessibility challenges and opportunities for scalable, immersive learning. Šajbanová et al. (2021) used a similar SWOT analysis to evaluate 3D printing in aircraft manufacturing. Despite its broad applicability, traditional SWOT has limitations in prioritization and depth. Enhancements, such as integrating the Analytic Hierarchy Process (AHP) or evolutionary SWOT models, have been proposed to generate more actionable insights (Liangrokapart and Sittiwatethanasiri, 2022; Vlados, 2019). These hybrid approaches allow for better decision-making, particularly when navigating complex technology implementation scenarios in aviation.

Human factors are central to aircraft maintenance safety because procedures, tools, operating conditions, and organizational systems shape technician performance. Models such as SHELL and Reason’s Swiss Cheese show how latent conditions combine with active failures to produce maintenance errors (ICAO, 2021; Wiegmann and Shappell, 2017). Common stressors include fatigue, time pressure, weak supervision, and poor communication, all of which increase slips, lapses, and violations (Rankin et al., 2014; Padil et al., 2018). HFACS is widely used to classify maintenance errors across unsafe acts, preconditions, supervision, and organizational influences (Hulme et al., 2019; Li et al., 2020). Studies show that maintenance events often arise not only from technician mistakes but also from supervisory and organizational conditions that contribute to missed inspections, documentation lapses, and procedural deviations (Harper and Bliss, 2023). However, HFACS alone is mainly diagnostic and offers limited support for intervention prioritization or implementation feasibility.

Related work on performance-shaping factors (PSFs), such as fatigue, workload, and training, helps explain how these conditions affect human error probability (HEP) in maintenance tasks (Franciosi et al., 2019). Recent mitigation efforts emphasize FRMS, proactive screening, and just-culture reporting systems to support learning rather than blame (Miller et al., 2023; Mrusek and Douglas, 2020). Overall, the literature supports a systems view of maintenance error and underscores the need to pair HFACS-based diagnosis with a feasibility-oriented approach to guide actionable interventions.

HFACS is valuable for classifying maintenance-error mechanisms across system levels and shifting attention from individual blame to systemic causes. However, in document-based maintenance research, the evidence is often text-based, uneven in detail and context-sensitive, which can limit coding consistency and causal interpretation. More importantly, HFACS is primarily diagnostic: it identifies where failures occur but does not assess implementation feasibility, such as workforce readiness, digital maturity, regulatory clarity, vendor support, and resource constraints.

SWOT, on the other hand, looks at how well an organization can adopt new technologies and how outside factors affect that process. It doesn't say where mistakes happen or which systems need to be fixed. The gap, therefore, is that HFACS-only studies diagnose error pathways without prioritization under constraints, while SWOT-only studies assess readiness without linking strategy to specific human-error mechanisms. To address this gap, the present study integrates SWOT and HFACS to connect error diagnosis with feasibility assessment and support implementation-oriented prioritization of smart technologies in resource- and regulation-constrained MRO contexts.

The HFACS is a validated framework for systematically identifying the underlying human factors that contribute to errors and accidents. Originally adapted from Reason’s Swiss Cheese Model, HFACS organizes causal factors into four hierarchical tiers: Unsafe Acts, Preconditions for Unsafe Acts, Unsafe Supervision, and Organizational Influences (Small, 2020; Wiegmann and Shappell, 2003). This taxonomy has been applied across diverse sectors, including aviation, maritime, health care, and unmanned aerial vehicle (UAV) operations (Hulme et al., 2019; Widyanti and Reyhannisa, 2020).

In aviation maintenance, HFACS supports root cause analysis by identifying systemic contributors to human error, such as fatigue, poor supervision, or flawed organizational processes. Extensions of HFACS have incorporated tools like Bayesian networks and decision-making models to enhance analytical depth (Lyu et al., 2019; Al-Rabeei et al., 2022). While its reliability is generally high, researchers acknowledge that rater interpretation and contextual adaptation may influence its application (Oncu and Yildiz, 2014).

SWOT analysis is a classic strategic tool for assessing an organization’s strengths, weaknesses, opportunities, and threats. It is widely used in the aviation, health care, education, and technology sectors to support evidence-based planning and the adoption of innovation (Benzaghta et al., 2021; Siddiqui, 2021). In aviation, the SWOT analysis has been used to evaluate MRO strategies, assess the implementation of smart technology and inform AI integration in safety-critical environments (Šváb et al., 2024; Javaherikhah and Sarvari, 2025).

While traditional SWOT is qualitative, researchers have proposed enhanced variants, such as SWOT-AHP or evolutionary SWOT, to improve prioritization and decision support (Liangrokapart and Sittiwatethanasiri, 2022; Vlados, 2019). Its strength lies in its adaptability and capacity to integrate contextual insights for strategic alignment.

This study applies a dual-framework approach that integrates HFACS and SWOT to examine human error and smart-technology adoption in Philippine MROs. As shown in Figure 1, the framework links maintenance error evidence with technology capabilities to generate strategic recommendations. Maintenance error reports, including incident logs, safety audits, and SMS manuals, are analyzed through HFACS to classify active failures and latent conditions across unsafe acts, preconditions for unsafe acts, unsafe supervision, and organizational influences (Wiegmann and Shappell, 2003).

At the same time, smart technologies such as AR, AI, RFID, and digital twins are assessed through SWOT to evaluate internal capabilities and external adoption conditions (Benzaghta et al., 2021). The findings from both frameworks are then synthesized in a mapping stage that aligns specific error types with appropriate technology interventions, identifies strategic gaps, and supports recommendations for reducing human error, improving procedural compliance, strengthening technician performance, and promoting a more data-driven maintenance culture.

This study used a document-based multiple-case design to examine how smart technologies may mitigate human error in selected MRO settings in Clark, Philippines. It integrates HFACS to classify error pathways across unsafe acts, preconditions, unsafe supervision, and organizational influences, and it uses SWOT to assess internal capabilities and external constraints that affect technology adoption. Data were drawn from a bounded corpus of operational and governance records, including incident narratives, audit findings, SMS manuals, training and supervision records, and technology deployment logs. Qualitative analysis relied on thematic coding and HFACS tagging to identify recurring error mechanisms and latent organizational drivers.

To complement the qualitative analysis, the study applied lightweight quantitative triangulation through: (1) descriptive frequency profiling of HFACS-coded factors across cases; and (2) semi-quantitative estimation of how documented PSFs, such as workload, fatigue, procedural complexity, and supervision adequacy, may shift HEP when a technology is plausibly present. This approach is grounded in PSF–HEP logic in maintenance human reliability research (Franciosi et al., 2019) and is consistent with HFACS-based extensions that support probabilistic reasoning when direct experiments are not feasible (Meng and Lu, 2022). Findings were synthesized in a technology–error mapping matrix linking each technology to its targeted HFACS factor(s), estimated direction or magnitude of HEP reduction, and relevant implementation constraints typical of aviation digitalization (Verhagen et al., 2023).

This study used two complementary evidence streams: (i) organizational documents and (ii) targeted expert inputs. Documentary sources included maintenance error and occurrence logs, SMS manuals, safety audit findings, corrective-action records, technology deployment logs, and relevant CAAP regulations and circulars. Together, these formed the primary corpus for HFACS-based identification of active and latent error mechanisms, consistent with HFACS-ME applications in maintenance records (Illankoon et al., 2019).

To strengthen interpretation and reduce misreading of incomplete or institutionally shaped texts, expert clarification was obtained from quality assurance (QA) officers, safety managers, and senior line supervisors. Their inputs helped validate local reporting terminology, resolve ambiguous narratives, and distinguish routine conditions from exceptional events, consistent with recommended document-analysis practice and qualitative rigor safeguards (Dalglish et al., 2020; Johnson et al., 2020). After source identification and screening, data were collected through the structured three-phase procedure described below.

Data collection proceeded in three phases. First, relevant documents were obtained from participating MRO facilities in Clark and nearby aviation hubs, including internal maintenance error reports, SMS manuals, audit records, and applicable CAAP circulars. Document authenticity was verified with QA personnel, confidentiality procedures were observed, and records were screened using inclusion criteria aligned with the study’s focus on human error and technology adoption.

Second, selected texts were deductively coded using HFACS. Each maintenance event was classified across unsafe acts, preconditions for unsafe acts, unsafe supervision, and organizational influences. To strengthen coding reliability, two coders used a shared HFACS codebook, piloted a subset of cases, and refined decision rules before full coding. Disagreements were resolved through coder reconciliation and, when necessary, adjudication by a third reviewer (QA/SMS focal person) using documentary evidence as the primary basis.

Third, technology-readiness content was analyzed through SWOT. Evidence statements were grouped into strengths, weaknesses, opportunities, and threats, with attention to enablers and barriers shaping feasible smart-technology integration. Optional feedback sessions with QA heads and SMS officers were used to clarify ambiguous narratives, validate borderline HFACS placements, and confirm the plausibility of SWOT interpretations in the operational context. A concise summary of this expert clarification process is provided in  Appendix.

The study used a dual-framework analytic workflow combining deductive thematic analysis with structured cross-tabulation. Thematic analysis was selected because it supports theoretically driven coding while remaining flexible for heterogeneous text-based records (Braun and Clarke, 2022). Validated documentary sources, including incident narratives, audit findings and SMS manuals, were coded into HFACS tiers, while text on smart-technology capabilities and constraints was coded into SWOT categories, consistent with SWOT’s role in organizing readiness and external pressures for strategic decision-making (Benzaghta et al., 2021).

To connect both lenses, a technology-error matrix was constructed linking each technology, such as AR, AI analytics, RFID tool control and digital twins, to the HFACS-coded mechanisms it may interrupt. This step follows HFACS-based maintenance research showing the value of structured taxonomies for aggregating and mapping causal factors from deviation records (Illankoon et al., 2019). A final synthesis matrix then traced the pathway from error mechanism to technology option to SWOT condition, showing how each intervention aligns with feasibility and strategic implications in the Philippine MRO context.

To complement the qualitative HFACS–SWOT synthesis, the study included a lightweight quantitative module that generated bounded estimates of how smart technologies may reduce maintenance human error. HEP was operationalized through PSFs commonly documented in maintenance records, including workload or time pressure, fatigue, procedural complexity, communication quality, and supervision adequacy (Franciosi et al., 2019). For each HFACS-coded error type in the corpus, dominant PSFs were identified and scored on a five-point ordinal scale, from very favorable or low risk to very unfavorable or high risk. A baseline HEP index and a technology-adjusted index were then derived through scenario-based rescoring under an assumed “with-technology” condition.

Technology effects were estimated through scenario-based pre/post scoring. For each mapped technology, such as AR, AI decision support, RFID tool control, and digital twins, an “as-is” and a “with-technology” scenario were specified using documented use cases and feasible integration assumptions. Expected PSF changes were converted into relative ΔHEP values. This approach is consistent with HFACS-based probabilistic reasoning where controlled experiments are impractical (Meng and Lu, 2022). To preserve implementation realism, ΔHEP estimates were interpreted alongside SWOT constraints and then incorporated into the Technology–Error Mapping Matrix as numeric or bounded estimates rather than purely qualitative labels (Mendes et al., 2022).

This study followed established ethical standards for qualitative research involving organizational records. All documents obtained from participating MRO facilities were handled confidentially, and identifiable names, company codes, and sensitive operational details were anonymized during extraction and analysis. Access to internal records, including SMS manuals, error logs and audit reports, was obtained through formal requests and informed consent from authorized organizational representatives, typically quality assurance managers or safety heads.

Where expert validation sessions were conducted, participation was voluntary and supported by clear disclosure of the study’s purpose, scope, and intended outputs. The research protocol was reviewed in accordance with institutional ethical guidelines and aligned with the data-handling policies of participating organizations. No data were used beyond the agreed research purpose, and no interviews were conducted without prior approval. These procedures were intended to preserve confidentiality, transparency, and respect for both institutional stakeholders and documentary sources.

The analysis of maintenance documentation revealed that human errors in Philippine MRO environments predominantly fall under the “Unsafe Acts” category of the HFACS, accounting for 41 documented instances. These include skill-based slips, such as improper torque application and rule-based mistakes, like deviations from standard inspection protocols. The second most common category was “Preconditions for Unsafe Acts” (n = 28), which primarily involved fatigue, miscommunication, and environmental distractions such as noise or poor lighting.

“Unsafe Supervision” accounted for 15 occurrences, often associated with insufficient oversight, ambiguous task delegation, or failure to identify high-risk behaviors. Finally, “Organizational Influences” (n = 10) were observed in cases involving resource constraints, policy lapses, or conflicting priorities between safety and efficiency.

This distribution emphasizes that while individual actions are the most frequent, latent systemic and supervisory factors also significantly contribute to error incidence. The results validate the applicability of HFACS as a tool for tracing human error back to deeper organizational roots. Table 1 summarizes the identified errors by HFACS level, while Figure 2 visualizes the relative frequency of each error category.

Cross-case triangulation of maintenance error reports, SMS manuals, technology logs, and expert feedback was used to map smart technologies deployed or considered in Philippine MROs to HFACS-coded error mechanisms. HFACS localized where each technology may plausibly intervene across the four tiers (Unsafe Acts, Preconditions for Unsafe Acts, Unsafe Supervision, and Organizational Influences), while the quantitative module generated scenario-based bounded ΔHEP estimates for each technology–error pairing.

Across the mapped technologies, AR aligned most consistently with unsafe acts, particularly where real-time overlays and step prompts may help reduce omission risk, procedural deviation, and documentation slips. AI and predictive analytics are aligned primarily with preconditions for unsafe acts and, to a lesser extent, with unsafe acts, especially in cases involving fatigue, attention, and workload-related PSFs. RFID tool tracking aligned most closely with unsafe supervision and organizational influences by strengthening traceability, accountability, and documentation integrity. Digital twins, although less prevalent in current practice, align mainly with organizational influences, where they may support planning, coordination, resource allocation, and scenario-based decision support for latent systemic conditions. Wearables and sensors were similarly associated with preconditions for unsafe acts through their potential to detect fatigue, stress, or reduced alertness.

Table 2 presents these technology–error relationships in a more explicit form by showing the targeted HFACS tier, key error types addressed, dominant PSFs affected, and the corresponding bounded ΔHEP estimates derived from the procedure described in subsection 5.5. These values should be interpreted as analytic, scenario-based estimates intended to support comparative reasoning rather than as directly observed empirical effects. Figure 3 complements this mapping by visualizing the relative pattern of estimated ΔHEP ranges across technologies and HFACS tiers.

This study synthesized internal and external dimensions influencing the adoption of smart technologies in Philippine MROs through a SWOT-based lens. Internally, strengths include the operational presence of well-established safety management systems (SMS), which serve as procedural backbones for integrating new technologies. Several MROs have demonstrated adaptive capacity through successful pilot deployments of AR for line maintenance tasks and RFID for tool tracking, indicating a baseline digital readiness and positive workforce receptivity.

In contrast, weaknesses were evident in the fragmented IT infrastructure and siloed data systems, which hindered the seamless integration of technologies such as Digital Twins and predictive analytics. For instance, while some facilities maintain electronic error logs, others rely on manual processes, reflecting inconsistent digital workflows. Additionally, limited investment in technician upskilling for smart tools, such as AR-based troubleshooting or AI-based diagnostic platforms, creates gaps in the diffusion of innovation and operational scaling.

Externally, opportunities stem from the increasing global demand for aircraft maintenance services, particularly in Southeast Asia. This is coupled with the Civil Aviation Authority of the Philippines' (CAAP) regulatory shift toward digital oversight and the emergence of private sector interest in MRO digital transformation (e.g., OEM-led collaborations on AI diagnostics). These trends present opportunities to fund technology, align policy, and develop capabilities.

However,threatss persist. Regulatory ambiguity – especially regarding the use of AI and data in aviation maintenance – poses legal and compliance uncertainties. Resistance from segments of the maintenance workforce, particularly senior technicians accustomed to manual processes, may delay adoption. Additionally, the lack of harmonized digital standards across MRO operators leads to interoperability issues, complicating industry-wide innovation efforts.

As presented in Figure 4, these SWOT elements collectively map the current strategic landscape. Figure 5 visualizes this positioning, placing Philippine MROs in a “Tech-Aware but Infrastructure-Limited” quadrant. This suggests that while the strategic intent and operational interest in smart maintenance are evident, full-scale implementation remains constrained by infrastructural inconsistencies, training gaps and policy uncertainties. Addressing these barriers is critical to transitioning toward a more digitally mature and resilient MRO ecosystem.

Applying the quantitative module described in subsection 5.5, the mapped technologies produced scenario-based bounded ΔHEP estimates that vary by HFACS tier and are intended for comparative prioritization rather than empirical validation.

As summarized in Table 3, the larger bounded estimates appear where a technology plausibly addresses dominant PSFs linked to a given HFACS tier. AR was associated with unsafe acts, with estimated ΔHEP ranges of 20%–35%, reflecting its potential role in step guidance, omission control, and procedural support. AI showed its strongest estimated alignment with Preconditions for Unsafe Acts (20%–35%), where fatigue exposure, workload variability, and vigilance-related factors are more prominent, while its secondary alignment with Unsafe Acts was smaller (5%–10%). RFID tool tracking was associated with unsafe supervision (20%–35%) and, to a lesser extent, organizational influences (10%–20%), reflecting its potential contribution to stronger traceability, accountability, and control routines. Digital twins were mapped primarily to organizational influences, with bounded estimates of 20%–35%, where planning quality, coordination, and resource visibility are relevant. Wearables and sensors showed more moderate estimated effects for Preconditions for Unsafe Acts (10%–20%), particularly in relation to fatigue, stress and physiological strain monitoring.

Taken together, these outputs suggest that AR and selected AI applications may offer nearer-term value for task-level and precondition-related risk management, whereas RFID and digital twins may be more relevant for strengthening supervisory and organizational control structures. However, these ΔHEP values should be read as modeled comparative indicators within the study’s analytic framework, not as validated effect sizes from experimental or operational intervention studies.

This study interprets human error and smart-technology adoption in Philippine MROs through an integrated HFACS–SWOT lens. HFACS localizes where maintenance errors cluster across system tiers, while SWOT clarifies whether technology controls are realistically implementable under current operational and regulatory conditions. The findings show that unsafe acts—particularly skill-based slips and procedural deviations—were most frequently documented, followed by preconditions for unsafe acts such as fatigue, miscommunication, and inadequate task planning. Supervisory and organizational contributors were also evident in audits and safety records, indicating that maintenance error is sustained by latent system conditions rather than by frontline actions alone.

When these patterns are mapped to technology options, AR and RFID align most strongly with execution control and traceability, whereas AI and digital twins align more with planning, forecasting, and system-level support. SWOT findings, however, show why these alignments do not automatically produce realized benefits. AR, for example, may help reduce procedural deviations, but its effectiveness depends on workflow standardization and IT integration. Likewise, AI-enabled fatigue or forecasting tools depend on reliable data architecture and consistent reporting quality. In this sense, Philippine MROs appear technologically aware but operationally constrained: the same infrastructure gaps, capability limitations, and regulatory uncertainties that affect adoption also reinforce the error conditions identified through HFACS.

Taken together, the findings suggest that technology adoption alone is insufficient without concurrent investment in enabling conditions such as training, data governance, standardized workflows, and change management. The study’s central contribution is therefore not only the identification of error mechanisms but also the conversion of HFACS from a classification tool into a feasibility-aware decision framework through SWOT integration and bounded ΔHEP reasoning. This supports more credible prioritization of what to implement first in a resource- and regulation-constrained MRO environment.

The findings provide practical guidance for aligning safety improvement with digital transformation in Philippine MROs. First, technology deployment should be tied directly to HFACS-identified error categories rather than introduced as a stand-alone innovation. AR and RFID, for instance, are likely to be most useful when embedded in targeted error-control routines and linked to documented maintenance vulnerabilities.

Second, MROs should build internal safety-technology road maps that match dominant weaknesses, such as fragmented supervision, procedural deviation, or limited traceability, to appropriate smart solutions. This can improve investment focus, reduce redundant deployments, and strengthen the feedback loop between safety diagnosis and intervention.

Third, regulators and training institutions have enabling roles. CAAP can support wider harmonization by incorporating digital-readiness and data-governance expectations into oversight, while training providers can expand curricula to include digital tool literacy, human factors awareness, and technology-assisted diagnostics. Technology vendors should likewise move toward interoperable, ecosystem-based partnerships with MROs and regulators so that digital tools reinforce, rather than bypass, existing SMS processes.

The study extends HFACS research by repositioning HFACS from a classification taxonomy toward a decision-oriented logic for maintenance intervention. HFACS remains effective for explaining what errors occur and where they propagate, but by itself, it does not indicate which controls are feasible under actual organizational and environmental conditions.

By integrating SWOT, the study adds a feasibility layer that links error pathways to the capabilities and constraints governing adoption. This combined framework explains not only why errors persist, but also why some interventions remain pilot-bound while others are more scalable. In theoretical terms, SWOT + HFACS creates a bridge between safety causation and implementation capability, allowing intervention priorities to be both mechanism-aligned and constraint-aware.

Methodologically, the study also adds bounded-impact reasoning through the ΔHEP module. Although these estimates are not empirical effect sizes, they provide a structured way to compare plausible intervention value across HFACS tiers while conditioning interpretation on readiness and implementation constraints. This makes the framework transferable to other high-reliability contexts where evidence is largely documentary and controlled experimentation is difficult.

Although the empirical setting is the Philippine MRO, the broader pattern is relevant to other emerging aviation markets where digitalization is uneven, and safety evidence remains largely documentary. Similar barriers—variable documentation quality, weak interoperability, limited analytics readiness, workforce upskilling gaps, and regulatory uncertainty—can likewise constrain the scaling of AR-, AI-, and RFID-based solutions beyond pilot deployment.

The SWOT + HFACS logic is therefore potentially transferable because HFACS provides a stable structure for locating ermechanisms,isms and SWOT provides a practical structure for evaluating feasibility under constraint. Even so, transferability is not automatic. Replication in other settings would require recalibration of PSFs, adaptation to local reporting quality and taxonomy and adjustment to the regulatory and vendor environment that determines what constitutes feasible implementation.

The proposed technology-safety road map faces several implementation barriers. Fragmented IT systems and siloed data can weaken interoperability and limit the usefulness of AI and digital-twin applications. Inconsistent maintenance procedures and documentation quality can also reduce the effectiveness of AR guidance and RFID traceability, both of which depend on stable task structures and reliable records.

Resource limitations, particularly in budget and training capacity, may further slow workforce upskilling and long-term sustainment. Cultural resistance among experienced technicians may also reduce adoption fidelity unless change management is accompanied by strong just-culture messaging. Finally, regulatory ambiguity and uneven vendor ecosystems may delay approvals, data-governance alignment, and system integration. These barriers reinforce the need to sequence implementation with readiness-building rather than assume that technology introduction alone will improve safety performance.

Based on the integrated SWOT + HFACS findings and bounded ΔHEP estimates, Philippine MROs should adopt a staged implementation pathway that prioritizes feasible, high-value controls while building readiness for more complex tools. First, HFACS coding should be applied systematically to incident narratives, audit findings, and corrective-action records to localize dominant error mechanisms by tier. Second, technologies should be matched to those mechanisms and sequenced accordingly: AR for task-execution control, AI or analytics for fatigue- and workload-related preconditions, RFID for supervisory traceability and tool accountability, and digital twins as planning and data maturity improve.

Third, SWOT findings should be used to remove adoption barriers before large-scale rollout. This includes standardizing workflows, strengthening documentation quality and data governance, expanding training capacity, and embedding change-management and just-culture routines. Finally, implementation should be aligned early with CAAP requirements on validation, oversight, and data handling. These steps are summarized in the five-step technology–safety road map presented in Figure 6.

This study has several limitations that affect scope, generalizability, and evidential strength. Access to internal materials depended on institutional permissions, producing uneven document completeness across sites and limiting the inclusion of confidential operational records. Because the evidence base was largely documentary, the analysis may also reflect self-presentation bias in how events and corrective actions were recorded. Although these risks were reduced through triangulation, structured coding, dual-coder checks, and expert validation, some interpretive subjectivity remains inherent in HFACS- and SWOT-guided analysis. As a document-based multiple-case study, the work prioritizes contextual explanation rather than statistical inference and may not capture the full variability of maintenance practice across Philippine MROs.

A further limitation concerns the ΔHEP values reported in the study. These are scenario-based, PSF-derived bounded estimates intended to support comparative prioritization within the SWOT + HFACS framework rather than directly observed empirical effects. They should therefore not be interpreted as validated causal effect sizes. Their reliability depends on documentary quality, the plausibility of scenario assumptions, and the consistency of PSF rescoring under technology-enabled conditions.

Future research should strengthen the reliability, reproducibility, and external validity of the method through larger incident data sets and more rigorous quantitative validation. Computer simulation approaches, including Bayesian-network modeling, Monte Carlo sensitivity analysis, and digital-twin-based scenario testing, could refine PSF–HEP relationships and examine how technologies affect maintenance risk under varying conditions. Experimental and quasi-experimental studies, such as controlled task-based trials, before-and-after evaluations, or longitudinal field implementations, would also help establish more robust evidence on the actual effects of AR, AI, RFID, digital twins, and wearable systems on maintenance error reduction. The framework should also be tested in other settings, including military, UAV, and other resource-constrained maintenance environments, to assess transferability and recalibrate the ΔHEP logic to different regulatory, technological, and organizational contexts.

This study used an integrated HFACS–SWOT framework to examine human error and the adoption of smart technology in Philippine MROs. The findings show that maintenance risk is concentrated mainly in Unsafe Acts and Preconditions for Unsafe Acts, while supervisory and organizational factors continue to shape the conditions under which those errors occur. Procedural deviation, fatigue exposure, miscommunication, and control weaknesses remain prominent concerns. At the same time, the analysis indicates that although Philippine MROs are increasingly aware of technologies such as AR, AI, RFID, digital twins, and wearables, broader adoption remains constrained by fragmented digital infrastructure, uneven workflow standardization, limited investment in training, and regulatory uncertainty. These patterns indicate that maintenance error is best understood as a systems-level issue shaped by the interaction of task performance, organizational capability, and implementation readiness.

A key contribution of the study is the technology–error mapping supported by the quantitative module in subsection 5.5. However, the reported ΔHEP values should be interpreted as scenario-based, PSF-derived bounded estimates intended for comparative prioritization within the SWOT + HFACS framework, not as directly observed empirical effects. They therefore do not represent validated causal effect sizes. Rather, they provide an analytic basis for identifying which interventions may plausibly offer greater relative value under the conditions documented in this study. Within this framing, AR and selected AI applications appear more relevant for near-term task-level and precondition-related risk reduction, whereas RFID and digital twins appear more relevant for strengthening supervisory control, traceability, and organizational planning. Overall, the study suggests that the safety value of smart technologies in Philippine MROs will depend less on technological availability alone than on whether enabling conditions are developed to support credible and sustained implementation. The integrated SWOT + HFACS framework therefore contributes a transparent and context-sensitive basis for linking error diagnosis with feasibility-aware intervention prioritization in resource- and regulation-constrained maintenance settings.

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

To improve interpretive transparency in the document-based analysis, targeted expert clarification and optional feedback sessions were conducted with selected personnel from participating MRO-related functions. These inputs were used to clarify terminology in local reporting, resolve ambiguous incident narratives, validate borderline HFACS classifications and assess the plausibility of SWOT categorizations in the operational context. Documentary evidence remained the primary analytic basis; expert input was used only to refine interpretation and reduce ambiguity.