Introduction

AI has nowadays been established as an informative technology in aviation maintenance aiming for enhanced safety, optimistic operation, and cost saving. Through the application of AI-PMD, the aviation industry can envision decreasing equipment failure rates, as well as improving the efficient use of resources (Tursunbayeva and Chaluz-Ben Gal, 2024). However, there are various challenges inherent in integration of AI in aviation maintenance for instance; issues to do with data quality; issues to do with regulatory requirements; issues in integrating with legacy systems; and issues concerning skills gaps. This literature review categorises these problems to give a clear presentation of existing literature and proposes the perquisites for bolstering the AI applicability in the aviation maintenance domain.

Methodology

This paper employs systematic literature review to identify and discuss the varied nature of integrating AI into aviation maintenance processes. To achieve this, this qualitative research employs a comprehensive analysis of academic publications, industry reports, and technical documentation from 2022 to 2024 to identify and explore the complex issues, novel approaches, and strategic deployment factors of AI-supported aviation maintenance systems through a literature review approach.

The literature collection has included technical and academic databases like IEEE Xplore, Science Direct, Google Scholar in addition to aviation databases and industries. Advanced search protocols have combined primary aviation maintenance terminologies with specific AI implementation components, including predictive maintenance systems and data quality management, which has made it possible to pinpoint significant peer-reviewed articles, conference papers, and trade reports that tackle several aspects surrounding AI deployment in aviation maintenance contexts. After carrying out essential searches and reading through the various literature sources that have been collected, it is now possible to filter down the works that directly relate to current aviation maintenance challenges and AI implementation considerations.

The key research issues explore the critical relationships between the necessary data quality and the corresponding predictive maintenance performance, the regulatory compliance concern for using AI systems in maintenance applications, the integration of the AI-based maintenance systems with traditional systems in existing maintenance organizational structures, and the workforce preparedness needed for change in favor of AI-assisted maintenance. In this way, this study focuses on the interdependence of these problems and demonstrates how changes or advancements in one area can significantly affect others within the aviation maintenance ecosystem. Efforts have been made to pay particular attention to practical and realistic aspects of the theoretical frameworks, especially, in the cases where there are numerous implementation issues. The study focuses on considering the impact of data quality on the compliance processes that regulate maintenance, the role of system integration in increasing the operational efficiency of maintenance activities, and the potential of the workforce when it comes to AI adoption in maintenance environments.

All the selected works have been examined for methodological validity, quality of empirical evidence, and relevance to aviation maintenance operations. Empirical papers providing wide-ranging data, cases, or industry applications have been especially highlighted, ensuring findings maintain strong relevance to practical aviation maintenance challenges. The focus of the review process has been on papers that contain real outcomes from operational aviation maintenance contexts, including the difficulties of implementing AI solutions into traditional maintenance frameworks. These criteria keep the essence of theories tied to the operational context of aviation maintenance.

Based on the analysis, preference has been given to the studies examining large-scale AI implementations in aviation maintenance settings, particularly focusing on those addressing multiple challenge areas simultaneously. With the help of this approach, it becomes easier to understand the multiple implementation barriers and how they impact the overall system. The methodology’s focus is on the comparison of patterns between different implementation contexts, helping to establish common challenges and successful solution strategies that transcend specific organizational boundaries. In this way, by comparing the findings systematically, it is possible to identify both regularities and differences in the implementation of the challenges in various contexts of aviation maintenance.

This study also reveals some important methodological concerns related to this rapidly evolving field. The analyzed field is characterized by the dynamic nature of AI technologies and this creates inherent limitations in long-term applicability of specific technical solutions. Also, differences in legal regulation and the practical application of solutions in different countries contribute to the degree of difficulty in the analyzed field. Industry conditions and operation characteristics also differ geographically and this therefore limit the generalization of conclusions in various contexts of aviation maintenance. However, these boundaries have been kept in mind throughout the analysis and results presented in this work to make the analysis reasonable and doable while at the same time considering the dynamic nature of AI technology in the context of the aviation maintenance environment. The methodology also understands the need of variations of maintenance practices and regulations in different regions and hence the findings remain valid in the various operations.

Thus, this research adopts a comprehensive methodological approach to offer a systematic approach to examining the challenges in the implementation of AI in aviation maintenance without losing sight of the real-world application of the concept. These steps also play the important role in maintaining the validity and reliability of the study results as well as make the study academic legitimate and practical for applying artificial intelligence into aviation maintenance operations.

Data Quality and Availability

Data availability and quality are crucial enabling factors for the use of AI in aviation maintenance, as they are for many other AI applications. In order for the AI model to properly fit into the predictive or suggesting apparatus, it is important to have the relevant data (Scaife, 2023). However, insufficient and inconsistent data collection, missing information, or even low-quality data from certain sources might occasionally derail the plan of implementing a predictive maintenance program. Elahi et al. (2023) claim that poor data processing is also harmful to the application of AI since it introduces risks into decision-making, which is especially riskier in the safer industries like aviation. Furthermore, Mendes et al. (2022) note that the need to unify the data used in such algorithms in accordance with particular protocols is another significant issue when utilising AI. In order to address the aforementioned issues, Praxie (2024) has suggested using big data techniques in conjunction with the predictive maintenance concept. This would open the door to higher data quality standards, which would improve the likelihood of fault finding and reduce the frequency of a breakdown. Research on the direction of DM standardisation is still comparatively lacking, despite the literature’s indication that a systematic approach to DM treatment is necessary.

Regulatory Compliance

The major difficulty of using AI in aviation maintenance is regulation compliance. The aviation industry has many stringent rules of compliance, especially when it comes to implementing new technological change as AI is (Stroeve et al., 2022). Hossein and Amirhesam Abedsoltan also argue that the rapid advancement of AI continues to rapidly expand, and thus outstrips changes in safety regulation (2024). Hryniewicz (2024) also note that integration of AI must also operate within regulatory framework that are functional and legal and this may take time and costs. Ucar et al. (2024) equally expound on the reliability question, as superintendence authorities demand explanation of how AI systems arrive at particular decisions where human life is at risk. However, despite the evidence of the requirement for new forms of regulation, specific effective strategies to address the problem of the gap in regulatory AI integration are scarce at the time of writing this article and may require further research and potentially policy formation.

Integration with Legacy Systems

Due to various different technologies and infrastructures already in use in aviation maintenance systems, the incorporation of AI in these environments is still a problem because of the need for large amounts of funds. During interview with Kabashkin et al. (2023), it was revealed that aviation is greatly constrained by existing systems that cannot support new AI solutions and services. Likewise, Merlo (2024) has it that equipping old systems with new advanced features of AI calls for significant costs and oversolicitous reconstruction that often involves making overhual or replacement of even hard and software parts. This is where Elahi et al. (2023) accomplish the job by stressing on the concept of modular AI systems, where implementation can go step by step rather than an overhaul. But according to the findings of Hryniewicz (2024) most organizations are reluctant to employ an AI solution because of the costs that are involved as well as the interruptions in business operations. The literature points to a need for continued research in how AI can be implemented in an effective and inexpensive manner, including methods that can be implemented in present organizational sources without the need for expensive new infrastructure.

Skill Gaps in Aviation Maintenance Teams

Technically proficient individuals who can manage and examine AI-automated systems are required for the application of AI in aviation maintenance. According to Sanders and Wood (2023), it is essential to teach employees how to use AI because they cannot fully benefit from it if they lack the necessary understanding. According to Kabashkin, Mišņevs, and Zervina (2023), the biggest barrier to integration is the lack of skill distinction in AI training because the majority of maintenance staff are not prepared to comprehend how AI-based systems in aviation organisations work. Additionally, in 2024, Tursunbayeva and Chaluz-Ben Gal created a training framework for digital leaders that can be used in the aviation industry. It ensures that staff members are trained to acquire the necessary abilities that are necessary for the AI implementation. Furthermore, in order to create specialised training programs that would produce a new generation of maintenance professionals, Hossein and Amirhesam Abedsoltan (2024) suggested partnerships with educational institutions. In light of these recommendations, there is a deficiency in the creation of sophisticated training solutions for the use of AI in the aircraft maintenance sector, as evidenced by the lack of research conducted to provide industry-specific, trustworthy training solutions.

Synthesis of Findings

An examination of the literature on AI applications in aviation maintenance reveals both opportunities and difficulties. The listed OE important studies’ forcing factors all agree that the primary problems are skill gaps, legacy system integration, regulatory compliance, and data quality. High-quality data and sound regulations are widely agreed upon (Elahi et al., 2023; Stroeve et al., 2022), but the issues around their resolution are not as well defined. Furthermore, there aren’t many useful implementation suggestions in the body of existing literature, despite the fact that the practical solutions resulting from current research include the usage of modular AI systems and targeted employee training (Merlo, 2024; Sanders & Wood, 2023). This evaluation calls for more studies on efficient administration of large-scale data for large corporations, impacts of new rules, cost-efficient approaches to integrate AI in enterprises, and sufficient training programs in the aviation sector.

Conclusion and Purpose Statement

Finally, it is suggested that there exist promising outcomes when AI is applied to aircraft maintenance and generally, increasing the degree of AI usage in aviation increases its efficiency and safety levels. However, in order to unlock this potential several areas of challenges related to data quality, regulations, communication between legacy systems and lack of skills before data science is initiated need to be solved. In an attempt to encourage proper direction for future research and policy, these barriers are outlined in this literature review. This capstone project aims to provide recommendations that would help eradicate these problems and guide the aviation industry in profiting from the use of AI in maintenance management. It suggests that a systematic way to such concerns can help in building the better future of aviation maintenance.

References

Aizhan Tursunbayeva, & Hila Chaluz-Ben Gal. (2024). Adoption of artificial intelligence: A TOP framework-based checklist for digital leaders. Business Horizons, 67(4). https://doi.org/10.1016/j.bushor.2024.04.006

Elahi, M., Afolaranmi, S. O., Lastra, M., & Pérez, A. (2023). A comprehensive literature review of the applications of AI techniques through the lifecycle of industrial equipment. Discover Artificial Intelligence, 3(1). https://doi.org/10.1007/s44163-023-00089-x

Hossein Abedsoltan, & Amirhesam Abedsoltan. (2024). Future of process safety: insights, approaches, and potential developments. Process Safety and Environmental Protection. https://doi.org/10.1016/j.psep.2024.03.034

Hryniewicz, R. (2024, February 13). 7 Main Problems in Aviation Maintenance (and How New Technologies Will Solve Them). Nsflow.com. https://nsflow.com/blog/problems-in-aviation-maintenance

Kabashkin, I., Mišņevs, B., & Zervina, O. (2023). Artificial Intelligence in Aviation: New Professionals for New Technologies. Applied Sciences, 13(21), 11660–11660. https://doi.org/10.3390/app132111660

Mendes, N., Geraldo Vidal Vieira, J., & Patrícia Mano, A. (2022). Risk management in aviation maintenance: A systematic literature review. Safety Science, 153(1), 105810. https://doi.org/10.1016/j.ssci.2022.105810

Praxie. (2024, January 22). Smooth Sailing in the Sky: The Role of Predictive Maintenance in Aviation. https://praxie.com/predictive-maintenance-in-aviation/

Sanders, N. R., & Wood, J. D. (2023, November 3). The Skills Your Employees Need to Work Effectively with AI. Harvard Business Review. https://hbr.org/2023/11/the-skills-your-employees-need-to-work-effectively-with-ai

Scaife, A. D. (2023). Improve predictive maintenance through the application of artificial intelligence: A systematic review. Results in Engineering, 101645. https://doi.org/10.1016/j.rineng.2023.101645

Stroeve, S., Smeltink, J., & Kirwan, B. (2022). Assessing and Advancing Safety Management in Aviation. Safety, 8(2), 20. https://doi.org/10.3390/safety8020020

Tereza Raquel Merlo. (2024). Emerging Role of Artificial Intelligence (AI) in Aviation. Advances in Mechatronics and Mechanical Engineering (AMME) Book Series (Print), 25–41. https://doi.org/10.4018/979-8-3693-0732-8.ch002

Ucar, A., Karakose, M., & Kırımça, N. (2024). Artificial Intelligence for Predictive Maintenance Applications: Key Components, Trustworthiness, and Future Trends. Applied Sciences, 14(2), 898. mdpi. https://doi.org/10.3390/app14020898

‌