Artificial Intelligence in Aviation: A Review of Machine Learning and Deep Learning Applications for Enhanced Safety and Security

Waqas Ahmed
Air University, Islamabad, Pakistan
Correspondence to: waqaskhattak99@gmail.com

DOI: https://doi.org/10.70389/PJAI.100013

Additional information

• Ethical approval: N/a
• Consent: N/a
• Funding: No industry funding
• Conflicts of interest: N/a
• Author contribution: Waqas Ahmed – Conceptualization, Writing – original draft, review and editing
• Guarantor: Waqas Ahmed
• Provenance and peer-review:
  Commissioned and externally peer-reviewed
• Data availability statement: N/a

Keywords: Predictive maintenance, Anomaly detection, Aviation cybersecurity, Flight path optimization, Explainable AI.

Peer-review
Received: 5 January 2025
Revised: 21 January 2025
Accepted: 28 January 2025
Published: 5 February 2025

Introduction

Overview of Safety and Security Challenges in Modern Aviation

The aviation industry is critically dependent on safety and security. Despite stringent regulatory frameworks and advanced technological systems, challenges such as mechanical failures, human errors, cybersecurity threats, and airspace congestion continue to pose significant risks. Strengthened by aging aircraft, weather conditions that are hard to predict, and growing air traffic volumes, safety concerns are ratcheted up. In addition, as communication and onboard systems in aviation networks intensify, a cyberattack can continue to result in disastrous consequences.¹ To solve these challenges, we have to move from reactive to proactive means of safety, which requires a reliance on predictive tools and automated systems to anticipate and contain risks before escalation.

Artificial Intelligence (AI) and Its Transformative Potential in Aviation

The industry by which machines learn, reason, and make decisions (a role humans have traditionally handled) is being disrupted by AI. In aviation, AI’s potential is to process huge datasets and look for patterns humans cannot see. AI is impactful, but the subfields of machine learning (ML) and deep learning (DL) are the impactful subfields of AI.² ML algorithms train on historical and real-time data to improve outcomes, just like DL, which uses neural networks to imitate complex decision-making processes. These technologies bring transformational solutions that enable the optimization of design for maximum safety, expand the scope of security protocols, and reduce wasteful operations. Examples of AI applications are predictive maintenance, anomaly detection, and cybersecurity, all of which are now becoming essential parts of the aviation industry to the advanced level by which it can operate as intelligently and autonomously as possible.³

Objectives and Scope of the Review

Primary Objective

A review and analysis of ML and DL applications for increasing aviation safety and security.

Secondary Objectives

• To explore the key use cases of ML/DL in aviation that include anomaly detection, predictive maintenance, and cybersecurity.
• To evaluate the potential for AI-driven solutions to overcome the problems faced by Aviation Safety and Security today.
• To identify the gaps in existing research and advocate for future directions for integrating AI into aviation.

Exploratory Objective

To discuss emerging trends and potential advances in AI technologies that can lead to an even further revolutionization of the aviation industry.

Scope of the Review

In reviewing, we will present new AI trends and technology that may bring about disruption to aviation as we know it today. It treats explainable AI (XAI) as a key topic. The application of quantum computing to enhanced AI performance, followed by the application of collaborative AI systems for complex decision-making in air traffic management (ATM) and urban air mobility, respectively, conclude.

Methodology

This systematic review depended on select criteria to identify relevant studies. Articles from peer-reviewed journals and industry reports from 2013 to 2023 were selected to study the latest ML and DL applications in aviation safety and security. The research team searched for specific studies that explained accurate predictions in maintenance work plus advancements in air traffic control systems security and aviation risk evaluation techniques. Literature from four essential databases was retrieved: IEEE Xplore, SpringerLink, ScienceDirect, and PubMed. The review employed Boolean search filters with terms such as “ML in aviation” and “DL for aviation safety” to obtain focused research outcomes. Some papers from the review were removed because they did not present measurable data, used theory only, or were not in the English language. The titles and abstracts were evaluated first, followed by all the papers that qualified through our review. The detailed selection method made sure we included only top-quality research papers that fit our paper’s purpose.

Background

AIS Evolution in Aviation: You can quickly say that the history of AI in aviation is the history of progress. Early implementations relied upon rule-based systems that did not require reprogramming yet could only perform very simple control tasks. However, this consistency was non-adaptable or learning.² The algorithms have become more sophisticated, the data available easier to store and analyze, and the computation power has increased by leaps and bounds through the decades to what AI is today, a powerful tool that can dynamically decide and solve complex problems.

Overview of ML and DL Methodologies: ML is a broad field and encompasses many things that might not always be obvious to people who are new to the field. However, a core component that we will cover in these tutorials is ML, sometimes referred to as “learning from data and improving performance without explicit programming.” It can spot the patterns and relationships in data and use that data to perform predictions or decisions. There are three types of ML methodology. In short, supervised learning involves using a model that pairs the input and output of the dataset, and the technique used is based on labeled datasets.⁴

A subset of ML known as “deep learning” (DL) involves artificial neural networks, which are inspired by the brain’s structure and function. These deep neural networks consist of multiple layers, each more and more abstract of the extracted features of the data from the data, to do such sophisticated tasks as image and speech recognition.⁵ DL has revolutionized image processing with a powerful definition of spatial patterns, and convolutional neural networks (CNNs) have successfully defined spatial patterns. In aviation, bag X-ray image security systems are dependent upon CNNs for safety improvement as they analyze X-ray images to detect prohibited items.⁶ Recurrent neural networks (RNNs) are another DL architecture that is designed to learn sequential data and has been utilized in predictive systems using time series data—determining flight parameters at each instant to identify anomalies.⁷

Applications of AI in Aviation Safety

Predictive Maintenance

One of the best applications of AI in aviation safety is based on predictive maintenance. Airlines and maintenance teams are allowed to switch from reactive maintenance to proactive interventions by leveraging the opportunities when economics allows the incorporation of the use of ML or DL algorithms. In predicting when things are going to break, AI models use historical performance data and real-time sensor input from airplane components to anticipate injury taking place.⁸ Improvement in this predictive capability increases operational safety by reducing downtime and catastrophic equipment failure and enabling timely repair. For instance, sensor data streams in use for engine health monitoring systems are continuous feeds of information trying to find tiny deviations in performance that proclaim the origin of a fault.⁶ For example, there are cases of how AI has been employed in the Rolls-Royce application of AI for predictive maintenance case studies, which helps improve both operational efficiency as well as safety and, hence, proves that AI can work in such a way by saving some considerable costs.⁷

Anomaly Detection

Yet AI also has a significant role to play in another major application, which is anomaly detection, a technique that significantly enhances aviation safety by discovering departures from expected performance patterns. ML algorithms regularly parse through vast quantities of flight data to identify flaws that point to mechanical failure, sensor inaccuracies, or process changes.⁸ Real-time anomaly detection systems enhance situational awareness for both pilots and ground control by alerting pilots and ground control to potential hazards as they occur. Systems can, for example, monitor flight trajectory data to find unusual patterns associated with mechanical problems or navigation errors. Moreover, AI does what it does best in ATM: AI-driven anomaly detection detects abnormal traffic patterns that could heighten collision risks (Figure 1).⁹

Flight Path Optimization

On the path of becoming an essential AI-inspired innovation for aviation efficiency and safety is flight path optimization. Static routing used in standard flight planning can lead to undesirable fuel consumption and exposure to hostile weather. Real-time factors such as weather conditions, air traffic congestion, and turbulence risks are taken into account by AI algorithms, and route recommendations are made based on them.¹⁰ AI-based optimization integrates to reduce travel time, fuel consumption, and environmental impact while potentially improving passenger comfort by avoidance of turbulence-prone flights. Examples include satellite weather data combined with ML model predictions, of course, to personalize adaptive routing recommendations.¹¹

Pilot Assistance and Automation

Pilot assistance and automation, as well as human decision-making and operational control, have been greatly advanced by AI. DL models enable pilots to quickly analyze complex scenarios that occur in real time with actionable insights and risk assessments which can support critical decision-making. For example, AI-driven tools check takes off, landing, and emergency risk factors and provide recommendations to improve situational awareness.¹² In addition, DL algorithms are being used in modern autopilot systems to handle demanding maneuvers, as well as to preserve flight stability in a safe manner that reduces pilot workload. AI-driven pilot support and automation provide a well-combined method offering a balance between relying on human expertise augmented with intelligent systems to achieve the safest outcome (Figure 2).⁹

Applications of AI in Aviation Security

Cybersecurity

With the increasing use of digital systems for aviation communication, navigation, and control, aviation cybersecurity has become an increasingly important consideration. The security of aviation networks is provided by AI-based solutions against ever-increasing cyber threats. ML models are used to detect anomalies in the network traffic and system behavior related to potential attacks. Further, DL is powerful in deepening cybersecurity capabilities in pattern recognition that uniquely detects malicious Automatic Dependent Surveillance-Broadcast (ADS-B) messages.¹⁴ Spoofing attacks are possible against ADS-B, one of the core tools of modern air traffic surveillance, leading aircraft tracking and control. AI-based algorithms can monitor message authenticity and detect irregularities that might represent tampering.¹³

Passenger and Baggage Screening

Passenger and baggage screening at airports have benefited from AI-driven advancements in security processes that are both efficient and accurate. Manually inspecting luggage scans was, traditionally, a complicated, time-consuming, and error-prone process. With computer vision and DL models, the process has been automated using AI solutions, and prohibited items have been substantially detected.¹⁴ These models, trained on large amounts of X-ray images, can identify weapons, explosives, and restricted material at higher precision than human operators. AI-equipped automated screening systems can screen images in real time, cutting down on wait times for passengers while maintaining the highest levels of security (Figure 3).¹¹

ATM

Another critical area where applications of AI improve safety and efficiency is ATM. As data shows, the time necessary to avoid, deflect, or immobilize an aircraft exceeds the time it takes to collide with other aircraft, so real-time aircraft decision-making is needed to prevent collisions and mitigate congestion.¹² ML models find potential bottlenecks and suggest a route adjustment so that traffic continues to flow smoothly; meanwhile, DL algorithms help collision avoidance systems constantly assess aircraft trajectories and give alerts when separation minima are at risk. While it does make driving safer, AI also helps increase operational efficiency by automating key components in traffic management: fewer delays and less fuel consumption.¹³

Drone Surveillance and Management

New security challenges have resulted from the proliferation of drones near airports and other critical infrastructure. Unmanned drones are unauthorized and can disrupt flight operations, compromising airspace integrity and presenting serious safety hazards. Since AI-based drone detection and management systems rely on computer vision and a category of DL technology, they offer robust solutions.¹⁶ These systems analyze video feeds and sensor data, looking for and tracking drones, distinguishing harmless from potentially harmful devices in real time. Beyond detection, AI-driven systems enable the automation of threat mitigation strategies, e.g., the use of counter-drone technologies that destroy or divert unauthorized drones without causing collateral damage.¹⁷

Challenges in AI Implementation

Data Availability and Quality

The availability and quality of relevant data are some of the biggest hurdles to the widespread use of AI within aviation. DL and ML models heavily rely on massive datasets to train and generate good results.¹⁸ The aviation industry has also created enormous amounts of unstructured and heterogeneous data from sensors, logs, and CE systems that must undergo massive preprocessing and labeling before intelligence models can be trained. Income from a lack of access to high-quality, standardized datasets can cause biased or inaccurate AI predictions that destroy the reliability of safety-critical applications.¹⁹

Scalability and Computational Demands

The other formidable challenge for aviation stakeholders when it comes to AI models is scalability and computational requirements. Training even complex DL architectures like CNNs or RNNs eats up a tremendous amount of computing: it requires powerful GPUs or special-purpose hardware such as TPUs. These demands are further exacerbated by the real-time processing needs of AI systems in applications such as anomaly detection and autonomous flight control.²⁰ The big thing about investment is that implementing AI at scale across the entire fleet or network is pretty much nuts and bolts of giving it to you cold, but it is economically prohibitive for smaller airlines or smaller aviation service providers.²¹

Regulatory and Ethical Concerns

Accountability and transparency issues bubble to the surface when AI systems take control or make autonomous decision-making about flight safety.²² To prove, certify, and subsequently get validated, AI systems must meet strict safety standards, and the regulatory bodies must set clear guidelines for these systems as to what needs to be done to validate and from where. Meanwhile, the transparency problem of the black box nature of many DL models makes it difficult for the regulators to understand what is happening inside. This lack of explainability also raises ethical concerns about fairness, bias, and trustworthiness.²³

Integration with Legacy Systems

AI integration demands large-scale reworking of legacy infrastructure: updating the existing communication of data, software interfaces, and system interoperability. This is on top of the fact that strict safety regulations are preventing any modifications to aircraft systems, and hence, every modification needs to be rigorously tested and validated before being used in an aging aircraft. Additionally, integrating also introduces new risks of introducing new vulnerabilities.²⁴

Emerging Trends and Future Directions

Improved Trust Using XAI

ML and DL models, with their complex structure, leave aviation professionals struggling to find the practical applications they need. The rise of XAI helps aviation practitioners understand the processes behind AI model predictions and decisions. To understand AI model predictions better, stakeholders can use feature attribution methods like Shapley Additive Explanations to find which parameters matter most, such as identifying engine failure markers in predictive maintenance.²⁵ Saliency maps allow AI model visualization to identify specific image parts that help object detection systems find unauthorized drones in restricted airspace. These advancements increase trust between engineers and regulators by showing them safety rules clearly and helping them follow regulatory requirements.²⁶

Advanced AI Capabilities Using Quantum Computing

Quantum computing technology lets AI systems resolve aviation problems in ways faster than current methods allow. Traditional computers handle binary data through single bits, but quantum computers deliver parallel operations using qubits. Flight route optimization can run many times faster through quantum algorithms when it examines millions of factors, including fuel usage, weather patterns, and air traffic patterns. During severe weather disruptions, quantum AI programs can instantly adjust flight paths for all aircraft in an airline fleet to cut down on flight delays and fuel usage.²⁵ Predictive maintenance systems can leverage quantum ML to examine aircraft sensor information, which helps them forecast component breakdowns and automate maintenance scheduling. Early quantum computing advancements will soon help improve how we manage air traffic and detect security risks in aviation systems.²⁷

Collaborative AI for Multi-Stakeholder Decision-Making

The success of aviation operations needs all stakeholders, including airlines, air traffic controllers, maintenance teams, and regulators, to function as one unit. New AI platforms unite multiple data sources to create comprehensive views that support smarter decision-making.²⁷ An AI network combines instant weather updates with flight maintenance plans and air traffic status to make dynamic airport schedule updates. By using AI, the European airport improved terminal capacity planning through flight data analysis, resulting in smoother passenger processing and enhanced peak-hour operations. AI technology allows unmanned drones and flight crews to exchange real-time updates that protect airspace users as flight traffic increases.²⁸

AI in Next-Generation ATM

ATM systems need immediate modernization because travel demand has reached new heights. AI-driven systems create future traffic forecasts and find solutions for airflow control as they monitor current traffic conditions. AI technology scanning airspace continuously helps detect and prevent air traffic disturbances before they happen. Automatic systems powered by AI make better flight route choices for drones and air taxis to work well within standard air traffic control networks in urban areas. As an example of an AI application NASA successfully tested autonomous drone flight management in its ATM eXploration project. AI technology needs to transform aviation management systems before autonomous aircraft multiply in our busy airspace.²²

Discussion

Integrating AI into aviation has significant benefits, including safe, efficient, and cost-efficient integration. Predictive maintenance systems lower downtime along with the elimination of costly failures, anomaly detection, and cybersecurity solutions enhance situational awareness and threat response. All these advantages lead to some limitations of AI.²³ Another significant problem is data dependency—many models require enormous (sometimes felt qualitative) datasets, which are not always provided in aviation. More constraints on scalability include computational complexity and resource demands. Furthermore, regulatory hurdles and problems of XAI models impede adoption in safety-critical domains.²⁴ The rules and regulations-based system of traditional aviation safety is coupled with historical trends and manual oversight. While effective, these methodologies are not as flexible as they could be and do not provide a similar degree of predictive power as the methods offered by AI.²⁶ This is where AI-driven systems are markedly excelling in processing large volumes of real-time data to detect the slightest anomaly or predict future failures that traditional inspection methods may miss. For example, sensor readings from machines also allow ML models to forecast wear in aircraft components better than scheduled maintenance approaches that are based on fixed timeframes.²⁹

To accelerate the rate of adoption of AI, a multi-faceted collaboration strategy among researchers, industries, and policymakers is needed.³⁰ Therefore, regulatory frameworks need to catch up and define clear guidelines for AI validation, safety assurance, and accountability. Those investments in data-sharing initiatives, standardization efforts, and XAI will fortify the foundation for broader implementation.³¹ Furthermore, advancements will be driven by an innovation-friendly environment that will promote partnerships between aviation organizations and technology firms. Another important part will be training programs to upskill aviation professionals in AI competencies toward transition.

Limitations and Challenges in AI Adoption in Aviation

Technological Constraints

AI systems use ML and DL technologies but need substantial data inputs to work effectively. The aviation industry struggles to unite data from many systems across aircraft models while standardizing it for different operational regions.²⁸ Predictive maintenance systems need exact and complete component data from sensors to work appropriately.

Regulatory and Certification Hurdles

Air transportation standards demand new technologies to work perfectly without safety issues. AI systems must complete extensive tests before receiving FAA and EASA certification to operate safely. Air traffic control and aircraft automation systems must confirm their ability to maintain safety across different operating conditions.

Ethical and Trust Issues

Aviation stakeholders find it difficult to have faith in AI systems due to their mysterious decision patterns. Air traffic controllers and pilots want to trust automated systems but need to know the logic behind emergency flight reroutes before they activate them. An AI passenger screening system could accidentally harm specific groups, helping to shape complaints from both ethics and law experts.²²

Integration Challenges

AI systems face technical and operational barriers when we try to add them to current aviation platforms. Traditional aviation systems, like old air traffic controls and maintenance processes, do not work well with new AI technology.

Cybersecurity Risks

AI technology implementation in aviation brings fresh threats to aviation cyber security. AI systems that use cloud computing for data processing have cybersecurity weaknesses that cyberattacks can exploit to damage vital functions and expose confidential data.²⁰

Economic and Skill Barriers

AI systems need considerable financial support to develop and operate, yet small aviation companies often struggle to afford these costs. Making AI work properly in aviation requires people who know AI technology and aviation systems and who have experience with regulatory guidelines. The inability to find enough people with AI expertise stops many companies from adopting AI systems across industries.

Conclusion

Aviation safety and security are being transformed by AI, which is bringing innovative solutions to long-standing challenges. AI has shown its potential to transform the industry through predictive maintenance, anomaly detection, and greatly enhanced cybersecurity. Continued research, cross-sector collaboration, and proactively developed regulation are important to maximize the benefit derived from AI. As aviation turns to a data-driven future, AI will remain among the main drivers of progress, helping make air travel safer, more secure, and more efficient. That leaves AI’s transformative power to be fully realized in years to come and a collective effort among technology developers, aviation authorities, and policymakers.

Practical Implications and Recommendations for Integrating AI into Aviation Systems

AI systems bring critical benefits to aviation by making operations more efficient and increasing both safety and predictive power.

Data Infrastructure and Standardization

To introduce AI into aviation, we need to first create an effective data system that keeps all information together. The aviation sector maintains its data across multiple isolated platforms used by airlines, air traffic control, and maintenance groups. The scattered nature of the data makes it hard to use AI effectively. Air transport must establish consistent data standards that ICAO directs and regulates.

Regulatory Alignment and Certification Pathways

AVMC does not yet have strong rules to certify adaptive AI systems for aviation use. Aviation stakeholders need to partner with regulatory agencies like the FAA and EASA to establish certification rules for emerging technology. Digital twin models create realistic environments to prove that AI systems work safely. Simulation testing shows AI systems in multiple operational conditions to speed up and lower certification expenses. The creation of XAI models represents a key component in these efforts. Showing AI systems’ decision-making logic with XAI enables regulation compliance and helps stakeholders trust AI systems.

Workforce Development and Upskilling

To make AI systems work in aviation, we must teach aviation staff the proper techniques for using these advanced tools. For AI to work well in aviation operations, staff, including pilots, traffic controllers, and maintenance personnel, need specific education about these tools. AR and VR systems can build virtual training spaces that let staff practice with AI systems in real environments.

Cybersecurity Measures

Air transport operators must strengthen their digital protection as AI controls increase and linked systems become more common. AI systems must watch computer networks for safety holes and stop threats instantly. Blockchain technology provides secure data interchange protection for linked AI systems to maintain data authenticity.

Pilot Projects and Incremental Integration

The introduction of AI technology should start inaccessible operations, such as airport resources and baggage systems, to validate their effectiveness before applying them to critical tasks. Narrow AI projects help stakeholders discover system weaknesses and optimize functions before rolling out AI across large-scale operations.

Collaborative Ecosystems and Stakeholder Engagement

Multiple aviation organizations need to partner up because flying operations require many different teams. Several interested parties must work together in dedicated teams to ensure consistent AI plans and spread good practices among all groups. AI systems that combine data from all sources enable organizations to make linked decisions simultaneously throughout the aviation industry network.

Advanced Use Cases and Continuous Improvement

The aviation sector sees substantial growth potential through advanced AI applications for flight prediction maintenance and automated airspace operation alongside self-operating systems. AI technology in aviation tracks real-time flight information about weather patterns, flight routes, and traffic flow to improve airspace usage and shorten delays between planes.

References

1 Alreshidi I, Moulitsas I, Jenkins KW. Advancing aviation safety through machine learning and psychophysiological data: A systematic review. IEEE Access. 2024. Available from: https://ieeexplore.ieee.org/abstract/document/10379685
https://doi.org/10.1109/ACCESS.2024.3349495
 
2 Helgo M. Deep learning and machine learning algorithms for enhanced aircraft maintenance and flight data analysis. J Robot Spectr. 2023;1:90-9. Available from: https://anapub.co.ke/journals/jrs/jrs_pdf/2023/jrs_volume01/JRS202301009.pdf
https://doi.org/10.53759/9852/JRS202301009
 
3 Alqahtani H, Kumar G. Machine learning for enhancing transportation security: A comprehensive analysis of electric and flying vehicle systems. Eng Appl Artif Intell. 2024;129:107667. Available from: https://www.sciencedirect.com/science/article/abs/pii/S0952197623018511
https://doi.org/10.1016/j.engappai.2023.107667
 
4 Ucar A, Karakose M, Kırımça N. Artificial intelligence for predictive maintenance applications: Key components, trustworthiness, and future trends. Appl Sci. 2024;14(2):898. Available from: https://www.mdpi.com/2076-3417/14/2/898
https://doi.org/10.3390/app14020898
 
5 Khan ZU, Nasim B, Rasheed Z. Generative AI-based predictive maintenance in aviation: A systematic literature review. 2024. Available from: https://www.researchsquare.com/article/rs-5277729/v1
https://doi.org/10.21203/rs.3.rs-5277729/v1
 
6 Batyha RM, Krishna MH, Kumar MS, Tiwari T, Mishra D, Ranjit PS. Predictive maintenance using machine learning in the aviation industry. In Recent Technological Advances in Engineering and Management 2024;(pp. 253-8). CRC Press. Available from: https://www.taylorfrancis.com/chapters/edit/10.1201/9781003531395-51/predictive-maintenance-using-machine-learning-aviation-industry-radwan-batyha-hari-krishna-sai-kumar-tripti-tiwari-deepti-mishra-ranjit
https://doi.org/10.1201/9781003531395-51
 
7 Ain QU, Jilani AAA, Butt NA, Ur Rehman S, Alhulayyil HA. Anomaly detection for aviation cyber-physical system: Opportunities and challenges. IEEE Access. 2024. Available from: https://ieeexplore.ieee.org/abstract/document/10749801
https://doi.org/10.1109/ACCESS.2024.3495519
 
8 Goriparthi RG. AI-driven predictive analytics for autonomous systems: A machine learning approach. Rev Intel Artif Med. 2024;15(1):843-79. Available from: https://redcrevistas.com/index.php/Revista/article/view/218
 
9 Chaudhry ZJ, Fox KL. Artificial intelligence applicability to air traffic management network operations. In 2020 Integrated Communications Navigation and Surveillance Conference (ICNS) 2020;(pp. 5A1-1). Available from: https://ieeexplore.ieee.org/abstract/document/9222889
https://doi.org/10.1109/ICNS50378.2020.9222889
 
10 Drishti. Application of AI in Predictive Maintenance of Vehicles. Analytics Vidhya. 2022. Available from: https://www.analyticsvidhya.com/blog/2022/07/application-of-ai-in- predictive-maintenance-of-vehicles/
 
11 Gülcü A. AI-enhanced cybersecurity solutions for protecting US aerospace manufacturing infrastructure: Techniques and real-world applications. J Bioinform Artif Intell. 2023;3(2):192-212. Available from: https://biotechjournal.org/index.php/jbai/article/view/113
 
12 Mızrak F. Integrating AI with aviation cybersecurity: The strategic role of human resources. Bus Manag Stud Int J. 2024;12(3):610-28. Available from: https://www.bmij.org/index.php/1/article/view/2396
https://doi.org/10.15295/bmij.v12i3.2396
 
13 Maaz A. Autopilot: A Complete Guide. Simple Flying. 2022. Available from: https://simpleflying.com/autopilot-complete-guide/
 
14 Jones T. Investigating the opportunities and limitations of artificial intelligence and cybersecurity in aviation. Master’s thesis, Eastern Michigan University. 2024. Available from: https://www.proquest.com/openview/39b84df3a74dd8c6cf2121f6614c9707/1?pq-origsite=gscholar&cbl=18750&diss=y
 
15 Vukadinovic D, Anderson D. X-ray baggage screening and artificial intelligence (AI). Publications Office of the European Union. 2022. Available from: https://www.researchgate.net/publication/361305570_X-ray_Baggage_Screening_and_Artificial_Intelligence_AI
 
16 Sutthithatip S, Perinpanayagam S, Aslam S. (Explainable) artificial intelligence in aerospace safety-critical systems. In 2022 IEEE Aerospace Conference (AERO) 2022;(pp. 1-12). Available from: https://ieeexplore.ieee.org/abstract/document/9843612
https://doi.org/10.1109/AERO53065.2022.9843612
 
17 Wiggerthale J, Reich C. Explainable machine learning in critical decision systems: Ensuring safe application and correctness. Artif Intell. 2024;5(4):2864-96. Available from: https://www.mdpi.com/2673-2688/5/4/138
https://doi.org/10.3390/ai5040138
 
18 Perez-Cerrolaza J, Abella J, Borg M, Donzella C, Cerquides J, Cazorla FJ, et al. Artificial intelligence for safety-critical systems in industrial and transportation domains: A survey. ACM Comput Surv. 2024;56(7):1-40. Available from: https://dl.acm.org/doi/full/10.1145/3626314
https://doi.org/10.1145/3626314
 
19 Reddy BR, Chokkalingam SP. Enhancing transportation security: Automated prohibited object detection for baggage inspection leveraging YOLO-NAS. In 2023 Seventh International Conference on Image Information Processing (ICIIP) 2023;(pp. 721-6). Available from: https://ieeexplore.ieee.org/abstract/document/10537641
https://doi.org/10.1109/ICIIP61524.2023.10537641
 
20 Philip JJ, Panda S, Meesala KR, Garad J, Bandhari O, Rani R, et al. Object detection in airport security checks. In 2024 International Conference on Advances in Computing Research on Science Engineering and Technology (ACROSET) 2024;(pp. 1-6). Available from: https://ieeexplore.ieee.org/abstract/document/10743507
https://doi.org/10.1109/ACROSET62108.2024.10743507
 
21 Kahan N. Enhancing security: Millimetre-wave imaging for concealed object detection and localization in passenger screening. 2024. Available from: https://www.researchgate.net/publication/387326665_ENHANCING_SECURITY_MILLIMETRE-WAVE_IMAGING_FOR_CONCEALED_OBJECT_DETECTION_AND_LOCALIZATION_IN_PASSENGER_SCREENING
 
22 Steno P, Alsadoon A, Prasad PWC, Al-Dala’in T, Alsadoon OH. A novel enhanced region proposal network and modified loss function: Threat object detection in secure screening using deep learning. J Supercomput. 2021;77:3840-69. Available from: https://link.springer.com/article/10.1007/s11227-020-03418-4
https://doi.org/10.1007/s11227-020-03418-4
 
23 Xu W. From automation to autonomy and autonomous vehicles: Challenges and opportunities for human-computer interaction. Interactions. 2020;28(1):48-53. Available from: https://dl.acm.org/doi/abs/10.1145/3434580
https://doi.org/10.1145/3434580
 
24 Bridgelall R, Tolliver D. Autonomous aircraft: Challenges and opportunities. 2023. Available from: https://rosap.ntl.bts.gov/view/dot/72546
 
25 Efe A. Assessment of the artificial intelligence and quantum computing in the smart management information systems. Bilişim Teknolojileri Derg. 2023;16(3):177-88. Available from: https://dergipark.org.tr/en/pub/gazibtd/article/1190670
https://doi.org/10.17671/gazibtd.1190670
 
26 Veeramachaneni V. Optimizing renewable energy integration in AI-driven data centres using quantum algorithms. J Netw Secure Data Mining. 2025;8(1):36-48. Available from: https://www.researchgate.net/publication/387262890_Optimizing_Renewable_Energy_Integration_in_AI-Driven_Data_Centers_Using_Quantum_Algorithms
 
27 Sood SK. A scientometric analysis of quantum driven innovations in intelligent transportation systems. Eng Appl Artif Intell. 2024;138:109258. Available from: https://www.sciencedirect.com/science/article/abs/pii/S0952197624014167
https://doi.org/10.1016/j.engappai.2024.109258
 
28 Rajkumar N, Viji C, Latha PM, Vennila VB, Shanmugam SK, Pillai NB. The power of AI, IoT, and advanced quantum based optical systems in smart cities. Opt Quantum Electron. 2024;56(3):450. Available from: https://link.springer.com/article/10.1007/s11082-023-06065-0
https://doi.org/10.1007/s11082-023-06065-0
 
29 Curtis TL. AI-driven avionics: Innovations in real-time data analysis for enhancing air travel experience. Aerosp Syst Technol J. 2023;17(2):310-25. Available from: https://aerospace-techjournal.com/article/view/25
 
30 Hendershot WE. Avionics systems of the future: Exploring the role of AI in safety-critical applications. J Aerosp Saf Technol. 2023;28(1):56-72. Available from: https://link.springer.com/article/10.1007/s00251-023-01478-7
 
31 Arafat MY, Pan S. Urban air mobility communications and networking: Recent advances, techniques, and challenges. Drones. 2024;8(12):702. Available from: https://www.researchgate.net/publication/386093820_Urban_Air_Mobility_Communications_and_Networking_Recent_Advances_Techniques_and_Challenges
https://doi.org/10.3390/drones8120702

Cite this article as:
Ahmed W. Artificial Intelligence in Aviation: A Review of Machine Learning and Deep Learning Applications for Enhanced Safety and Security. Premier Journal of Artificial Intelligence 2025;3:100013

Export to EndNote Export to Mendeley