A Bibliometric Mapping of Airport Security: Research Trends, Thematic Structure, and Practical Implications

Every time a traveler steps into an airport, they enter a complex system designed to balance two competing needs: the absolute necessity of safety and the desire for a smooth, efficient journey. Behind the scenes of these security checkpoints, a quiet revolution has been taking place in how researchers understand and improve the process. For decades, the focus has been split between two distinct worlds. On one side, there is the human element: the training, psychology, and cognitive limits of the screeners who look at X-ray images to spot dangerous items. On the other side, there is the technological frontier: the rapid development of computers and software that can automatically recognize threats in baggage. While both sides aim to keep the skies safe, they have often operated in separate lanes, rarely speaking to one another. This separation raises a critical question for the future of travel: how do we best combine human skill with machine intelligence to create a security system that is both effective and efficient?

To answer this, a researcher named Tarık Mildan from Kocaeli University in Türkiye decided to map the entire landscape of academic research on airport security. Instead of reading every single study one by one, which would be impossible given the sheer volume of work, he used a method called bibliometric analysis. Think of this as a way to take a snapshot of a vast library and draw a map showing which books are most often read together, which authors talk to each other, and which topics are rising in popularity. By gathering 1,207 academic documents published between 2010 and early 2025 from a major database called Scopus, Mildan created a visual representation of the field's intellectual structure. He looked at who wrote the papers, what words they used most often, and how the topics changed over time. The goal was not just to count articles, but to understand the hidden patterns that dictate how knowledge in this field is built and where the gaps remain.

The map revealed a clear story of two parallel paths that have been running side by side. One path is deeply rooted in human psychology. At the center of this group is a researcher named Adrian Schwaninger, who leads a network of scholars focused on how people see and interpret images. These studies look at the mental processes of security screeners, how they are trained to spot threats, and how they interact with the machines they use. This tradition has been steady and consistent, building a strong foundation of knowledge about human performance and the cognitive challenges of looking at X-ray images for hours at a time. The other path is driven by engineering and computer science. This side of the map has exploded in recent years, particularly after 2019. It is dominated by research into deep learning, a type of artificial intelligence that allows computers to "see" and identify objects. The most frequent terms in this cluster include "deep learning," "object detection," and specific software tools like YOLO, which stands for You Only Look Once. This technology allows machines to scan baggage and flag potential threats in a fraction of a second, a capability that has become a standard in modern research.

What makes this map so valuable is that it shows how these two paths have largely ignored each other. The researchers studying human psychology rarely cite the papers written by the computer scientists, and vice versa. The data shows that while the human-factors group has a tight-knit community of collaborators, the technology group is a massive, rapidly growing field that has attracted thousands of citations. The most prolific authors in the technology sector are experts in computer vision and artificial intelligence, while the human-factors experts are psychologists and applied researchers. This split suggests that the field is currently divided into two separate conversations. One conversation is about how to train a person to be better at their job, and the other is about how to build a machine that can do the job for them. The study suggests that this separation might be a problem, as the future of airport security likely depends on integrating these two approaches rather than keeping them apart.

The timeline of the research tells another important story. For the first eight years of the study period, from 2010 to 2018, the number of publications grew at a slow and steady pace. However, starting in 2019, the output surged dramatically, reaching a peak in 2023 and 2024. This sudden jump coincides with the widespread adoption of advanced computer vision techniques. The surge was not just a general increase in interest; it was specifically driven by a flood of new studies on deep learning, X-ray imaging, and automated threat detection. The subject areas of these papers confirm this shift. The majority of the work now comes from engineering and computer science departments, with a significant portion also coming from mathematics and physics. While social sciences and psychology still contribute a substantial number of studies, their share has become smaller relative to the booming technical literature. This indicates that the field is becoming increasingly technical, with a heavy focus on building better algorithms and detection systems.

Despite the impressive growth in technology, the map also highlights what is missing. The research is heavily weighted toward the mechanics of detection: how to make the machines faster, more accurate, and better at spotting weapons or explosives. However, the experience of the passenger is barely mentioned. There is very little research connecting the security screening process to how travelers feel, how long they wait, or how satisfied they are with the service. Similarly, the broader policy and economic aspects of airport security are underexplored. While the technical side of the equation is being solved with great speed, the human side of the equation—how these systems fit into the daily lives of travelers and how they are regulated—remains a blind spot. The study suggests that future research needs to bridge these gaps, looking at how automated detection can be integrated with human training and how security measures can be designed to be both safe and passenger-friendly.

For the people who actually run airports and design security policies, these findings offer a clear guide. The persistence of the human-factors research tradition suggests that machines are not a replacement for human screeners, but rather a tool that must be used alongside them. The data indicates that even as we develop better artificial intelligence, we must continue to invest in training people, because trust between the screener and the machine is a complex psychological issue that cannot be assumed. Furthermore, the rapid pace of technological change means that the rules and regulations governing airport security need to keep up. If the technology is advancing every year, the certification processes for new equipment should not lag years behind. Finally, the study points out that the most effective security strategy will likely come from connecting the dots between risk-based screening programs, which treat passengers differently based on their risk level, and the new automated detection tools. By bringing these separate streams of research together, the aviation industry can move toward a system that is not only smarter but also more cohesive and effective for everyone involved.