Artificial intelligence progressions have the potential to alleviate bottlenecks in highly capacity-constrained switch security. Industry experts anticipate significant advantages from these advancements, along with certain challenges that may arise. In an article initially featured in the September 2023 edition of Passenger Terminal World magazine, Alex Grant explores five key applications.
As global economies rebound from the disruptions caused by Covid-19, there is a surge in air travel demand. The outlook is optimistic, with the ICAO Council projecting a return to pre-pandemic customer volumes by 2023 and an estimated increase in annual demand from four billion passengers in 2019 to eight billion by 2040. This growth adds pressure to already strained airport capacities.
Developers are exploring lucrative commercial applications, leading to increased investment in AI-based solutions as a means to alleviate these pressures. Investors recognize that the current excitement surrounding AI is not just a passing trend. Financial services company UBS forecasts substantial growth in the industry, from \(28 billion (€22 billion) in 2022 to \)300 billion (203.6 billion) by 2036, positioning AI as one of the fastest-growing segments in the technology sector.
It is crucial to note that AI is not a distant technology; organizations often underestimate its current role, as highlighted by independent safety and security consultant Donald Zoufal from CrowZnest Consulting. While AI is commonly used for tasks like license plate recognition for traffic and parking surveillance, its potential extends to streamlining security screening processes and enhancing the overall travel experience for customers.
Zoufal mentions that terminals and airlines are testing genetic solutions to advance security measures, particularly focusing on bag-drop checkpoints. Understanding and utilizing AI technologies effectively is essential, as AI has been integrated into various operations for many years. Transparency and comprehension of these AI systems are key to their successful implementation.
Efficient Personality Assessments
AI plays a significant role in streamlining passenger control processes, exemplified by Dubai International Airport’s implementation of 122 smart gates in 2019. These gates allow travelers to pre-register for expedited procedures using iris recognition, reducing passport control time to just five seconds. Despite a 161.9% increase in passenger numbers during the first half of 2022 (reaching 27.9 million), 97% of passengers experienced wait times of fewer than five minutes, with 96% waiting for under three minutes.
IATA’s One Card program emphasizes early passenger information sharing and biometric identification for seamless terminal processing. While AI is not a prerequisite for flights, its integration can offer advantages. Matthew Vaughan, chairman of aviation security and cyber operations at IATA, highlights the shared goal of airports and governments to streamline testing and facilitation processes.
Facial recognition solutions based on AI, however, must address concerns regarding algorithmic bias and discrimination, as highlighted by Zoufal. Machine learning algorithms heavily rely on available data, potentially posing challenges with minority groups due to limited data availability. Additional strategies may be required to accommodate individuals not well-represented in the dataset.
Zoufal notes, “Algorithmic bias is inevitable, akin to human decision-making biases.” Evaluating the error rates based on machine bias is distinct from human error rates, suggesting that machines may exhibit lower failure rates compared to humans in certain scenarios.
SmartGates in Dubai offer document-free boarding. Enhanced Baggage Screening
While baggage screening is a critical security measure, it can be time-consuming due to the extensive list of prohibited items and the utilization of two-dimensional scanning techniques. Despite the efficiency improvements brought by CT systems, such as not requiring the removal of electronic devices and liquids before scanning, challenges persist.
Alan Tan, Senior Vice President of Port Protection and AVSEC at Changi Airport, highlights the benefits of automated prohibited-item detection systems (APIDS) in expediting checkpoint clearance and image analysis. The transition from manual to automated detection can enhance the speed and accuracy of threat identification, ultimately improving security protocols.
Regulatory bodies are paving the way for broader APIDS implementation, with the EU approving three APIDS algorithm standards and ECAC establishing screening guidelines for fully automated systems. Airports like Schiphol are conducting tests using the Project Dartmouth engine, powered by Google Cloud and Pangiam, to enhance security screening capabilities.
Tan suggests that further integration of machine learning can refine APIDS functionality by reducing false alarms and expanding the system’s recognition capabilities. Training APIDS to distinguish between threat items and benign objects can optimize security screening processes and minimize unnecessary rechecks.
Tan explains, “Expanding the list of recognized benign items through machine learning can reduce false alarms, allowing for quicker processing of baggage without compromising security standards. Additionally, machine learning can aid in identifying emerging threats by leveraging a comprehensive database of known risks and benign items.”
AI collaboration between Schiphol and Pangiam enhances security screening. Optimized Working Conditions
The implementation of AI solutions raises questions about its impact on the workforce. According to a report by the UK’s Organization for Economic Co-operation and Development (OECD), 27% of jobs, including those in banking and manufacturing, are at risk of automation, including AI. However, a survey conducted in 2022 revealed that over half of respondents reported improved emotional well-being, with 63% expressing greater job satisfaction since adopting AI technologies.
Billy Shallow, Senior Director of Stability, Innovation, and Engineering at ACI, anticipates similar benefits for security personnel with the deployment of advanced screening solutions. APIDS algorithms can serve as additional analytical tools, potentially automating image analysis tasks in the future, allowing human operators to focus on critical decision-making processes.
This shift can alleviate the burden on security screeners, enabling them to concentrate on threat assessment and resolution rather than mundane tasks. The integration of APIDS into security operations requires careful consideration of the interface between AI systems and human operators to ensure seamless adoption and global consistency.
Tan from Changi Airport emphasizes the importance of safety monitoring applications in addressing the rising demand for air travel amidst ongoing manpower challenges. Integrated testing solutions can enhance operational efficiency and security protocols, enabling operators to respond swiftly to emerging threats while maintaining a high level of passenger safety.
AI has the potential to streamline security personnel’s workload. Enhanced Surveillance at the Terminal
Christopher McLaughlin, Executive Vice President of Operations at Dallas-Fort Worth International Airport, envisions significant potential for AI applications at one of the world’s busiest airports. Beyond passenger and luggage handling, AI technologies can enhance risk monitoring within the terminal, identifying potential threats to visitor safety.
McLaughlin asserts, “Our focus is on control.” Dallas-Fort Worth International Airport is actively developing several AI/ML-driven projects aimed at enhancing surveillance capabilities and passenger experience. These projects encompass both covert security measures and customer-facing initiatives, aiming to address security challenges effectively.
Schiphol in the Netherlands utilizes AI for belt monitoring, employing the Strong Turn system to predict delays and streamline turnaround processes based on real-time and historical data. This system not only enhances operational efficiency but also serves as the foundation for tracking passengers’ luggage from aircraft to baggage claim areas.
IATA’s Vaughan highlights perimeter breach detection programs, citing recent airport protests as a driving factor for enhanced security measures. Various vendors and technology firms offer solutions for perimeter surveillance, staff access control, and other security-related tasks to bolster airport security protocols.
Improved Passenger Flow
AI technologies offer unprecedented capabilities in analyzing passenger movement within terminals, particularly at security checkpoints. Dallas-Fort Worth introduced line-wait technology in 2018, utilizing 3D optical sensors to track passenger activity, streamline security procedures, and provide real-time updates on wait times.
McLaughlin emphasizes the use of AI/ML to empower passengers with travel information, enhancing their overall airport experience. By leveraging AI-driven technologies, airports can mitigate stress associated with uncertainty during travel, providing passengers with real-time updates and greater control over their journey.
Furthermore, AI/ML applications extend to areas that may not be immediately apparent to passengers, such as exit-lane technologies that enhance security and streamline passenger flow. AeroCloud Optic, a computer vision-based system, enables airports to monitor passenger traffic discreetly, anticipate security checkpoint bottlenecks, and expedite passenger movement to commercial areas, benefitting both travelers and businesses.
