London and Singapore test AI systems to assist air traffic controllers

Aviation authorities in London and Singapore are testing AI systems that could assist or potentially replace human air traffic controllers, following a series of high-profile accidents including the recent collision near Reagan National Airport that killed 67 people. The trials aim to determine whether artificial intelligence can reduce human error and improve safety in increasingly congested airspace, though experts remain divided on the risks and benefits of automation in air traffic control.

What you should know: AI-powered systems are being tested at major international airports to enhance air traffic control operations through continuous monitoring and early conflict detection.

• The U.K.’s NATS (National Air Traffic Services) is testing “Aimee” (Artificial Intelligence for Managing Integrated Environmental Elements) at London’s Heathrow Airport, which uses 360-degree panoramic vision to monitor multiple aircraft positions simultaneously.
• Singapore’s Changi Airport is also conducting trials with similar AI-assisted ground traffic management systems.
• These digital towers can provide high-fidelity views of entire airfield operations through arrays of fixed cameras, potentially eliminating the need for controllers to continuously scan in all directions.

The big picture: Current air traffic control systems rely heavily on decades-old technology and human oversight, creating potential vulnerabilities as air traffic increases and staffing shortages persist.

• Runway lights use technology from the 1980s, and some control towers still use paper to track aircraft movements.
• Controllers must balance flights in airspaces ranging from just a few cubic miles at busy airports to midflight sectors spanning more than 30,000 cubic miles.
• Most airline accidents occur during taxiing, takeoff, or landing phases when controllers face the most intense workload.

How it works: AI systems analyze multiple data sources to provide enhanced monitoring capabilities that exceed human limitations.

• “Once we digitize what controllers monitor, we can hand that data to an AI engine,” says Andy Taylor, chief solutions officer at NATS and former air traffic controller.
• The system processes live video feeds, ground environment data, and transcribed voice commands from pilots to monitor aircraft movements.
• AI can provide audible warnings to controllers about problematic plane movements and potentially alert pilots directly.

Collision avoidance improvements: MIT’s Lincoln Laboratory is developing ACAS X, an AI-informed upgrade to the current Traffic Alert and Collision Avoidance System (TCAS).

• The new system aims to reduce false alarms that occur when planes fly close together, which happens more frequently than when TCAS was designed in the 1980s.
• Unlike TCAS, which only directs planes to climb or descend, ACAS X can warn aircraft to move laterally in the sky.
• The system has been tested through millions of simulated near-miss scenarios.

What experts are saying: Aviation professionals express mixed views on AI’s role in air traffic control, citing both potential benefits and significant concerns.

• “The system can be trained to look for exactly the same things that a controller is looking for,” Taylor explains, noting AI’s ability to check runway clearances and scan tarmac areas in real time.
• “Automation is heralded as the solution, but it can actually make things worse,” warns John Leahy, former British Airways chief pilot and Royal Aeronautical Society member.
• “If you start depending on automation, you lower your guard,” cautions Shem Malmquist, a Boeing 777 pilot and Florida Institute of Technology instructor.

Key limitations: Experts highlight several risks associated with increased automation in air traffic control.

• AI currently lacks the creativity, intuition, and adaptability needed to handle emergencies that deviate from historical flight data.
• Over-reliance on automation could erode controllers’ and pilots’ ability to make quick decisions during critical situations.
• Increased digitization makes air traffic control systems more vulnerable to cybersecurity threats.
• Legal and ethical questions remain about accountability when AI systems are involved in accidents.

Why this matters: The aviation industry faces mounting pressure to modernize air traffic control systems as global air traffic increases and controller shortages worsen, but the transition to AI-assisted operations raises fundamental questions about safety, reliability, and human oversight in critical infrastructure systems.