Urban air mobility – flying taxis – will revolutionize personal transport. This is a promise being made and also a massive challenge to the aviation industry. Whether autonomous or pilot-equipped the dramatic increase of vehicles in the sky requires a new and groundbreaking approach to controlling their motions. And nearly everyone agrees that the projected volume of traffic is simply too much for humans to effectively manage. Is Artificial Intelligence-based coordination and control the answer?
For Thales Flight Avionics Strategy Director Vincent Megaides the answer is a resounding yes. The densification of air traffic is “unprecedented” and cannot proceed without automation at multiple levels. Megaides sees an evolution of operations that integrates pilots, air traffic controllers and flight operations planners into a more tightly coordinated group. Their decisions are bolstered by better data and better data analysis. For Thales this is one of many visions for how AI will transform air travel. There’s just one problem: It almost certainly won’t happen fast enough.
The urban air mobility companies are talking about fleets of aircraft operational in 5-10 years. Early demonstrator vehicles are flying today. The timeframe in which these aircraft need to be integrated into existing airspace is rapidly approaching. Current controllers and technologies cannot handle the volume increase in that time. Megaides’s most optimistic forecast for AI taking over systems that are safety-critical is more than a decade away. How will the two sides of the industry bridge that gap?
We believe that intelligent assistants will be in service in the next decade. When we forecast how AI could reach the first certification level we are thinking about the 2030-2040 timeframe. For non-safety critical operations like airline operations centers it will come more quickly, used for optimizing things like slots, crew resources, aircraft available and passenger counts. but for safety-critical operations we don’t expect anything certifiable before 2030-2035. – Vincent Megaides, Strategy Director, Thales Flight Avionics Business Line
Getting AI-based ATC help today
Fortunately there are some AI-based advances that are in place today and others in advanced testing stages. Megaides describes these as Intelligent Assistants, support for pilots and controllers to help speed decision-making processes and improve the quality of choices. The assistant could be listening to the conversation between controllers and pilots, for example, while also processing weather data from sensors on the plane and other sources such as the IATA turbulence platform. Integrating the data and the conversations, with a dash of AI-based processing, could allow the Intelligent Assistant to recommend a deviation in course to improve the ride. That same work could be performed wholly by pilots and controllers today, but the extra data integration can avoid pilots requesting a new course or altitude that is known to have a conflicting flight path. That reduces workload on everyone involved and speeds the successful solution.
Another example is AIMEE, the AI-supported assistant being trialed at London’s Heathrow Airport. The use of ultra high definition cameras and digital processing software can identify when aircraft have exited the runway, allowing controllers to clear the next arrival, even if the tower is fogged in. NATS hopes to recover some 20% of capacity in low visibility operations once the system is fully operational. AIMEE also aims to coordinate controller and operations moves related to parking stand assignments, helping reduce bottlenecks on the taxiways or optimize ground handler efforts by processing more data more quickly than what either party could do on their own manually.
These smart assistants will evolve over time to handle more and more of the work. Eventually functions will become more automated, as is necessary for the continued growth of the industry. Taking that time is important both for regulatory adoption and for passenger comfort. It is also necessary to ensure that the systems function consistently and safely; losing the reputation that the air travel industry has earned through the decades would be a terrible step backwards. Getting the technology to a level that can be certified in the aviation space is not going to happen quickly.
That may be disappointing for some of the flying taxi companies as they struggle to grow with limited operations available, but it appears a necessarily pragmatic approach to ensuring that the systems integrate safely in to the global air traffic ecosystem rather than interfere with it.