By: Aditya Tangirala RIG Intern Researcher
After a long artificial-intelligence winter and spring is finally here. New applications of artificial intelligence have been sprouting for the past several years, and business interest in harnessing AI and machine learning in various workflows has been growing significantly. Among the various use cases of machine learning, one particularly noteworthy application is that of air-traffic control in airports.
Air traffic control in the United States has three primary functions, First, it’s used to prevent collisions between aircraft operating in an airport system. Second, it organizes and speeds up the flow of traffic to and from the airport. Finally, it provides additional support for U.S. national security and homeland defense (Chapter 2. General Control 2). As the world opens back up and people fly at pre-pandemic levels, the demand for higher-performance ATC systems is projected to increase significantly; Eurocontrol projects an 84% increase in global flights by 2040, justifying this demand (“Thales”).
There are various components of a successful ATC system, the most significant one being a radar in a central control tower, used to track and identify aircraft and other vehicles and objects on taxiways and runways (Searidge). These can be difficult to implement in certain conditions, such as in extraordinarily large airports with many runways to manage, or small regional airports without as much funding to build a dedicated control tower.
In recent years, AI has slowly been integrated into this aspect of air traffic control. For instance, Searidge Technologies of Canada has designed an “advanced neural network framework for the development of … [AI] based solutions for air traffic control and airport efficiency” (Searidge). This network harnesses camera feeds spread throughout the airport to track objects on runways and build an augmented-reality view of the entire airport, helping organize ground crews without the need for a direct sightline.
Another aspect of ATC, aside from ground crew that could benefit from AI and ML is scheduling takeoffs and landings. For instance, an airport without a central control tower could harness a machine-learning program integrated with cameras and aircraft-communication systems. These could be used to determine which planes should take off or land before the others depending on a variety of factors, such as the plane’s destination, fuel capacity, or gross weight.
Revolutionary Integration Group’s Dynamic Trust could be utilized in both aforementioned scenarios, as well as many other use cases related to air-traffic control. In the first scenario, by controlling and organizing ground crews, the Dynamic Trust could check and validate whether an outgoing plane has all its systems in order, from having food on the plane to checking its fuel levels. In the second scenario, RIG’s Dynamic Trust could verify those same criteria; ground crews could be alerted to guide an incoming plane toward a terminal or prepare an outgoing plane for takeoff.
With newer advances in artificial intelligence and machine learning come new opportunities to apply these technologies in a myriad use cases, and Revolutionary Integration Group is a key player in this increasingly exciting space. To learn more about the Dynamic Trust, schedule a demo with RIG at https://rigroup.co/dynamic-trust/.
Chapter 2. General Control. 7 June 2010, https://web.archive.org/web/20100607105632/http://www.faa.gov/air_traffic/publications/atpubs/ATC/atc0201.html.
Searidge-Technology-Success-Story-Us-75066-R5-Hr.Pdf. https://www.nvidia.com/content/dam/en-zz/Solutions/design-visualization/customer-stories/searidge-technology-success-story-us-75066-r5-hr.pdf. Accessed 22 Apr. 2021.
“Thales Is Using AI to Augment Air Traffic Management.” Aviation Today, 24 Jan. 2019, https://www.aviationtoday.com/2019/01/24/thales-using-ai-augment-atm/.