ORO Navigacija: Can AI replace air traffic controllers?

Picture of Vincent Lambercy
Posted by Vincent Lambercy

Editor's note: the version of this article in Lithuanian language triggered a lot of comments on linkedin, including requests for an English version. This is a translation by FoxATM and it has been reviewed by ORO Navigacija.

 

Aircraft and air traffic control centers use state-of-the-art technology to facilitate the work of pilots and air traffic controllers. Innovation has in the past led to a reduction in the number of pilots from 4 to 2 and the recent question is whether there will even be air traffic controllers and pilots in the future or if the aircraft will only be flown artificial intelligence.

An air traffic control radar display.

“As the number of flights increases, the desired airspace capacity must be ensured. Currently, airspace is used by different users - commercial airlines, general aviation, military, balloons, paragliders and drones. Naturally, in order to ensure orderly, safe air traffic and to meet the needs of all users, it is also necessary to ensure safe airspace management. Those performing these functions could also from artificial intelligence. Thinking about automated communication between air traffic control and aircraft systems is only the beginning. The air traffic control clearances now transmitted by radio, could be transmitted automatically by the air traffic control system. The pilot could decide to accept or reject it”says Tomas Tamašauskas, Head of Strategic Development and Innovation Department at Lithuanian State Enterprise Oro Navigacija.

Up to 40,000 flights take place in the airspace of the European Union on an intensive summer day. In Lithuania three-quarters of all flights are in transit. As of 2023, according to Air Navigation representatives, within the European Union, each aircraft will have its own 4D flight path object, consisting of three spatial dimensions and time as a fourth dimension, which will allow better route planning.

Artificial Intelligence and Aviation

Recently, a large number of users want to fly in the same part of the airspace, increasing traffic complexity but even so, airlines can’t fly as many flights as they want. "Artificial intelligence would allow more efficient use of airspace, save time, prevent traffic jams, but it requires a lot of testing before it can be properly utilized," explains Vaidotas Kondroška, ​​Head of Competence Center at Oro Navigacija. He emphasizes that Oro Navigacija is currently completing the introduction of a modern air traffic control system iTEC, which will, through certain automation functions, alert air traffic controllers well in advance of possible airborne convergence, thereby helping to prevent dangerous situations.

The automated functions would help not only pilots but also air traffic controllers.

Typically, the Area Control Centre ATCO operates about 40 to 50 aircraft per hour. Depending on the season, an aerodrome air traffic controller manages 115-150 aircraft daily. By increasing the airspace capacity, artificial intelligence, by performing certain technical functions, would assist air traffic controllers in servicing more aircraft. “Air traffic control systems need to ensure efficient air traffic control, so they are modernized or replaced every 12 to 15 years. In addition, airspace management is conducted by the air navigation service provider in each country in which the aircraft flies, so it is important that all providers use efficient air traffic management systems that are capable of efficient data exchange with each other. In order to achieve a good result the planning should be performed not locally but at the Regional basis, so the implementation of new systems and functions is subject to close coordination among all stakeholders of the Aviation system, i.e. airlines, air navigation services providers, etc., ”says V. Kondroška

Automation in Systems

Automation already assists in the processing of data as well as attempts to free the controller from formal, routine functions such as establishing and maintaining communication, etc. Because air navigation services are provided during each phase of the flight, even while the aircraft is on the ground, good communication between airport ground operations, pilots and aerodrome control centers is essential. Artificial intelligence can do this well. Nearly half of airports are expected to deploy artificial intelligence tools in the next decade.

Already today, countries such as the United Kingdom, Sweden, France, as well as the United States and others, use some forms of automation to help them in their daily work: the Approach Path Monitor (APM), a ground safety warning system trajectories, an innovative "Follow the Greens" guidance system to help you choose the optimal taxiway, and the Runway Incursion Warning System (ARIWS), which warns of an unauthorized intrusion into the runway

Artificial intelligence or human?

Consciousness distinguishes humans from robots, so the air traffic controller - the person making the final decisions - will not be replaced by any machine, say Oro Navigacija’s experts. “Certain concepts of artificial intelligence must help a person to perform technical work, provide information, offer solutions. In the future, the specifics of the work of an air traffic controller will change. He will have to do much less technical work, more monitoring and decisive decisions” says V. Kondroška, ​​Head of the Competence Center of SE Oro Navigacija.

He says that safety is the most important thing in aviation and the cost of the error is too high, so that all the latest technology is first thoroughly tested, which takes time. Taking into account all complex environment: the big amount of data related to the traffic flows, the number of possible interactions between traffic on pre-tactical and tactical phases, also the key performance indicators applied to the Traffic Network and its components, most probably that Artificial intelligence firstly will be applied in Air Traffic flow management  processes.

In addition, people lack confidence in fully automated solutions. A study by the Swiss Banking Union (UBS) found that as much as 61% people would not dare to fly knowing that their plane is not operated by a pilot even if the cost of the service is lower. Therefore, in order to achieve a change in the deployment of artificial intelligence, it is first necessary for passengers to have confidence in autonomous devices, and then it is quite possible that we will fly without pilots in the future.