Today, AI is making its mark worldwide in various industries, especially in the aviation market. The global aviation industry is progressing at an unprecedented pace and air traffic is booming day by day. According to IATA (International Air Transport Association), the number of air passengers was 3.8 billion in 2016 that is projected to reach 7.2 billion by 2035. Hence, the passengers’ counts are projected to double within the coming few years. The dramatic increase in air traffic demands a groundbreaking approach to ensure efficiency and air transportation safety. Thousands of planes are flying in our skies all the time, and they are at high risk of safety due to ever-increasing congestion. The aviation industry needs to ensure the secure operations of air traffic management systems. AI is already revolutionizing the way air traffic management systems are manufactured and hence is believed to play a key role in optimizing air traffic flow.

Air Traffic Control (ATC) and its existing challenges

Air traffic control is extremely complex and a safety-critical process of decision-making in a stochastic and highly dynamic environment. With the fast-growing and high-density global air traffic, ensuring efficiency and air transportation safety becomes a critical challenge. Tactical ATC decisions are still being made by humans as it was around fifty years ago. A human air traffic controller has the responsibility to monitor and direct many aircraft and to ensure a safe distance between aircraft. He also has to adjust the trajectory of each aircraft in such a way that a minimum distance between the aircraft is maintained all the time. The objective behind this is to avoid the “loss of separation” (LoS) and to assist the aircraft to reach safely to their defined destinations. At the same time, the controllers also need to keep up with backup plans to ensure safety in case of some unexpected event. Further, there are other areas of interest such as orderly transfer of aircraft between sectors, environmental improvements and fuel efficiency, optimization of landing sequence, and many more.  These elements can be optimized by focusing on the potential future of each flight, predicting possible conflicts, and providing timely instructions. Some sources of uncertainty also need to be considered like individual pilot’s behavior and changes in weather conditions. A decision-making system in ATC has to consider these uncertainties to optimize the air traffic flow and ensure a safe flight for the passengers.

Therefore, the development of an ultimate, safe, and reliable air traffic control system has become a critical challenge due to increased complexities in the airspace. Unfortunately, the existing air traffic control systems are incapable to accurately foresee the traffic evolution and use tactical interventions to ensure the safety of flights. That is why; the planes are habitually penalized in real traffic conditions because of redundant safety standards. However, on account of severe congestion in most significant terminal areas, this management criterion is no longer valid. Air traffic is continuously increasing at an exponential rate leading to flight delays and pollution due to significant congestion. Moreover, stratospheric balloons and drones are the new air vehicles flying in high or low-altitude airspace. This high-density air traffic has increased the complexity of managing airspace.

Keeping in view the existing and forthcoming challenges, and the growing complexity in low-altitude and traditional airspace, an autonomous system using an AI-powered platform is required to optimize ultra-dense air traffic and secure a safe distance between aircraft. Heinz Erzberger and his colleagues from NASA presented the Advanced Airspace Concept (AAC) for the first time to increase airspace capacity and safety, hence proposed an Autonomous Air Traffic Control system. They designed automation tools such as TSAFE and Autoresolver to augment human controllers in case of any conflict resolution. TSAFE (Tactical Separation-Assisted Flight Environment) predicts imminent conflicts like loss of separation, hence alerts the controller while Autoresolver attempts to resolve the predicted conflicts. Hence, AI platforms are needed to ensure adequate safety levels and plan functions to improve the capacity and efficiency of the system.

AI-powered platforms to Optimize Air Traffic

AI is one of the top drivers of technology at present and is influencing the aviation industry as well. AI applications are being applied to the flight planning environment of Airlines and can help address the existing challenges regarding capacity, connectivity, forecasting, pollution, delay management, and safety. Several technology innovators have designed and introduced AI-based platforms to the aviation industry to cope up with the existing and upcoming challenges as a result of increased congestion in the airspace. Fortunately, some AI-based advances are in use today and others are in testing stages. Some of them are highlighted below:

Searidge’s AI platform (Aimee) – Searidge Technologies with over thirty-five sites in twenty countries is recognized as a global leader and the most preferred partner for digital airport solutions and Remote Tower. It is a technology innovator that improves efficiency and safety in the aviation industry.

Seardige AI platform named “Aimee” bears the ability to simplify the procedure of configuring and training ANNs (Artificial Neural Networks) with huge and complex data sets. Hence, AI makes possible the aircraft’s image and geolocation tracking, ensuring the availability of runway for the next aircraft to land in sequence. It is already being used at London’s Heathrow International airport to replace the runway views of air traffic controllers when the clouds and fog make the tower’s vision unclear. At this airport, the world’s earliest 4K digital tower lab was established by Searidge. The objective behind this lab was to recover the landing capacity loss that was around 20% because of low visibility conditions. AI plays its role when inclement weather obstructs the runway view for a controller from the 100 feet tall tower. Twenty zoomable cameras are installed on the north and south sides of the tower to provide runway coverage. The cameras’ images are then fed into Aimee. Aimee, after interpreting the images, and tracking the aircraft, informs the controller once it has cleared the runway successfully that helps the controller in making further decisions. According to Ruckert, the radar’s update rate is 1 Hz while that of the visual system is 25 Hz, so in case of triangulation, the radar source is sometimes behind or a little inaccurate. So radar is used as the initial source, while AI is used to correct.

According to the head of innovation at Searidgetechnologies, Mr. Marco Ruckert, their objective is not replacing human air traffic controllers rather enhancing his capabilities by providing situational awareness and the tools to support their decisions.  Hence, by taking away the mundane tasks, the human can focus on just complex decisions and can do their job more efficiently.

Thales AI-based platform

Thales – a frontrunner in air traffic management (ATM) solutions, delivers safer airspace management systems with an AI-enabled platform. It exists in 85 locations worldwide and provides the controllers with a clearer prediction of trajectories, air traffic flow, and estimated times of flight take-off and arrival. Thales believes in predictive maintenance, therefore, it has a paradigm shift from corrective to predictive maintenance. Rather than repairing defects, it ensures to avoid downtime. Thales AI algorithms, using smart sensors to capture operational data, can predict upcoming problems, assisting planners to make correct maintenance decisions at the right time. Thales also helps airlines to offer the passengers a seamless experience. AI and Big Data are used to keep a record of passengers’ preferences and offer them personalized recommendations including endless inflight entertainment and targeted advertisement along with shopping.

Flyways AI platform

Flyways AI, developed by Airspace Intelligence, is an innovative technology that involves 4D mapping to make the flight more efficient.It provides an AI-based predictive and recommendation platform for commercial air operations. It has billed itself as the world’s first AI-powered platform for flight monitoring and routing. Just like the apps that the drivers use on the roads, it assists the dispatchers in navigating the skies and optimizing the routes.

American airline Alaska has recently announced a partnership with Airspace Intelligence and signed a multi-year contract for its Flyways AI platform. The prevailing airline computer systems are incapable to pile up all information and changing conditions into one cohesive map. Contrary, Flyways can predict forthcoming situations and manage exceptions by processing huge data inputs quickly and more accurately.

The system of Flyways AI, autonomously evaluates the compliance of Air Traffic Control (ATC), operational safety, and efficiency of planned and active flights of the airline. When it finds a more efficient route around turbulence, it furnishes the flight dispatchers with actionable recommendations. Then the dispatcher makes the decisions whether to implement the recommended solution. Flyways AI assists the dispatchers to perform smartly by proposing the most efficient and safest routes to the pilots. It also assists in streamlining traffic flow hence reducing greenhouse gas emission (GHG) and fuel burn.

Pasha Saleh, director of flight operations of Alaska Airlines says that implementing the Flyways AI system in just six months resulted in the saving of 480,000 fuel gallons along with the reduced carbon emission of around 4600 tons.

Skywise AI platform

Skywise is an AI-based platform developed to collect data about aircraft operations. Airbus is using this platform to perform observation tasks such as computer vision, natural language processing, time series analysis, and to make predictions and decision making. Airbus is using vision-based AI and intelligent navigation. According to the Airbus head advisor on AI, Romaric Redon, Skywise helped Airbus during the COVID-19 pandemic by supporting them to analyze air traffic fluctuations and flight restrictions.

Artificial Intelligence (AI) leading the way to Optimize Air Traffic

The aviation industry, at present, needs to find ways to optimize air traffic and AI can be very influential in this endeavor. The European Aviation High-Level Group on AI has recently published a report considering the benefits of employing AI in air traffic management stating that AI can help optimize air traffic by providing critical decision-making tools and improving operational efficiency.

At present, many air traffic management (ATM) systems and airports are adopting AI technologies to optimize air traffic. Eurocontrol, in 2019, organized an inaugural forum regarding AI and aviation. In March 2020, they published their first report named “Fly AI”.In the report, they have outlined how all the stakeholders including ANSPs (air navigation service providers), airports, airlines, and others see the likely use of AI to optimize the aviation data leading to more sophisticated tools and more accurate predictions. As per the report, some ATM applications for AI are ATC planning and flow management. Eurocontrol had already performed several trials concluding a 30 percent upgrading in trajectory predictions. As per Eurocontrol, AI also has the potential to help developing surveillance technology for drones that are operated on a large scale beyond visual sightlines in the airspace.

AI is being infused into the workflows at many airports. It can be used to simultaneously monitor diverse areas of concern across an airport that the humans are incapable of doing physically, such as runway exit points. The objective is to focus on making the turnaround process of an aircraft more efficient for airlines as well as airport workers. This focus on AI does not reduce the significance of humans in the process, rather supports optimizing human performance. AI is also used to predict the number of passengers an airline can have in a given term. The OptiScorer engine developed by OptiWisdom that contains artificial intelligence analyzes the historical passenger data and projects the number of passengers of an airline in the coming months.

The ATM application of AI is using ML and neural networks and is believed to improve tracking and detection competency in image processing. Neural networks, in the first phase, analyze datasets to “train” and develop an understanding of the normal operations. In the next phase, followed by training, the marginal data is highlighted in “anomaly detection”. This capability to detect operational events beyond normal parameters is the main distinguisher between the traditional system and machine learning. Hence, by using AI in ATM, the time between development and operational deployment is reduced from years to months.

For air traffic controllers, the most crucial task assigned is to avoid collisions and to plan the flight routes ensuring on-time flight take-off and arrivals at its predefined destination. These challenging tasks are error-prone and time-consuming being fully dependent on human decisions. The advanced AI-based tools help the air traffic controllers to prevent aircraft collisions. These tools support the controllers in making quick decisions based on the historical data, hence, save time and ensure safe and on-time flight operations.

AI also helps in reducing the impact of exterior factors like weather and supports the air traffic controller in making quick and right decisions at the right time. Hence AI can enhance the operational capability, as well as safety. In case of extreme weather conditions or heavy fog, there is very low visibility at the airport. The air traffic controllers have to rely on the radar to make sure that the arriving airplane has left the runway or not. That is a time-consuming process affecting airfield punctuality. To avoid this, several AI-based automated solutions are used to furnish controllers with more accurate and authentic information earlier, improving the landing efficiency at the airport.

AI can also be used to ensure smooth communication between controllers and pilots. At present, the primary communication between both is through the radio channel that is too hard to hear especially when several different accents are involved as in Scotland alone, there are 50 different accents. AI, with an automatic speech recognition system, assists the controllers by making their jobs easier.

Maintenance of an aircraft is a tough task that can cost a fortune if done incorrectly. Today, several airlines are shifting from corrective maintenance to predictive maintenance. Thanks to AI, airline companies can predict potential maintenance failures before they happen in actual. AI can use the data from in-service aircraft and can predict the potential issues. AI helps to trigger real-time repairs and hence reduces the requirement of routine maintenance.

By using AI, airline companies can also optimize their labor and operational costs. AI can also increase the fleet availability of airliners by up to 35% and decrease the labor cost by almost 10%. AI-powered tools can predict future flights, and resolve customer queries. It can predict airplane faults and flight delays and can better facilitate both airliners and airports by avoiding critical issues that could disturb air traffic, revenue, and customer satisfaction.

So the researchers believe that AI can enable air traffic controllers to cope with the anticipated growth in air traffic and the complexity of incorporating new vehicles. It will ensure secure, safe, and highly efficient airspace management, with improved predictive capabilities, resulting in fewer delays and less gas emission due to less burned fuel.

Summary

We are living in an era of accelerating change. With every passing day, there is increasing pressure on aviation industry to introduce the new systems safely. In the future of air traffic management, AI technology is projected to play a great role by making people free from monotonous tasks and allowing them to focus on decision-making. Researchers believe that AI will complement and augment human competencies by reducing their involvement in low-value and repetitive tasks, so they will be free to concentrate on more critical and complex tasks where the human intrusion is crucial.

By employing AI techniques in the aviation industry, all the stakeholders involved in air safety – civil aviation authorities, ANSPs (air navigation service providers) airlines, and airports will enjoy improved operations and reduced operating costs. In recent years, AI has introduced new control concepts, optimization models, and algorithms to optimize air traffic. Airlines companies keep experimenting with AI to analyze how it can help optimize air traffic by ensuring faster, safer, and more convenient flights. However, these experiments are at the initial phases right now, but they believe that AI has the potential to reshape the aviation industry in the future. AI-based routing and flight monitoring platforms are being employed to assist in critical decisions and optimizing air traffic flow. AI and ML techniques assist dispatchers in optimizing the routes hence resulting in more sustainable and efficient flight operations by improving the predictability and airline traffic flow. It helps the flight operations to plan, monitor, and also offers recommendations to reroute flights so that the bad weather and congested airspace issues are avoided. It also saves time and carbon emissions and mitigates congestion resulting in a delightful experience for the passengers, hence, AI has the potential to transform travel from ordinary to extraordinary and is a definitive solution to optimize the complex, dynamic and ultra-dense air traffic.