Implementing Artificial Intelligence in the Aviation Network


Why does aviation need AI?

As technologies continue to advance, Artificial Intelligence (AI) will play bigger roles in our futures. This might be concerning to some, but it is a part of our lives. AI has already been incorporated in many industries; the aviation industry has already successfully implemented the early stage of automation in daily flight operations with flight management systems (FMS) or autopilot.

With the limited capacity of airspace and aerodromes, the aviation industry is faced with the challenge of accommodating surging air traffic demands. New kinds of flight operations, such as drones, commercial space travel and the personal air vehicles of the future, are making the problem even more difficult. The modern aviation system dynamically adapts and evolves on a regular basis to meet demands with real-time information sharing. But as the system rapidly increases its complexity, we are reaching our limit. Traditional automation cannot handle such ‘information explosions’. Human cognition and performance may need to be further augmented with the support of artificial intelligence.

How can AI contribute to aviation?

Safety is one of the key features in this industry. Aviation is the safest mode of transportation, given that the industry demands the highest level of safety standards and risk management. The validation process for machine intelligence had to be rigorous, so not all early attempts were successful. It had to be proven that AI operates safely; is interoperable with the current system; supports the human-centric system; and is applied through smooth and stable transition with globally harmonized manner. The key to achieving well-guided AI is to establish a human-machine coexisting environment, where the machine becomes a ‘sidekick’ that supports humans, instead of being a prospective ‘rival’.

While early automation was providing support with simple and repetitive tasks, today AI is expected to deliver further capabilities by learning and mimicking human behaviours. AI is carrying out human tasks and in certain cases, even out-performing them. This can be achieved from the data, which fuels AI. Whether it be for individual flights or systems, the aviation industry has gathered standardized data that creates and develops centralized, streamlined and globally accessible information exchange systems. The industry has already worked to build the cornerstone for AI, and with open data sharing, AI could change the direction of how we make decisions.

 Where are we and where are we heading?

In our previous AI article we illustrated an optimistic view of AI implementation in aviation. But we have many challenges to overcome. We do not fully understand what AI is capable of, and how AI will impact the industry. Concern is growing that AI would bring new opportunities as well as new problems, just as historically, other inventions have. Moreover, in considering the huge impact will have on our society, we need to work together to achieve common goals, and not represent the interests of individuals or corporations. The ‘positive use of AI for all’ is an emerging topic, with AI helping people lead enhanced lives and helping to accelerate the United Nations Sustainable Development Goals (SDGs).

Aviation plays an important role in connecting the world, we need to remember those pioneers and adventurers who proactively explored the new horizon with their innovative thoughts and actions.

About the author

Hyuntae Jung is an intern in ICAO’s Air Navigation Bureau.  With skills in Air Traffic Management and Machine Learning, his ambition is to bridge the gap between academia, industry and regulatory agencies, so he pursuing his studies in futuristic operation concepts and artificial intelligence.