Digitalisation in aviation is transforming the way the industry operates, and one area where it has made a particularly significant impact is in air traffic management (ATM). With the proliferation of data and technology, it’s crucial for ATM systems to be able to process and utilise this information effectively in order to improve efficiency and safety. Digitalisation in ATM will rely heavily in the use of automation and artificial intelligence (AI). Leveraging these technologies will enable air traffic controllers (ATCOs) to handle a larger volume of flights and make more efficient use of airspace (see our blog on how AI Digital Assistants for Air Traffic Controllers will revolutionize operational rooms). However, with all of these new data and technologies comes the risk of information overload.

Information overload is a common problem in the aviation industry, where pilots, ATCOs, and many other professionals in the sector are frequently faced with an overwhelming amount of data to process. This can lead to decreased productivity, errors, and, in worst case scenarios, even accidents. In this blog post, we’ll explore what information overload is, how the aviation industry is affected by it, and the role that AI solutions play in this issue.

So, what is information overload?

Essentially, it occurs when an individual (or organization) has more information to process than they can handle effectively and efficiently. In the aviation industry, information overload can occur at various stages and is sometimes referred as sensory overload, occurring when an individual is overwhelmed by the amount of sensory information they are receiving, such as visual, auditory, and kinesthetic stimuli. One example of information overload in aviation is when pilots are faced with too much data to process. These can include weather, route changes, and warning messages, among other things. If a pilot is trying to sift through all of this information while also trying to safely fly the aircraft, it can be easy for them to miss key information and/or make wrong decisions. Similarly, ATCOs can experience information overload while on the job. With several aircraft in the air at any given time, controllers must process and manage a large amount of information to safely coordinate operations. If complexity rises quickly in a given airspace, or if there are issues with communications or surveillance equipment, it can be easy for ATCOs to become overwhelmed.

So why is it important to address information overload in the aviation industry?

First and foremost, it’s a safety issue. When pilots and controllers are overwhelmed by too much information, they are more likely to make mistakes, which could have serious consequences. For example, in 2009, an Air France Airbus 330 crashed into the Atlantic Ocean, killing all 228 people on board. The investigation found that the pilots were confronted with multiple alarms and warning messages, which is believed to have contributed to their confusion and increased stress, ultimately leading to the accident. In addition to safety concerns, information overload can also lead to decreased productivity and efficiency. When professionals constantly try to keep up with a never-ending stream of data, they have less time and energy to focus on other important tasks.

So how can AI solutions help alleviate the issue of information overload in aviation?

One way is through the use of data analytics and visualisation techniques. These tools can help pilots and ATCOs more easily identify patterns and trends, and make more informed decisions based on the data. With these tools, they can quickly and easily see all of the relevant information they need in a clear, concise way. For example, an AI-powered dashboard could display real-time updates on weather, flight routes, and other important data in a way that is easy to understand and interpret. Another potential use of the these technologies is through the use of predictive and prescriptive analytics. These algorithms can also be used to predict and prevent potential problems, such as equipment failures or delays, by identifying early warning signs and alerting pilots and ATCOs to take action. However, the application of these technologies will always have to be studied with caution, especially so that they do not possibly worsen a problem they were intended to solve, or generate new ones of equal or worse consideration. For example, we must be careful that the use of AI does not introduce unwanted bias or additional complexity into the decision-making process of the intended users, as well as ensure that it does not create unwanted dependency on these systems. Over-reliance on AI solutions can create situations where if something goes wrong with the algorithms or the data they use, it could have significant consequences for operations. It is important to have adequate backup systems and contingency plans in place to ensure that operations can continue safely in the event of a problem.

What is our (Innaxis and DataBeacon) work in relation to information overload and AI?

Innaxis has coordinated and participated in several EU funded projects related to the use of AI in the field of ATM where issues such as user information overload have been addressed. In the SafeClouds project, which focuses on data-driven research addressing air safety intelligence, one of the case studies explored was the study and prediction of unstable approaches using mainly FDM data. In its development, different dashboard visualisations were built for airline safety officers to help them understand and explore the multiple parameters collected by the aircraft in unstable approach situations. In addition, predictive models were developed to theorise about the possibility of alerting pilots in real time about an unstable approach situation and its possible causes so that they could correct or perform a go-around to increase safety in the approach phase. The SafeOPS project focused its research on ‘from prediction to decision’, a common decision-making paradigm in digitisation and predictive analytics. It investigated how predictions based on AI models affect the decision making process of an ATCO. For this purpose, a use case related to go-around prediction was studied where following a human-centric design methodology. A prototype tool was developed where, among other things, the possible information overload of the controller was taken into account. In DataBeacon’s Victor5 solution for a digital assistant platform for air traffic control operations, we also take a human-centric approach to integrating AI; special attention is paid to the design of UXs to deliver flexible and modern UIs for any aviation use case, operational scenarios or operational staff requirements. All of this helps reduce possible information overload for the user. Finally, Innaxis, DataBeacon and other partners have recently been funded in Horizon Europe, a part of the SafeTeam project. The main focus of this project is on the human factors aspects of the use of digital aviation assistants, including a deeper understanding of the technology and processes that will facilitate the adoption of AI tools and their integration into operations, enhancing human cognitive capabilities and potentially automation.

Overall, while information overload is a common and serious problem in the aviation industry, there are ways that AI and ML solutions can help mitigate its effects. By using data analytics, visualisation tools, predictive and prescriptive models, we can help pilots and ATCOs stay on top of all of the information they need to do their jobs safely and effectively, while also reducing the risk of errors. I hope you have enjoyed the blog and don’t forget to visit https://datascience.aero/ for more interesting posts!