Digitalization is the super trend in aviation along with the environmental sustainability. Many airlines, airports and other aviation stakeholders are talking about IoT, machine learning, augmented reality, virtual reality, robotics and other similar solutions. At the same time they are talking about big data and how to lead business with data. The goal is to get cost savings and operational improvements in various areas. But the honey pot in the end is the artificial intelligence (AI) in a wider perspective and everything it might be able to provide.
What is AI? It is a algorithmic tool that can learn and create suggestions based on data. AI based solutions have become a norm in the society and we are using them rather widely in our every day life. As the amount of data and the calculation capacity in our computers and smart devices have improved, the possibilities that can be gained through AI are improving as we speak.
Algorithmic leadership can lead to automatized management in the passenger services side. Forming AI related solutions at that area in aviation has been a trend for a long time already. Examples for these are personalized offerings, chat bots or localized services. These applications are similar to the customer service solutions in the other service industries like banking or retail. If these solutions don't work well, they just make the customer annoyed, but there is typically no safety risk involved. The fact that there is no safety risk involved has lead in many cases to a situation where the whole process is automatized. This gives cost savings, but if this is done poorly, it can certainly lead to a lousy passenger experience.
On top of the basic algorithms generative AI, like ChatGPT by Open AI or Copilot by Microsoft, can create new content based on the data with which it has been trained. The challenge is that training of the algorithm is based on data created by a human being and it can contain inconsistencies or unethical points. This can be very dangerous in aviation since we are talking about a safety oriented industry that can not tolerate safety deviations. Situational awareness, resilience and the ability to think alternative options are crucial.
In the operations side the aviation stakeholders like airlines, airports or air navigation service providers have a much more challenging situation compared to the passenger service side. The processes are complex, there is plenty of operative data and the situations are changing all the time. And the processes are certainly safety critical. In these kind of situations there is plenty that can be done with AI, but leaving the final decision to the algorithm can lead to a catastrophic result. Processes like optimizing the flight route plan for fuel consumption, vectoring flights for an ideal approach to the airport or creating dynamic aircraft parking plan can be good examples as processes that would benefit from AI. However, these processes might include unexpected circumstances that haven't been trained to the algorithm or the algorithm makes poor interpretations. This could cause severe safety challenges unless controlled by a human being.
Artificial intelligence has gained big hype in the media. Everybody is talking about AI and has some kind of an opinion about how it will change the world. This applies also to the business leaders. At the same time aviation industry is in a cost savings mode especially after the financially challenging years of pandemic. It would be tempting for a leader to order the transformation of the processes to AI based as widely as possible. However, the understanding of the potential of AI varies based on the background and ability to critical and risk analyzing thinking of the business leader. And, how big the urge to get cost savings is.
Leading AI in aviation is leading the change. Therefore the leader has to have discipline and ability to also understand the core functions of aviation. The transformation process has to be done in a good cooperation between the process owners and the AI solution designers. Processes have to be analyzed throughly since they probably have to be transformed and redesigned from many angles. Too often the risk analysis and creating various scenarios and levels for applying AI is forgotten. The end result has to be always effective, safe and resilient.
No matter how much we want the world to be fully digitalized, we still have passengers. Those people pay for their flights so they deserve good service. We also have physical aircrafts that fly in the air carrying passengers. No AI solution can be allowed to compromize this.