Airplanes are “data factories” for artificial intelligence… a major revolution that does not include pilots technology

Artificial intelligence will in no way be a substitute for airline pilots, but this rapidly developing technology will be an assistant in improving the flow of operations at airports, customer services, and aircraft maintenance management.

For this purpose, airline heads from all over the world are gathering in the Emirati city of Dubai to attend the 80th annual general meeting of their main global association (IATA), to discuss the future of this sector, which is full of projects centered on artificial intelligence, some of which are under development, and others have already been implemented.

Accustomed to complex operations, incidents that require urgent handling and tight financial margins, companies in the aviation sector are looking for new reserves of productivity to increase their competitiveness, and big data processing algorithms can help them do this.

“Data and AI are great tools for the aviation sector,” says Julie Pozzi, head of data and artificial intelligence at Air France and KLM.

The Franco-Dutch group says it has “more than 40 projects using generative AI,” which, like the popular GBT chat, creates new data that feeds it.

Among these tools is a response tool in 85 different languages ​​to various customer requests, installed on touch-screen tablets for Air France customers and expected in 2025 at Charles de Gaulle Airport in Paris.

Also, the ADP Group, which operates this airport, has launched several initiatives using artificial intelligence, in cooperation with emerging companies.

One of these initiatives, “Alloprin”, relies on generative artificial intelligence based on voice recognition, and has allowed to reduce “the number of unanswered phone calls from 50 to 10%,” according to Alban Negrete, head of the group’s innovation department.

With Wintex software, which specializes in extracting data from real-time video surveillance images, ADP hopes to simplify queue-and-ride routes or shuttle bus rotations.

Reducing waiting times is one of the crucial challenges, explains Jerome Bouchard, a specialist in the aviation sector at Oliver Wyman.

“We are witnessing an increasing rise in the number of passengers, in increasingly restricted spaces, and we are still traveling as we did in the 1970s, and there is progress to be made,” says Bouchard, mentioning the development of “digital services” such as facial recognition when going through immigration.

Aircraft are data factories

He adds, “But all of this requires massive coordination and synchronization of data,” which is still incomplete: “We have not yet reached the use of artificial intelligence, but rather we are in the stage of structuring the value chain.”

For his part, Jeffrey Weston, head of airline activities at Bain & Company, points out that artificial intelligence “undoubtedly represents a new frontier, with the extraordinary acceleration of technologies and capabilities it provides.”

“When we are faced with dynamic situations, in the case of connecting flights, artificial intelligence is very useful in greatly accelerating the process of transmitting the right information to the right people,” says Weston.

Companies can also use AI “to manage their inventory or further automate supply chains, which is very important for maintenance in operational conditions,” according to Weston.

Air France-KLM plans to use a generative artificial intelligence tool among its technicians to speed up the process of identifying the faulty part out of thousands of parts.

For its part, ADP Group, in cooperation with the emerging company We Maintenance, has equipped its elevators and escalators with sensors supported by artificial intelligence to detect weak signs of malfunctions before they occur.

Modern aircraft, with their advanced self-diagnosis and electronic driving systems, are real data factories that can be used, enhanced by artificial intelligence, when planning future flights.

Equipment manufacturer Thales is also working to integrate this technology into its air traffic management tools.

But when it comes to actually flying the plane, there is no room for giving control to algorithms, because issuing permits for aircraft and their systems, the cornerstone of the aviation industry, must rely on elements that can be tracked and replicated, far from “fuzzy logic.”

Ultimately, “it is up to humans to take responsibility for the decision,” Thales CEO Patrice Kane explained last March.

He considered that the matter is “beyond just artificial intelligence. I would like to talk about assistive intelligence. It is intelligence that comes to help humans.”