Worldwide Customer Service & Service Center Network
Dassault Aviation
I’ve recently learned that one of the early uses of AI’s predictive capabilities might be to avoid causing harmful contrails.
This tidbit emerged from a fascinating conversation with Dr. Patrick Fabiani, head of Dassault’s roadmap for artificial intelligence. Patrick earned a PhD in Artificial Intelligence from ISAE-Supaero, one of France’s “grande écoles,” devoted to aerospace engineering. Later, at ONERA, France’s leading national aerospace research center, he studied drone and autonomous technology.
Patrick has been guiding the introduction of AI at Dassault since 2020—before many of us even heard of AI. He says the move to applied technology has been rewarding.
The rapid advance of AI
A brief overview of AI, which is relatively new to us all, may be worthwhile before we discuss flight deck and maintenance specifics. Things are moving fast. Patrick said that in 2005, it would take about 8-9 months to train a small neural network (a computer program that learns by example). Today, a “large language model” can be trained in a few days.
We briefly discussed military applications, such as linking Dassault’s new uncrewed combat air vehicle, announced at the Paris Air Show, which can act as a “loyal wingman” to the next-generation F5 Rafale.
In civil aviation, Patrick’s roadmap is to proceed incrementally, demonstrating benefits and allaying concerns about “super-intelligence” scenarios. As we learn more about AI capabilities, we want to identify and enjoy the benefits that simply help us do our jobs better.
AI and the environment
So, contrails first. Did you know (I didn’t) that two to eight percent of contrails contribute to nearly 100 percent of contrail-induced global warming? Most have a cooling effect by reflecting solar radiation during the day. But persistent night contrails can trap heat.
AI can help build models for predicting areas to avoid, sometimes by making modest adjustments to altitude or route changes, or by adjusting departure timing, if feasible for operators. Business jets, which operate on more flexible schedules and routings, can potentially contribute more than airlinersto avoiding contrail-induced global warming .
AI and maintenance
As you would expect, we’ll likely see more AI applications on the flight deck to help pilots, not replace them. In terms of product support, we expect significant benefits from AI. The reason for this is the increasing amount of maintenance data that must be managed. Already, an aircraft such as the 6X generates more data than a mechanic can efficiently study.
AI will be more likely to detect drifting parameters and identify the right time to alert a maintenance team to a potential failure. In part, the promise with AI is the ability to anticipate a maintenance requirement before an actual failure and address it directly . Maintenance can be performed precisely when needed, ideally during normal downtime, while helping avoid a potential AOG. So one benefit will be greater operational availability.
Such tools are already being used to maintain Dassault military aircraft. As these tools are refined, they’ll migrate to the civil side, as so much military DNA does.
We’re already running AI tests on past datasets and using AI tools with flight test teams before introducing them to fleet operators. We have an upcoming discussion with the OAB to ensure they are adequately briefed and to gather more customer input as we consider how to implement AI technology. The bottom line, as Patrick emphasizes, is to help humans make better decisions. As we consider how to implement these new tools, our objective will be straightforward: finding ways to improve support and client operations and delivering an even better experience for everyone.
Senior Vice President, Worldwide Falcon Customer Service & Service Center Network
Dassault Aviation
