Dassault Aviation and Artificial Intelligence

Digital technology is at the core of our innovations, our transformation and our work organization. Our capacity as an industrial architect draws on all leading digital solutions, from 3D creation to Big Data.

A pioneer in digital technology and AI

Digital technology is at the core of our innovations, our transformation and our work organization. Our capacity as an industrial architect draws on all leading digital solutions, from 3D creation to Big Data. As pioneers of a new industrial revolution, we are the crucible from which Dassault Systèmes emerged, and we have a long-standing collaboration with the latter for close to 40 years. These close ties to the world leader in Product Lifecycle Management solutions give us experience and processes for adapting these advanced technologies to our industrial operations. The Falcon 7X, for example, was the first plane in the history of aviation to be developed entirely using a digital model, from design, to manufacturing, to maintenance.

Dassault Aviation has been integrating intelligent systems into its aircraft for some 30 years:

  • 1990s: development of “electronic copilot” for two-seat fighter aircraft;
  • 2000s: data fusion on Rafale;
  • 2010s: nEUROn combat drone, UCAV and MALE projects.
  • 2020s: predictive maintenance, new Rafale standards, New Generation Fighter.

In the future, intelligent systems will obviously be used increasingly in aviation, whether as ground assistance and maintenance systems, as embedded systems onboard manned aircraft, or as unmanned systems such as drones.

An AI philosophy

These systems will still be controlled by human beings, even when no one is on board. The nature of military action and the aeronautics context require the design of systems over which the operator retains control. With its capacity for adaptation, initiative and anticipation, only the human brain is able to provide the flexibility needed for conducting complex missions. Our use of artificial intelligence is designed to leverage our human capabilities through a complementary approach. Take the example of a combat drone. Its operation necessitates control over two major functions: flying the aircraft, which requires rapid automatic actions, and conducting the mission, which requires reasoned medium/long-term decisions. With AI, humans will be freed from the “flying the aircraft” function but will remain crucial to the “conducting the mission” function. In fact, today already, aircraft pilots give orders to the flight controls which then calculate the optimum configuration for the flight control surfaces. The same will hold true for future aircraft: they will be more autonomous thanks to AI, but remain under human control.

Towards French and European AI

The above challenges are considerable and particularly sensitive when it comes to military aviation. That’s why Dassault Aviation is investing in an advanced, sovereign information system with French or European providers and platforms. That’s why Dassault Aviation is a signatory of the Manifesto for Artificial Intelligence dedicated to industry. As an architect of aeronautical solutions working in partnership with our industrial partners and our government customers, we are building the collaborative combat systems of the future.

AI projects at Dassault Aviation

In terms of AI, Dassault Aviation focuses on concrete, targeted projects that are intended to lead to operational aeronautical applications. In addition to our own studies, we collaborate with startups and partners such as the Institute for technological Research (IRT) SystemX and IRT Saint-Exupéry on certain generic research programs.

Artificial Intelligence put to the test with aeronautical data

The data available to the aircraft manufacturer is wide-ranging in its sources and nature.

Several unique challenges are looming such as protecting military and commercial aircraft data, analyzing and learning from the time series data of thousands of flights, and implementing a reinforcement learning method.

One example of a challenge is the analysis of a set of flight parameters over an aircraft’s entire lifecycle in order to identify failures and implement predictive maintenance. More than 100 time series with an average of 200,000 data points per flight must be analyzed for each aircraft. Extracting the quintessence of this time information is one of the complex tasks faced by our data scientists.


To tackle it, our data scientists have developed cutting-edge knowledge in these fields to resolve the issues. This knowledge and our know-how as an aircraft manufacturer are shared with our ecosystem of partners through hackathons, challenge events and preliminary study programs. Algorithms are then custom-developed to implement products and services dedicated to our teams and customers.