Air France Enhances Operations with Cutting-Edge Artificial Intelligence
A History of Innovation
The airline’s journey with data and innovation dates back to 1958, when it established an operational research department to promote technological advancements. By the 1990s, Air France expanded its revenue management by utilizing historical data and trends. In the early 2000s, the airline developed Prognos, a predictive maintenance solution now used by over 80 airlines worldwide.
Integrating AI Across Operations
AI now plays a crucial role in various stages of the customer journey at Air France. From chatbots to predictive tools for baggage and meal counts, and eco-piloting systems for optimizing in-flight trajectories, AI helps optimize resources and anticipate needs, all aimed at enhancing the customer experience.
Generative AI: The New Frontier
Since 2023, Air France has been exploring generative AI, a type of AI that generates new data autonomously. Popularized by tools like ChatGPT, generative AI marks a significant technological breakthrough. Air France has launched over 80 projects involving generative AI across all business sectors, ranging from data management models to Proof of Concept stages.
Key AI Projects at Air France
-
TALIA: An internal ChatGPT tool that helps employees write emails, search PDF documents, organize events, and build to-do lists in a secure environment.
-
PAMELIA: A solution for airport agents to quickly respond to customer queries using an iPad, currently in the test phase and set to be deployed at Paris-Charles de Gaulle in 2025.
-
CHARLIE: A tool for maintenance teams to efficiently search for aircraft part numbers, aiding in timely repairs and replacements.
-
FOX: Analyzes customer feedback to understand expectations and concerns, capable of interpreting humor and irony, and detecting weak signals.
Principles Guiding AI Use
Air France’s approach to AI is guided by several key principles:
- Reasonable and Controlled Use: Prioritizing data protection, adhering to ethical principles, and complying with regulations.
- Promoting Existing Solutions: Supporting internal initiatives and guiding teams through AI projects.
- Market Solutions: Relying on existing AI models and solutions to maintain agility and control investment, rather than developing proprietary models.
Supporting Business Sectors
The operational research team, consisting of 150 employees, assists various business sectors in exploring AI possibilities. This includes guiding teams towards appropriate solutions, supporting Proof of Concept implementations, and participating in project rollouts.