KLM "Predicting Passenger Numbers with AI... Reducing In-flight Food Waste"

63% Decrease Over 3 Months

KLM Royal Dutch Airlines (hereinafter KLM) has reduced in-flight food waste by more than half by predicting the number of passengers using artificial intelligence (AI).

KLM announced on the 12th that it achieved this result by introducing the AI model 'Trace'.

Trace is an AI program developed last year exclusively for KLM catering services by the Dutch AI development company 'Kickstart AI'. Based on accumulated past data, it predicts the number of passengers in business, premium comfort, and economy classes for each flight and provides this information to the Meal On Board System (MOBS). This allows the system to analyze and determine the number of passengers from 17 days before departure up to 20 minutes before takeoff, loading only the necessary amount of in-flight meals.

After using Trace for three months, KLM reported a 63% reduction in food waste compared to preparing meals based on the number of booked passengers. In particular, on intercontinental flights departing from Schiphol, about 2.5 meals (approximately 1.3 kg) per flight were saved from disposal. This translates to an annual saving of 111,000 kg of in-flight meals.

Besides catering, KLM is utilizing AI in various fields such as aircraft maintenance, weather forecasting for flight schedule adjustments, and providing personalized travel information to individual customers. Marjan Rintel, CEO of KLM, said, "AI technology greatly contributes to improving flight operations and creating a sustainable flying environment," adding, "The introduction of Trace is a good example of how food waste, which is wasted annually, can be reduced by tens of thousands of times."