KLM implements AI to cut food waste and enhance sustainability
KLM has introduced an AI model named TRAYS to reduce food waste by predicting passenger no-shows. This initiative aims to improve sustainability by aligning meal provisions with actual passenger counts.
This innovative system is designed to predict the number of passengers who will not show up for their flights, thereby reducing the amount of food waste generated by uneaten meals. On average, between 3% and 5% of KLM’s passengers do not turn up for their flights, resulting in over 100,000 kg of food waste annually.
Historical data
According to a report by Business Traveler USA, TRAYS leverages historical data to forecast passenger numbers and adjust meal quantities accordingly. The AI model, which starts making predictions 17 days before a flight and updates them until 20 minutes before departure, was trialed for three months, resulting in a remarkable 63% reduction in food waste. This initiative is part of KLM’s broader strategy to enhance its sustainability efforts through digital technology and data analytics.
Predicting passenger numbers
The TRAYS system operates within the ‘Meals on Board System’ (MOBS), catering to three flight classes: Business, Premium Comfort, and Economy. By accurately predicting passenger numbers, TRAYS helps in preparing and loading the exact number of meals needed, thus minimizing wastage. Marjan Rintel, CEO of KLM, emphasized the importance of digital technology in improving flight operations and sustainability.
Accelerate AI adoption
Business Traveler USA reports that TRAYS was developed by Kickstart AI, a collaboration among several Dutch companies including KLM. Sander Stomph, CEO and co-founder of Kickstart AI, expressed satisfaction with the project’s success, highlighting their goal to accelerate AI adoption in the Dutch business community. KLM’s use of AI extends beyond reducing food waste; the airline also employs AI in aircraft maintenance, weather forecasting, and personalized travel tips for passengers.
