The age of AI-assisted flight provisioning is dawning

Several years ago, United Airlines announced that it would start using lighter paper for its in-flight publication, shaving one ounce off the weight of each magazine. The airline estimated that this seemingly minuscule change would save it 170,000 gallons of fuel a year across its fleet, accounting for $290,000 in savings.

Now imagine if airlines could remove heavy cases of soft drinks or unnecessary meal trays from their planes. The financial savings, not to mention the reduction in food waste and lower greenhouse gas emissions, would be significant.

Each day, tens of thousands of flights take off around the world with more food and beverage supplies than they need. The result is massive food waste, unnecessary costs, and inefficiencies that add up to hundreds of millions of dollars each year. Citing a 2023 study by Singapore’s Changi airport, the International Air Transport Association estimates that each passenger generates nearly 1 kg of cabin waste per flight, translating into 3.6 million tons per year. That figure is estimated to double by 2040 with current passenger growth rates. That’s a lot of fuel.

Despite operational costs that continue to rise, airlines have persisted with a rather rigid and simplistic catering system—stock every flight’s commissary as though it’s at capacity in order to ensure there’s plenty of drinks and snacks for everyone. But forward-thinking airlines are beginning to harness machine-learning and artificial intelligence to embrace dynamic provisioning, an approach that ensures aircraft are loaded with only what’s needed based on real-time consumption analytics.

It’s a monumental shift that has the potential to transform airline catering, reducing costs, improving sustainability, and streamlining operations.

Solving the complexity of in-flight catering with AI

To understand the scale of the problem, let’s consider the operations of a major international carrier. Imagine the airline flies to 140 destinations using 30 different aircraft types. Each of these types could have up to 10 seasonal or situational variations—region flown, spring break, Ramadan, Christmas, morning vs. evening flights, and more. Let’s say each aircraft has an average of 10 galleys.

The result: potentially well over 10,000 different galley plans, most likely managed through spreadsheets or legacy software. The sheer complexity means most airlines resort to the safest method—fully loading the galleys, no matter how many meals are expected to be served on each flight.

IFCS customer airlines will soon be implementing consumption analytics technology that allows them to track exactly what is being used on flights—at a level of granularity that was previously impossible to achieve.

Using systems integrated into the airline’s supply chain, this kind of in-flight catering software can adjust inventory levels based on real-world usage data. For example, the data might reveal that only 70% of an airline's typical stock of Coca-Cola cans will be consumed on certain routes. Instead of loading the usual full number of cases, the system would dynamically adjust for the expected consumption, ensuring that the next flight only carries precisely the number that’s needed.

No more guessing: provisioning flights in real time

Based on a continual stream of data on food and beverage consumption, including supplies that are either under or overstocked, machine learning algorithms adjust future loading plans in real time. Instead of updating thousands of spreadsheets, adjustments are automatically made in the next provisioning cycle.

One of the biggest advantages of in-flight catering software is the ability to analyze massive amounts of data, while keeping communications to staff simple to understand. Catering and onboard crews don’t need to manually count or reconcile stock; they simply scan a label and follow instructions. They will also be able to interact with the AI technology through natural-language queries.

In short, the airline catering space is poised for a revolutionary technological shift. Loading galleys in the traditional way will very soon start to seem badly outdated and airlines with the foresight to leverage the right consumption-tracking and provisioning tools will reap significant competitive gains.