Lufthansa Turns to AI “Tray Trackers” To Cut Food Waste and Redesign In-Flight Meals
- Lufthansa is using a new AI-powered 'tray tracker' with smart cameras to scan meal trays on arriving flights and identify how much food has been eaten and how much is going to waste.
German flag carrier Lufthansa is trying to cut down on food waste and save some money by using a new AI ‘tray tracker’ with smart cameras that scan meal trays and work how much food has been eaten by passengers.
The innovative AI-powered system is already in place at Lufthansa’s hub in Frankfurt, and it has apparently proven so successful that the airline plans to invest in more tray trackers for its larger base in Munich.
How Lufthansa’s Tray Tracker Technology Works
Special cameras are set up above a conveyor belt leading to the industrial dishwashers at Lufthansa’s catering facility and scan every meal tray that has arrived on a flight to work out how much food has been left.
The system not only detects when a meal has been fully eaten, partially eaten, or left untouched, but it can also tell what the meal was, what flight it just arrived on, and which class of travel it came from.
Within a short time, Lufthansa gets rich data-driven insights into which meals are popular and which items on the menu are proving a real turn. The airline can also quickly establish whether passengers routinely choose not to eat on particular flights – such as very late night or early morning departures.
How Will the Tray Tracker Impact Passengers
Lufthansa can then use that data to change the loading of meals on certain flights or swap out unpopular menu choices that are going to waste. The airline also says the AI software will help it ‘optimize’ portion sizes.
3% – 5%
of passengers don’t show up for their flight, meaning that their meals go uneaten and the food is wasted
This isn’t the only AI software that Lufthansa is using to reduce food waste. The airline is also using another computer model that analyses various data points like the flight duration, departure time and previous demand to guess which flights it can away with loading fewer meals than there are passengers.
A similar AI system has also been used by Dutch airline KLM, which attempts to calculate how many passengers will miss their departure or get misconnected from another flight.
The airline then uses this calculation to reduce the number of meals loaded in the hope that this will precisely match the number of passengers who boarded the plane – rather than the number of booked passengers.
Around 3% to 5% of passengers don’t show up for their flights on time, which can result in food going to waste, although the AI model doesn’t always work perfectly. Since introducing the system, KLM has been known not to load enough meals, leaving some passengers going hungry.
Matt’s Take – Food Waste Has Long Been a Big Problem For the Airline Industry
It should be said that an AI-powered tray tracker isn’t necessarily new. Back in early 2020, Etihad Airways trialed a very similar system, and the company behind it claimed that Etihad could see a return on investment of between 200% – 1000% within the first year of operation.
Food waste is, of course, a massive issue for airlines because strict international health and agricultural regulations require that the vast majority of unused food be incinerated upon arrival.
During my time in the airline industry, I’ve seen companies use far less sophisticated measures to reduce food waste, simply loading enough meals for 90% or even 80% of the booked passengers, hoping that not everyone would show up on time or wouldn’t want to eat.
Artificial Intelligence takes some of the guesswork out, but it is still a guess – no matter how calculated. The system isn’t infallible, and sometimes passengers might end up without a meal.
This all just sounds very strange. Why not just use the BA Model of inventing “customer feedback” and telling all their customers who paid £2-20k for a J cabin that they don’t want to eat a proper meal on a flight up to 11 hours duration departing after a certain time, despite none of those FFs being able to locate a single fellow pax who ever gave said feedback.
Why waste money on AI programming, smart camera’s etc when you can just make s**t up and gaslight your customers into feeling they’re in the wrong for wanting full service in exchange for their money.
It’s difficult to see any good coming out of this for passengers. If people don’t eat or finish an item because it looks/smells/tastes bad then the AI is likely to just conclude that the portions are too big. Then you have less meals being catered or Lufthansa deciding that they shouldn’t spend as much per meal because look how people aren’t eating what’s offered.
If the bigwigs at Lufhansa really want to know how the food is and the flow of service they should simply fly anonymously in all cabins but especially coach, and personally eat the food and observe and listen to what people say and do. This would also provide priceless insight on general comfort and service levels on short and long haul flights. Unfortunately, I don’t see Spohr and his ilk flying in the back themselves or subjecting their precious families to the horrors of economy, let alone for 10+ hours.