AI-enhanced garbage bins aim to solve food service’s waste problem

Leanpath’s waste-tracking technology is used in over 4,000 kitchens, including Google offices, Marriott hotels and university dining halls. Its line of trackers have scales and touchscreens on which chefs select the type of food, reason for disposal and time of day among other details. Some models use smart cameras to help employees visualize the accumulation of wasted food. The company estimates it saved 15 million meals from the garbage last year and prevented 55,000 metric tons of greenhouse gas emissions.

Dutch startup Orbisk’s waste calculators are being tested on two Carnival Corp. cruise ships after the operator found low-tech solutions — such as offering smaller buffet plates — only cut so much waste. Carnival slashed food waste across its entire fleet by 45% between 2019 and 2024, but it required onerous self-tracking methods.

“We’ve been able to consistently maintain that 45% over the last five or six months now, but we’re kind of at that plateau range again,” said Carnival’s food operations director Schalkie Badenhorst. “Which is why we’re looking at Orbisk and bringing in new methods to refine this a little bit more.”

Orbisk’s technology uses cameras and scales to detect the types of food hitting the bin — uncovering previously hidden opportunities for waste reduction. For example, in its first two months using Orbisk, Carnival found that its Bonsai Sushi restaurants have been scrapping a considerable amount of cucumber as chefs mainly used the outside of the veggie in dishes. Elsewhere, Carnival was shocked by the number of photos of unfinished fries from its burger bars — leading it to reconsider the size of serving scoops.

There are limitations to what all AI-driven technologies can do. Carnival’s Badenhorst says systems still aren’t smart enough to tell the difference between cuts of steak, a distinction that could provide cost savings for restaurants if they knew what specific beef dish customers or chefs were sending to the bin.

These systems also won’t help places that rely heavily on takeaway customers, considering how much waste is created after food leaves the kitchen. Dana Gunders, executive director of US nonprofit ReFED, says data suggests that 70% of food service waste comes from customers not finishing what they’ve been served or what they picked up at a buffet.

There are also challenges around complexity and cost. AI-driven waste systems create an extra step for cooks and kitchen managers — and the technology would be a big investment for smaller dining establishments. Orbisk’s systems cost between $2,000 and $8,000 annually depending on a customer’s needs. Winnow charges a monthly fee that ranges by option, but declined to give a price range, and offers a cheaper tablet-only solution targeted toward smaller kitchens. Leanpath would not comment on its prices.

Ultimately, those in the kitchen may be the strongest advocates for technology that reduces food waste. After all, chefs will say there’s nothing more they like to see than a clean plate.

“Chefs don’t want to throw away their food, they want to feel empowered that what they cook actually makes it to the plate,” said Hamish Forbes, a senior analyst at NGO Waste and Resources Action Programme. “AI is a great way of reducing the burden of that first step.”