• Hugo Walker

How AI can reduce perishable food waste

Updated: Feb 24

Food waste is prevalent in the food industry at every level of the supply chain. A colossal 1.3 billion tonnes - the equivalent of one-third of all food produced - is wasted annually. In 2019, 3.6 million tonnes of food was wasted in the UK - 2 million of which was still edible. These statistics stand in great contradiction with the 2 billion people worldwide who suffer from hunger or undernourishment. It is clear that there is a global disequilibrium between food supply and demand.

On the retail level, perishable food waste can arise due to pricing: there is a disconnect between the price of a perishable good and how the consumer values it, depending on how close it is to the best-before date. The closer a perishable product gets to its best-before date, the less it is valued. We’ve all bagged a bargain with a product that goes off tomorrow - but what about all the times that we’ve put a product down that has five days left of its shelf life, in favour of a product at the same price that has a week left? Would we have purchased the first product if it had been at a lower price? In other words, would retailers be able to reduce the amount of food waste if they had a more streamlined markdown optimization strategy?

Tech startup Wasteless has taken on the challenge of reducing waste and optimising the markdown price using AI. By using dynamic pricing that depends on how close a perishable item is to its use-by-date, rather than fixed pricing, the price more closely mirrors the value that consumers place on the good. The technology can be easily integrated with supermarkets’ existing systems.

As CEO and co-founder of Wasteless, Oded Omer, explains: “Let’s say that a store has ten packages of chicken thighs. They might sell seven and throw away three. On the last day, they offer a 40 percent markdown. We say: Don’t wait. Start with a 2 percent markdown six days before the expiration date. Then drop it another 7 percent or 11 percent. That way you’ll never reach 40 percent only on the last day. The product will sell well before the expiration date.”

A pilot run of Wasteless in a Spanish supermarket has reduced perishable food waste by a third. Mr Omer says that the algorithm is constantly learning and hopes to have that figure up to 80 per cent when the AI is fully developed. Wasteless is set to roll out in Poland’s Makro supermarkets in the coming months.

The UN’s Sustainable Development Goals include halving per capita global food waste at the retail and consumer level by 2030. Additionally, the SDGs seek to reduce food waste at the level of production and processing - a feat that may require an overhaul of our relationship to consumption and distribution entirely, beyond the level of pricing strategies. Yet startups like Wasteless give us an important demonstration of how AI and technology can be used to increase our organizational capacity within the market mechanism and lead to a less wasteful allocation of perishable items on the retail level.