The rise of AI has touched nearly every facet of consumers' lives, from music to maths, literature to logistics. It’s becoming the norm, and the food industry hasn’t escaped the phenomenon either.
Manufacturers are particularly interested in AI’s ability to increase supply chain efficiencies. Companies like the Bel Group and Danone, the latter in partnership with Microsoft, are leaders in this space.
Success of AI in food
The success of AI is like trying to get tomato ketchup out of a bottle: nothing comes out for a while, and then “all of a sudden, it all arrives at once”, said British Frozen Food Federation (BFFF) head of health & safety Simon Brentnall at the IFE show in London.
Alina Sartogo, co-founder of the Wonki Collective, agrees. “In a matter of years it’s going to be everywhere,” she says.
AI centres the digital rather than the physical. In many ways, modern innovation is about software, says Paul Wilson, CEO of Scorpion Vision. The important questions are not the manufacturing tools themselves, but what software goes along with it.
Optimising food processing
Philip Wilson’s company, Scorpion Vision, aims to make factory-level production more efficient using AI.
Its equipment performs a range of food packing tasks, such as de-coring lettuce and trimming leeks with waterjets.
AI’s ability to digitally model the food allows it to do this accurately. With the lettuce, for example, it creates a digital model of the lettuce, and uses this to know how deep the robot should go into the lettuce, and what angle to pierce it.
When this job is done by humans, there is substantially more wastage than with an AI-controlled robot, Wilson said.
The company is also used for quality control, for example sorting good, edible bacon from bacon which is burnt or unappealing.
Finding potential food waste
Sartogo’s company, the Wonki Collective, analyses the supply chains of its clients and finds food that has the potential to become food waste. Following this, it matches these companies with potential clients.
The company provides service-as-a-software, aiming in the long term to significantly reduce food waste. The company takes data from manufacturers and uses AI to identify potential food waste, its potential value and when it will be wasted, so that food manufacturers can make an effort to reduce it.
“Because food waste is not a priority, it helps manufacturers” by sorting out this data, explains Sartogo.
AI then makes a recommendation about whether it should be sold, donated, or another option.
Making health and safety more efficient
Within the BFFF, AI is being used to sift through CCTV footage in search of health and safety violations, including texting truck drivers.
While these trucks already have CCTV, finding violations like this is “like finding a needle in a haystack” for humans, due to the immensity of the footage recorded, says Brentnall. The AI, meanwhile, can find such behaviours quickly and easily.
The technology has led to a 40% reduction in road traffic incidents where the AI was implemented, says Brentnall. So powerful is the tool that driver prosecutions over six months in one depot have reduced from one a month to none.
On the factory floor, it has also detected incidents like blocked fire exits and incorrect manual handling techniques.
“AI works best when it supports people, not replaces them,” says Brentnall.
