Using AI, they installed cameras on a conveyor belt on an avocado production line where they scanned the fruit for freshness and ripeness.
This allowed them to work out how much dry matter was in the fruit to determine when it would ripen, and which ones would be best suited for longer trips, such as when exported.
“You can understand information about 100 per cent of the food products being processed on a conveyor belt rather than relying on sample-based testing,” Ramanan said.
She used the same technology while embedded in a commercial fish market in Spain to identify when red snapper was being substituted with tilapia, a much-lower-quality fish.
The food systems entrepreneur said agriculture was yet to fully benefit from ‘the explosion of digital technologies that has helped other sectors.
“I think there’s a number of different ways AI can shape the food systems landscape in general, starting at the very beginning of the supply chain,” Ramanan said.
Harriet Mellish, of Agrifutures, helped pull together the evokeAG event, and said it was important to get AI-driven innovation on the agenda, with a particular emphasis on what’s happening elsewhere in the world.
“To bring thought leaders and experience in this space to Australia is to ensure we are hearing from a global view,” she said.
But for Ramanan, AI has plenty of challenges too, like finding the right place in the supply chain to integrate the technology and working out who will pay for it.
“The big challenge with the food industry is that it’s a low-margin, high-volume industry that’s quite traditional; it doesn’t have the budgets of the pharmaceutical industry,” she said.
EvokeAG kicks off on Tuesday and runs for two days, bringing together innovators, producers and investors from the agrifood space.
Also on the programme are food security and the challenges posed by mining and agriculture, as well as how producers can quantify their greenhouse gas emissions.