A fruit yield forecasting tool startup has raised £2.8m ahead of a commercial launch next year.
FruitCast’s system uses a moving camera – which could be a GoPro on a moving trolley or handheld – to take thousands of photographs of crops, which are counted and analysed by AI to precisely forecast growth and predict yields.
The technology has been on trial at “at field-scale” with two UK soft fruit growers and processors, WB Chambers and Place UK, over the last two years.
The tool delivers the predictions through a user-friendly interface so production managers can allocate labour and market sales for up to six weeks in advance, FruitCast said.
“It’s a time-sensitive crop, and too much fruit in any week means waste and loss, and too little, you lose business,” said the company’s chief technical officer Dr Raymond Martin, who developed the technology on the back of his PhD research on optimising soft fruit processes.
The new capital was raised from private investors Ceres Agri-Tech, and Innovate UK through its investor partnership programme.
“Planning our labour needs and availability for our customers is wholly dependent on accurate estimation,” said Place UK production manager Dan Yordanov. “There’s a lot of calculation in forecasting, it takes a lot of time, a large amount of input from experienced staff and is currently subject to human error. This is a real challenge for the industry.”
“There’s plenty of very advanced AI out there, but the team at FruitCast understand what we need it to do rather than just what it can do,” he said.
FruitCast will be offered as a subscription system. Development will continue over the next six months to become commercially available for early access in the 2024 growing season.
Initially it will be used on strawberry and berry crops, but it is expected this capability will extend to other soft fruits, top fruit, brassicas and salads.