Choosing the right moment to harvest can be a make-or-break decision

When the first day of the apple harvest rolled around in Washington State last year, the fruit was ripe and the pickers were ready – but the weather had other ideas.

"It was like 38C… it's not safe for people to work in that heat," recalls Joel Carter at Okanagan Specialty Fruits. "We had to stop at 10 o'clock in the morning."

He says that artificial intelligence (AI) models that forecast ideal harvest dates, taking into account the weather, would be useful. "You need to know more than just when your fruit is going to be ripe. How long do you have to pick it?" he says. "That's where these models are really helpful."

Carter's company has more than 1,250 acres of apple orchards in Washington, and the fruit is grown for sliced apple portions – often sold to hotels and schools, for example. The firm is investing in technology with the hope of maximising productivity. Even the apples are genetically engineered so that they don't brown easily once cut.

But planning a harvest is tricky. New tools that count and analyse fruit on the tree or vine, and predict when crops will ripen, are emerging. It matters because prices for fruit, especially high-value berries such as strawberries or blueberries can fluctuate wildly. Getting the harvest date wrong means you book seasonal workers when you don't actually need them, and risk missing out on the biggest profits.

Okanagan Specialty Fruits is already experimenting with cameras from a Canadian firm called Vivid Machines. The cameras are mounted atop tractors and they scoop up imagery of the apple trees as the tractor trundles by. AI identifies buds, flowers or fruit in that footage.

"Right now, Vivid is telling us crop estimates and harvest dates," says Carter. He notes that the system is good at picking out very tiny flower buds, which are hard to see at a glance with the naked eye.

But the accuracy of forecasts is noticeably dependent on the quality of historical information fed in to the system, adds Carter. "This isn't something where an AI can scrape the internet and figure out what's the average [yield] for Granny Smith," he explains. "It's going to be bespoke to your farm."

Apples are at least somewhat forgiving – the harvest window for those Granny Smiths is three weeks long, says Carter. For other fruit, such as berries, you might only have a few days.

"If a strawberry crop is on, you have to harvest it – otherwise your entire crop gets diseased very, very quickly," says Raymond Martin, co-founder and chief operating officer of FruitCast, a UK company that offers harvest forecasts to fruit growers here.

His firm offers growers crop predictions for strawberries and also raspberries, blackberries, blueberries and tomatoes. "We're moving on to grapes next year," adds Martin.

Don't experienced farmers know when their fruit will be ripe, I ask? Martin says they generally do – but not necessarily across their entire farm, which might be many acres in size, or have both outdoor and indoor growing areas. "We do exactly what the farmers could do but we just do it on a scale that they can't."

The system analyses footage of ripening fruit that can be captured by drones, someone walking up and down a field with a smartphone, or a camera mounted on a farm vehicle.

This year has been challenging, notes Martin, with hot weather and severe drought hitting much of the UK. This has stressed many fruit plants, putting them into thermal dormancy and slowing down fruit production.

There's only a small window for picking berries says Raymond Martin

The FruitCast model accounts for weather and irrigation conditions in its forecasts. The company says, "our forecasts land within 10% of actual picked volume one week out (90% accurate) and within 17% three weeks out (83% accurate). We guarantee less than 20% error…"

Angus Soft Fruits is one company that has worked with FruitCast. AI tech for forecasting fruit ripeness is not yet "a finished solution" says operations director Neill Finlayson: "This journey is still ongoing and, whilst significant progress has been made, the industry remains some way from achieving a fully integrated forecasting ecosystem."

Driscoll's, a California headquartered fruit seller with a big operation in the UK, also confirmed to BBC News that some of its independent fruit growers in the UK have used FruitCast's technology.

Yasaman Ghasempour (left) is developing ripeness detection techniques

Researchers are also exploring new ways of analysing fruit in intense detail. Many farmers already use handheld brix meters to gauge the sugar content of their crop. These devices work by measuring how much light has changed direction, or refracted, after passing through a liquid – indicating the volume of solids present.

Some such meters use infrared light, meaning there's no need to cut into the fruit to take measurements. But Yasaman Ghasempour at Princeton University says millimetre waves, high frequency radio waves, can penetrate deeper "They basically respond very well to humidity, water [and sugar]," she says.

She and her students have come up with a millimetre wave-based ripeness detector, external that could be used by farmers – or even customers searching for perfectly ripe fruit in shops. Her students tried it out at a local market in New Jersey. "[Staff] there got kind of scared that we were doing something shady," laughs Ghasempour.

Such fine-grained analysis of fruit, if currently unripe, could also help inform forecasts of when it will be ripe.

Jing Zhang says new tech has to prove its worth to farmers

But while some growers, such as Carter, are interested in trying out new tech, many will wonder whether investing in it is really worthwhile. "Adoption is very complicated," says Jing Zhang at North Carolina State University. "The grower has to have confidence in the research and whether or not it works."

Zhang has been working on a system to automatically count the number of blueberries, external captured in smartphone images of blueberry bushes.

The simplest systems need not be expensive. Kevin Wang at the University of Florida has developed another crop-counting tool that can harvest imagery, external from $100 (£74) drones.

In the future, such information could feed in to ripeness predictions. However, he points out that some farmers might be uneasy about sharing commercially sensitive information about their growing strategies, such as fertiliser and irrigation plans, with third parties and AI models.

Ben Palone is senior director of automation and commercialisation at Western Growers, an association that represents farmers in the western US. He says harvest forecasts have some potential as "an optimisation tool".

But farmers will likely continue to rely on human intelligence. "Growers like to have people in the mix to make some of those very critical decisions – especially when it comes to harvests," he says.

The critical tech staying safe by going underground

Why airlines are warning over lithium-ion batteries