Why is automation useful? The Shropshire-UK-based agtech startup, which is using computer vision AI and farm-scale proprietary machinery to expand crop production, says it can harvest more broccoli than manual workers.
The efficiency of the unique AI-driven harvester makes it more efficient to harvest a conventional crop, but the process reduces waste – more protein can be extracted from a broccoli field without an army. More human workers to do this.
The smart machinery the app is building will allow broccoli farmers to harvest more of the plant than they could using human field workers because the AI-plus-tractor-tool combo does everything: sorting, cutting, lifting and carrying fully automatically. , up to 3 km per hour.
This AI-based approach allows farmers to “cycle” up to 80% of the broccoli plant (i.e. extra stems and leaves) from the field as waste per up and sell that as additional produce. The form it refers to is comparable to pea protein.
The startup’s concept system, which CEO and founder David Whitewood told TechCrunch is being helped by technology experts at the University of Lincoln, includes a tractor powered by a 3D camera and on-board computer vision AI. A model trained to identify when broccoli heads are the right size to pick (with better accuracy than humans, it claims), with a proprietary (patent pending) tractor-pulled cutting and harvesting device.
“The task of harvesting broccoli – first – you need to know which heads are ready to harvest. So we partnered with the University of Lincoln Agri Products Group, which develops machine learning and AI,” he explains. “We’ve been testing full cameras with them and dealing with the difficult problem of concealment. [where leaves may partly obscure the camera’s view of the broccoli head].
“They used a depth-sensing camera with a 3D slice inside to determine the size of that head. Because we don’t cut every head – we just cut the size required by the supermarkets… Then it says ‘cut’ and it sends a signal to our board computer and then we do our proprietary method. The plant – like a person grabs a plant stem – and then a very sharp knife flies in and cuts it in a fraction of a second. And then the plant rises.
The extra edible plant matter harvested in this way is not destined for supermarket shelves – the strict cosmetic standards grocery retailers apply to their suppliers to avoid stockpiling beneath perfectly-looking fruits and vegetables contributes greatly to food waste – but rather the idea is to produce protein- and nutrient-rich ingredients to sell to the food industry.
The app shows that dried broccoli protein is used in a variety of products – from sports-style protein drinks to prepackaged meals and baked goods.
Broccoli Bites for Bicycling contains 30% protein by dry weight, according to the startup’s website, and is also packed with nutrients (vitamins A, B, C, E, K, calcium, iron, potassium, phosphorus, zinc). It is also high in fiber.
App doesn’t seem to have any trouble finding early interest from the food industry for its up-and-coming edible plant-based protein — Whitewood says it already has three industry partnerships (he can’t name names yet, but says one is global; a functional drinks giant; another is the UK’s biggest food brand and the third is a specialist sweets bakery.
“They are very interested in the health aspects of broccoli,” he continued. “They want the fact that it’s clean and sustainable… so we’d say they’re happy. I don’t think we’ll have a problem with that market – once we get it off the ground.
On processing, App at Whitewood is working with experts at the James Hutton Institute in Dundee to “find out how to get parts of that plant that make it particularly suitable for the food industry”.
Highlighting, App is developing what it bills as a specialist “rounded plant protein” business, against a backdrop of growing demand for alternative, plant-based proteins as the food industry seeks ways to reduce its reliance on animal-derived proteins. Reduces the carbon footprint – under international pressure for all companies to hit climate targets.
So the startup predicts that AI-harvested broccoli protein could grow into a multi-billion dollar market in the coming years.
The marketing side also touts additional environmental benefits – broccoli protein is cleaner than pea protein (with 4x less carbon to produce) and also avoids the deforestation that has tarnished soy crops. So this is the conclusion More green Plant protein.
Automating broccoli production will inevitably displace some (human) workers.
Whitewood says the system will replace about seven field workers — but notes that “hot elements” are still needed in the packing house to package broccoli products for retail. He adds, “It’s hard to find seven people,” painting the grueling work field workers often have to do, arguing that these aren’t the kind of jobs anyone should miss. “No one wants to do this work. Even in China and India they are struggling to get people to do this… It’s the 21st century and we still expect people to do this. It’s just crazy.”
In the year Founded in 2022, the startup is gearing up for the next phase as it develops its technology to a concept stage – to develop a robust technology that can be commercially deployed – with a series of “field-to-protein” pilots planned for this year in the UK, Spain and California.
It’s expecting commercial production (which will generate first revenue) by late 2024 — projecting revenues to exceed £50M in three test markets by 2027.
The business was formed last year as a spin-off from another UK agtech business, Earth Rover – when Whitewood decided to split into two separate businesses as CEO before moving to the app as co-founder.
Today the startup announces a pre-seed investment of £500,000 Elbow Beach Capital, a decarbonisation, sustainability and social impact investor to support field trials – ahead of the planned commercial deployment next year.
Whitewood says the first commercial use of the technology could be in Spain or the U.K., due to seasonality, before YUP shuts down California broccoli growers on automating crop production.
Why has no one thought of extracting more good things from broccoli plants? Whitewood says people have been thinking about the potential to do this for years, but he points out that it is “quite difficult” – given the importance of picking and sorting the harvested crop, part of it goes to supermarkets (for fresh sale) and the rest requires further processing.
“It sounds simple – a lot of people have tried and a lot of people have failed,” he points out. “Only if you have a specialist who handles all the mass can you suddenly deal with the rest. You need automation—and it needs big automation. Small robots don’t deal with crops this size, this size… you need farm size machinery.