What is automation good for? Harvesting more broccoli than human workers, says Upp, using computer vision AI and its own farm-scale machinery to expand crop yields , an agTech startup based in Shropshire, UK.
Its pitch is to enable its specialist, AI-driven harvesters to not only harvest familiar crops more efficiently, but also extract more nutritious protein from fields of broccoli. And that process also reduces waste. extra human workers to do it.
With smart machines the company is developing, Upp is helping broccoli farmers harvest more plants than they can actually harvest using human field workers, because a combination of AI and tractor tools does everything. It is said that it will be , at speeds up to 3 km/h.
This AI-driven approach will allow farmers to “upcycle” 80% of broccoli plants (i.e., excess stems and leaves) that are typically left in the field as waste, according to Upp, and turn them into processable additions. can be sold as a product of The morphology it suggests is comparable to pea protein.
CEO and co-founder David Whitewood told TechCrunch that the startup’s concept system, which is being developed with the help of engineers at the University of Lincoln, has an onboard 3D camera and computer vision AI running. I use a tractor equipped with a computer. A model trained to identify broccoli heads of appropriate size for picking (claimed to be more accurate than humans) and a proprietary (patent pending) tractor-pulling cutting and Comes with a harvesting tool.
“The job of harvesting broccoli must first recognize which bulbs are ready for harvest. I was,” he explains. “We’ve tested a lot of cameras with them to address the difficult problem of occlusion. [where leaves may partly obscure the camera’s view of the broccoli head].
“They use a depth-sensing camera with a 3D piece inside to determine the size of that head. When you say “cut,” it sends a signal to the on-board computer to activate a patented mechanism to grab.Plants — this is the same way humans grab plant stems — and very A sharp knife comes flying and chops it up in a split second. Then the plant is lifted. ”
Extra edible plants harvested this way are not meant to be on supermarket shelves. In supermarkets, the strict cosmetic standards that grocery retailers typically apply to their suppliers are a major cause of food waste by rejecting inventory of fruit and vegetables that don’t look perfect. The idea is to process it into protein- and nutrient-rich ingredients to sell to the food industry.
Upp envisions using dried broccoli protein in a variety of products, from sports-style protein drinks to prepared meals and baked goods.
According to the startup’s website, the broccoli pieces the company targets for upcycling are 30% protein by dry weight and contain nutrients (vitamins A, B, C, E, K, calcium, iron, potassium, phosphorus, , zinc). As well as being high in fiber.
Upp seems to have no problem getting early interest from the food industry for its upcycled edible plant protein — Whitewood points out that it already has three industry partnerships. (He can’t name names yet, but he says one of them is global, another is a major British food brand, and a third is a confectionery specialty bakery.) .
“They are very interested in the health aspects of broccoli,” he continues. “They’re interested in the fact that it’s clean and sustainable … so you could say they’re excited. .”
According to Whitewood, when it comes to processing, App is working with experts at the James Hutton Institute in Dundee to find the best ways to “recover fractions from plants that are primarily suitable for the food industry.”
Upp is touting itself as a specialized “circulating plant protein” business amid growing demand for alternative plant-based proteins as the food industry seeks ways to reduce its reliance on animal-derived proteins. developing things. Reduce your carbon footprint—putting global pressure on all businesses to meet their climate goals.
As such, the startup predicts that AI-harvested broccoli protein could grow into a multi-billion dollar market in the next few years.
Marketing suggests that broccoli protein is cleaner than pea protein (it’s a quarter as carbon-intensive to produce), with added environmental benefits. While claiming it avoids the deforestation problem that is tarnishing the soybean crop’s reputation. So here’s the pitch more environmentally friendly vegetable protein.
Automating broccoli harvesting will inevitably result in a shift of (human) workers.
Whitewood says the system will replace about seven field workers, but says packhouses still need “warm bodies” to package broccoli products for retail. . “It’s hard to get seven people,” he added, sketching a picture of the grueling jobs field workers typically have to do, arguing that these aren’t jobs that anyone would miss. “Nobody wants to do this job. Even in China and India they have a hard time getting people to do this. It’s the 21st century and we still expect people to do this.” It’s just crazy.
Founded in 2022, the startup’s technology has been developed to the concept stage and is preparing for the next stage. To hone a robust technology that can be commercially deployed. England, Spain, California.
It is expected to enter commercial production (and generate its first revenues) in late 2024, with revenues projected to exceed £50m in 2027 across three pilot markets.
The business was set up last year as a spin-out from another UK agTech business called Earth Rover. Whitewood served as CEO before he moved to Upp as a co-founder when Upp decided to split him into two different businesses.
Today the startup is announcing a £500,000 pre-seed investment Decarbonization, sustainability and social impact investor Elbow Beach Capital will fund field trials ahead of plans for commercial deployment next year.
Whitewood said the first commercial use of the technology would probably be in Spain or the UK for seasonal reasons, after which Upp offered broccoli growers in California to automatically optimize crop yields. will be
Why hasn’t anyone thought of extracting more good stuff from the broccoli plant? Given the selective harvesting and the need to separate the harvested crops, it is “really difficult,” he suggests. sold fresh), the rest require additional processing.
“It sounds simple, but many have tried and many have failed,” he suggests. “It’s only when you have a specialized harvester that can handle most of everything that all of a sudden you can start processing the rest. , you can’t handle crops of this size, this amount of crops… you need farm-sized machines.”
