“You know more about your Uber driver than the food you’re eating,” says Karim Giscombe, the founder and CEO of a new company, Plant-Ag, about to launch in the coming weeks to close this peculiar knowledge gap. More and more consumers today want to know the origins of everything they buy, from what factory produced their T-shirt to where the gems in their jewelry were mined. Yet the decision-making process we use to buy fresh produce—a good purchased far more often, and that can also make us sick—really hasn’t evolved. Ever. In the store we’ll grab a tomato, spin it once in our hand, and hope for the best.
So Giscombe, a breezy but persuasive former director at Merrill Lynch whose social-media bios describe him as a “reformed capitalist,” has invested more than four years and is raising $9 billion to build what he calls the first fully transparent field-to-plate supply chain in which consumers can track a vegetable’s journey from where it sprouted as a seed to the grocery store shelf.
The food system’s opacity has allowed foodborne illnesses to morph into a bona fide health crisis. Food is now believed to sicken one in six Americans a year—130,000 of them end up hospitalized and 3,000 die. The year 2019 was the worst on record, beating 2018, the previous record holder, which beat 2017, the record holder before that. The latest data shows rising illnesses traced back to Listeria, Salmonella, Shigella, Campylobacter, E. coli, Vibrio, Yersinia, and the parasite Cyclospora.
The toll has health officials confounded. More stringent rules aren’t having the intended impact. Take lettuce: Since 2018—the year that the U.S. government declared that no head of romaine in America should be assumed safe to eat, instructing all U.S. citizens to destroy theirs and sanitize the fridge—a Shiga-toxin-producing E. coli strain found in leafy greens has continued producing a string of deadly outbreaks. (Last Thanksgiving, it forced 1,000 Walmart stores to recall bags of romaine yet again.)
Giscombe and his team believe that the industry’s big players are too entrenched to start over, and food-tech startups seem to be hacking away at small parts of the problem in isolation—next-gen DNA-sequencing (Clear Labs), “intelligent” food packaging (Primitives’ compostable blister pack with built-in spoilage sensors; SmartRiPE, a plastic container with a full RFID tag), and wifi-enabled miniature mass spectrometers (the SCiO Sensor)—often to middling results.
Instead, Plant-Ag is trying to fix the broken system in one fell swoop: by seizing blockchain, embracing greenhouse farming, building a platform to analyze billions of gigabytes of data, creating an agtech brain trust, making everything open source, and launching what appears to be the world’s biggest food infrastructure project. As you’ve imagined, it’s an exceptionally expensive undertaking. To fund it, Plant-Ag is raising $800 million from institutional investors, along with more than $8 billion in public-private partnership bonds brought to market by JPMorgan and Chicago-based investment bank Ziegler.
Plant-Ag is using that money to break ground on a network of high-tech greenhouses in April, and the first facility should be supplying a variety of produce identified as high risk by the Food and Drug Administration (FDA)—lettuce, tomatoes, eggplants, blueberries, strawberries, kale, basil—to the southeastern U.S. by fall 2022. Within four years, the company expects to be able to feed one-third of America with fresh produce that was growing just 72 hours earlier.
SHORTENING THE SUPPLY CHAIN
“There is absolutely no one,” Giscombe tells me, “who is delivering a product that fast.” According to his numbers, even if you include Walmart, whose supply-chain management is the envy of the business world, the industry average for produce is 11 to 12 days. For imports, it might be a month. (Half of the fresh fruit and a third of fresh veggies Americans buy now come from abroad.)
One reason is that vegetable production has become unusually concentrated. For example, 90% of U.S. lettuce is raised in California’s Salinas Valley from March until November, at which point the growers all “transition” to Yuma, Arizona, for winter. Nine in ten wintertime tomatoes come from the Immokalee region of Florida, and 70% of America’s annual apple supply is harvested by Washington state in the fall. (They’re kept in suspended animation, spritzed occasionally with fungicide or a product called SmartFresh that stops ripening.)
Plant-Ag plans to ship faster, but also shorter: Produce never physically travels more than six to eight hours from where it’s grown. A Plant greenhouse harvests at 4 o’clock in the morning. Distributors must load produce within 12 hours, then deliver it six to eight hours after that. By the 72-hour mark, orders must reach their point of sale in retail outlets, or their point of use at restaurants and foodservice providers.
This accelerated turnaround yields another benefit. “Americans don’t realize how much their produce has been genetically groomed for transportation,” says Plant-Ag’s Chief Operating Officer Richard Dent. Their continued susceptibility to pathogens notwithstanding, veggies and fruits were rewired by the food industry to withstand a beating. “Take the need for durability and a long shelf life out of the equation, and now we can leverage these really great minds to help with things like taste and yield,” Dent explains, referring to another Plant-Ag arm: an R&D lab of sorts that its starting with the University of Florida’s Institute of Food and Agricultural Sciences.
The first Plant-Ag greenhouse—dubbed Distributed Network Production Site 001—is strategically situated on Interstate 10 just outside Jacksonville, Florida. Within a six- to eight-hour radius are Florida, Georgia, South Carolina, and parts of Alabama, Mississippi, North Carolina, and Tennessee. When Site 001 starts selling fruits and veggies, it might be at a Pensacola, Florida, Publix; an Atlanta Whole Foods; or a Charlotte,North Carolina, Harris Teeter, and a host of restaurants and foodservice providers in between. The project’s first phase will put facilities in seven more key markets, such as Delaware and Texas, to supply the south-central U.S. and Eastern Seaboard by 2025. The plan is to be in the United Kingdom by then too, for a footprint totaling 3,000 acres of controlled-environment agriculture. Phase two tackles the West Coast after that.
Giscombe promises their margins are so good, there won’t be any $6-head-of-lettuce sticker shock: “We can sell it for a lot less than our competitors.” But unlike sneakers or even soft drinks, produce isn’t necessarily something consumers shop by brand. (Quick—what’s the last romaine lettuce brand you bought?) To prime people to start, Plant-Ag is also launching a nationwide billboard advertising campaign in a few weeks that will casually mention how little anyone in that market knows about their food.
Meanwhile, to keep tabs on that 72-hour countdown, Plant-Ag is working with IBM. The two companies hope to build a new iteration of IBM Food Trust, a blockchain system that’s established itself as the gold standard for tracking products across a supply chain, on top of the same open-source network on which all of Food Trust’s transactions operate: Hyperledger Fabric.
Food Trust already counts such brands as Walmart, Kroger, Unilever, Dole, and Nestlé as clients. For these juggernauts, IBM’s General Manager of Blockchain Services Jason Kelley tells me that shipping a crate may involve “200-plus documents,” alongside emails, phone calls, even a couple faxes, “because everybody has their own way of approving things as they travel to the consumer.”
In fact, the effect on food safety became apparent during a traceback exercise that Walmart organized when it joined Food Trust, back in 2017. IBM and the retailer both had to identify where a pack of sliced mangos originated. Walmart’s food-safety team took six days, 22 hours, and 15 minutes to snake back through the supply chain. IBM apparently found the mangos’ Mexican orchard in 2.2 seconds.
However, retrofitting enormous preexisting supply chains with blockchain requires effort (and several years of time, Walmart learned). Giscombe realized, for all its innovation, that blockchain was still falling short in two important ways: One, it’s a tool that reacts to outbreaks after they happen, rather than preventing them from occurring. Two, its reach always stops at the retailer or food distributor, instead of empowering consumers.
Being preventative meant making the process before shipment safer. Giscombe had watched an absurd amount of private capital be invested in greenhouses after America’s cannabis legalization wave started in 2012. He learned that indoors, you can eliminate over 95% of the exposure to bacteria and other pathogens.
So, Plant-Ag decided to make some really high-tech greenhouses.
GREENHOUSES OF THE FUTURE
The Dutch mastered indoor farming long ago, back in the 1970s, transforming their Maryland-sized country just 1,000 miles from the Arctic Circle into Europe’s top vegetable exporter. Their pioneering practices have been adopted by not just cannabis growers, but also ordinary U.S. produce suppliers that value predictability.
Most years since 2012, greenhouse farming’s compound annual growth rate has been upwards of 20%. In 2018, Wendy’s switched to greenhouse-grown tomatoes in the United States, and in Canada its lettuce is now, too. Walmart now sells BrightFarms, and Whole Foods offers Gotham Greens. “The demand is there,” says Giscombe. “And when you add our numbers in, it’s going to exponentially change that.”
To do that, Plant-Ag brought in VB Greenhouses Projects, a top Dutch builder, to build its controlled-environment agriculture sites. VB’s roster of clients runs from ag giant Syngenta to regional growers in Siberia, and the technology that heats FC Barcelona’s soccer field is even VB’s. But CEO Edward Verbakel says that no client has been driven by data and traceability like Plant-Ag. “Food needs to be traceable,” he tells me. “We have not always seen that driver in our industry. The focus is on quality and predictability, but not traceability.”
I was shown a rendering of Site 001 in Florida. It’s a tremendously large glass and metal structure that looks like a combination factory and science laboratory, with plants protruding from trays, conveyor belts, and automated arms. Insects and other external predators are kept out using an overpressure system. It uses hydroponics, therefore is dirtless, and Verbakel mentions that their closed-loop system will use 15 times less water than conventional farming. Both Plant-Ag and VB boasted about the site’s low carbon footprint, which consumers are also able to track.
The goal, understandably, is to eventually automate everything. Verbakel says they’re already there with lettuce: “No hand touches it. The seeding happens with an automated machine. And the cutting of the leaves at the end is also automated.”
On any farm, the humans themselves are a key contamination vector. In 2015, for instance, the FDA banned some Mexican cilantro after investigators found human feces and toilet paper in fields linked to hundreds of cases of cyclosporiasis. That’s a rather extreme example, and one that reflects the substandard conditions those laborers worked under, but the point is the fewer people in contact with your food, the lower the risk. And indoors, people are among the only threats. Entering a Plant-Ag greenhouse, therefore, will feel like going inside Fort Detrick: Workers change clothes, put on different shoes, and don gloves and hairnets.
I asked Verbakel what happens in the event that a pathogen does infiltrate this compound. “If we find something that could be a threat, we lock that specific zone off and try to insulate that predator,” he told me. Doors and walls between growing areas can be closed, sealing one off from the next, and workers in one zone can’t intermingle with the workers in other ones. Tools are also row-specific, meaning they can only be used on a tiny subset of any particular crop. Meanwhile, a central command is busy analyzing data to monitor plant vitals and yield outcomes. “If anything changes, for whatever reason,” says Verbakel, “you recognize it immediately and can react.”
Giscombe quips that they’ve created AaaS—agriculture as a service—meaning at root, a cloud platform linking the supply chain nodes. The model’s novelty is the data it possesses along the chain, end to end. A carton of blueberries or a tomato can be followed from germination to the last mile, which might stop at a fast-food order pickup, the grocery store register, or a home delivery by a courier from Instacart. Additional AI can be layered in the middle to get even more granular: Not just the stops a delivery truck made en route, but if the driver parked in the sun instead of the shade, and if so, how much warmer each corner of the trailer got.
Aggregating those data points is the “really exciting” next step, says IBM’s Kelley, and one reason the tech company signed on. He calls it the data-gathering equivalent of Walmart’s traceback “going from a week to 2.2 seconds.” And this gets back to Giscombe’s other critique of blockchain: that it typically stops short at retailers and distributors on the front end. “Very few platforms allow the consumer to have real-time knowledge about a product,” he says. Likewise, on the back end, it begins as items ship—too little, too late, in the case of food.
Plant-Ag’s front-end solution is simple mobile tagging on packages. When describing it, Giscombe was emphatic that “this is not going to be a QR code,” as he thinks everybody is tired of seeing those. Whatever its actual design, customers will scan the package with their phones, and details about the item’s journey will pop up. So will something more radical: a virtual 3-D tour letting consumers view where that piece of produce grew inside the greenhouse.
Before moving to Plant-Ag, Dent spent decades in retail, at companies like Bath & Body Works and Victoria’s Secret Pink, which he helped grow into a billion-dollar household brand in five years. Transparency wasn’t his old industry’s strongest suit. “There was always this place we kind of didn’t want to take you,” he says. “Here, it’s not, ‘I hope you won’t ask about my dyeing facility in Sri Lanka that I don’t want to talk about.’ We can literally bring you inside the facility.” This translates, from the outset, to greater accountability on the back end for practices that go beyond safety—like sustainability and labor conditions.
But data on the plants themselves is the true game-changer, Giscombe believes. He says the genesis for the company was a conversation he had around 2014, before he left investment banking. The individual he talked to had invented a tool that can analyze plant genetics to determine the perfect piece of produce for a market, then adjust the growing conditions to replicate its traits without genetic modification.
The tool “blew my mind,” Giscombe says. “I thought, ‘What if this technology could be a component of something bigger that reshapes that system?'”
Right now, Plant-Ag is in the process of acquiring this technology and integrating it into its platform, alongside the blockchain and its farming techniques. Once this triad is set, Giscombe says the plan is to make the platform open source, accessible to the entire produce industry.
He compares their decision to Tesla removing its electric vehicle patents. By putting Tesla’s battery technology out there, Elon Musk made progress collaborative. Giscombe argues the food system needs a similar shared, rapidly evolving platform. “That technology creates efficiencies for us that equate to more dollars and cents, and it allows us to build more advanced processes than other providers are currently utilizing,” he says. “And that’s why it should be a baseline for how the whole industry operates.”
Tesla revamped its mission in 2018, from just selling electric cars to convincing society to adopt more sustainable methods of energy production. Now, its valuation is also hung on how well it can dominate the entire energy sector—which is ambitious, to say the least. That isn’t intimidating to Giscombe, but he has taken a lesson from it. “What Tesla is after was too big of a picture to talk about out of the gate. So they phased it in,” Giscombe says. “But the truth is, if you move society forward? It can’t be just about you. It’s not, ‘How do we do this, how do we get that?’ It should be how do we change this for everybody.”