Weeds, the enemy of gardeners and farmers everywhere, could soon meet their robotic match.
FarmWise, a California-based weed-removal startup that works with some of the largest farms in the US, has raised $45 million in Series B funding, its co-founder and CEO Seb Boyer told Emerging Tech Brew.
FarmWise’s main product is Titan, an AI-powered robot trained to identify and cut down weeds without harming the desired plant. The company is part of a broader crop of precision-agriculture startups looking to harness tech to simplify processes for farmers while limiting the amount of pesticides and herbicides used to grow.
Around a decade ago, VC funding for agtech startups was ~$230.1 million. Since then, funding has grown to $10.5 billion in 2021, a 41.6% compound annual growth rate, per Pitchbook.
The round was led by Fall Line Capital and Middleland Capital, with participation from Taylor Farms and GV, as well as existing investors Playground Global and Calibrate Ventures.
In the field
In practice, Titan roams a farmer’s field, taking pictures of the ground and using machine vision and AI to separate the weeds from the crops, creating a 3D model of the field in real-time, according to Boyer. Titan’s AI system is trained on a database of over 500 million images of plants, Boyer said.
“Once you have the data, you need to make sense of it. You need to build AI models that are able to detect what crop species each plant is, and then to detect the exact geometry of the plant. It is likely not enough to only detect species; you actually need to detect more than that, like where the roots are, how many leaves, and so on to be able to get the precision that you need,” Boyer told us.
From there, Titan cuts the weeds out of the ground using blades similar to the ones used by hoes, but without the power of human labor, or the pesticides and herbicides typically used in the weed-removing process. Titan can be controlled through the use of a tablet or smartphone.
Keep up with the innovative tech transforming business
Tech Brew keeps business leaders up-to-date on the latest innovations, automation advances, policy shifts, and more, so they can make informed decisions about tech.
The company operates a weeding-as-a-service model, meaning farmers pay per acre for a Titan robot to drop by and weed their fields, typically committing to a certain number of acres per season, according to Boyer. FarmWise deploys the machine and an operator to client farms, taking care of the transportation, weeding, and maintenance of the equipment for the duration of the contract. Right now, FarmWise has a fleet of 12 Titan robots in operation, Boyer said.
The company said it works with some of the largest commercial farms in the US, including Taylor Farms, Mission Ranches, Dole Fresh Vegetables, and Ocean Mist Farms. FarmWise declined to share its revenue.
FarmWise’s robots can identify and cut a weed from the ground in less than half a second, removing weeds at 10x–15x the speed of a single farmer, according to Boyer. The company also claims its tech has comparable accuracy to manual processes, and “typically leaves less than 10% of the weeds behind while leaving intact more than 99.5% of the crops it works on,” per Boyer.
So far, FarmWise’s Titans are able to sort weeds out from crops like lettuce, broccoli, celery, cauliflower, cabbage, Brussels sprouts, and artichokes.
Looking ahead…This latest round of funding brings FarmWise’s total to over $65 million since its founding in 2017. Boyer said the money will be used to expand the scope of the 75-person company, training the Titans to be able to distinguish additional crops like tomatoes, peppers, and melons, while also expanding the company’s geographical scope into states like Iowa, Oregon, and Florida.
“We’re learning more crops, it takes time and money to capture images on different crops,” Boyer said. “Our dream is that any farmer in the world can pick up on this system and then use it the next day, that the systems are very easy to use, and they’re so robust that the person does not even need to ask himself whether this is going to work on this crop or this field with this configuration or not.”