Join the Newsletter

Stay up-to date with food+ag+climate tech and investment trends, and industry-leading news and analysis, globally.

Subscribe to receive the AFN & AgFunder
newsletter each week.

Dr. Vishnu Jayaprakash cofounder and CEO AgZen
Dr. Vishnu Jayaprakash cofounder and CEO AgZen. Image credit: AgZen

Meet the founder: AgZen’s Vishnu Jayaprakash on spraying, praying, and sticky droplets…

May 3, 2024

It’s no secret that not all the chemicals directed at crops hit their target, leading to runoff in waterways and buildup in soil, which is bad for the environment and even worse for farmers’ bottom line given the rising cost of crop inputs, says AgZen cofounder and CEO Dr. Vishnu Jayaprakash.

Multiple techniques have been deployed to tackle this ‘spray and pray’ approach, from spot spraying systems with sensors that turn selected nozzles on or off so herbicides are applied less indiscriminately, to high-tech smart-spraying systems attached to self-driving tractors that can precisely target weeds using computer vision and machine learning.

But the more basic problem, says Jayaprakash, who spent his PhD at MIT exploring how droplets behave on leaves, is that even when they are spraying on target, most farmers have no idea if they are spraying the right amount, and how much of that spray is staying on the plant and making a difference.

‘Feedback-optimized spraying’ from AgZen—a Boston-based spinout from the Massachusetts Institute of Technology (MIT) founded in 2020 by Jayaprakash and Professor Kripa Varanasi—aims to tackle this problem head on.

Its first commercial product (RealCoverage, available now) is a unit stacked with cameras that can be bolted onto a sprayer and link to a tablet in the cab to show the operator what is happening to the droplets in real time.

Based on that data and the conditions at the time, the AI-powered RealCoverage system can tell operators how to optimize everything on the sprayer, from which nozzle to use, to pressure, boom height and droplet size, to how fast to drive and how many gallons of spray per acre (GPA) is best for a particular chemical mix on a particular crop in a specific location. This can enable 30-50% reductions in chemical inputs, claims Jayaprakash.

Its second commercial product (EnhanceCoverage, launching next year) is a nozzle that cloaks droplets with adjuvants (rather than simply mixing them in), dramatically increasing the likelihood they will stick to plants, enabling farmers and growers to make further reductions in chemical use.

AgFunderNews (AFN) caught up with Jayaprakash (VJ) to talk about spraying, praying, and sticky droplets…

AFN: How did you get interested in optimizing spraying in farming?

VJ: I got introduced to spraying when I was a kid in India. I had a backpack sprayer and I would spray on my family’s farm near Chennai, where we grew mangoes and rice, and it was pretty awful. But growers do it because they have to.

Fast forward about a decade and a half and I got to MIT for graduate school and I met Professor Kripa Varanasi, who was working on making spraying more efficient because pesticide runoff is such a huge problem. So I spent my Master’s and PhD studying how droplets behave on leaves and how we can make spraying more efficient.

AFN: Tell us about the genesis of AgZen

VJ: AgZen was born out of more than a decade of research at the Varanasi research group at MIT, which has been studying how droplets behave when they interact with leaves and other plant surfaces since 2009. This involved creating new understandings of droplet dynamics and developing ways to make them stick to plants better without introducing new chemistries.

So growers commonly use adjuvants [substances added to a pesticide product or spray mixture to enhance the pesticide’s performance and/or the physical properties of the spray mixture], and we developed a new type of nozzle that cloaks droplets with adjuvants rather than mixing them in, and by coating them, they don’t bounce off the plants and end up as runoff.

So armed with this, we went out to growers and said, ‘Hey, we have this awesome new product that can save you a bunch of money because a lot fewer droplets are going to bounce off and you’re going to be able to spray a lot less!’

They came back to us and said, ‘That’s great, but what’s my coverage right now? Without knowing this, how would we even know how much better things would be [by using AgZen’s nozzles]?’ And that was the lightbulb moment.

When you spray an acre in a soybean field, you might generate 1.2 billion droplets. But where are all those droplets going and how many of them land on each leaf and stay there? No one really knows. I mean farmers know they’re probably wasting some product, but they have no idea how much. They just know that the label says you have to spray a certain number of gallons per acre, and they go out and spray that amount.

So that’s what inspired our first product RealCoverage, which is a bolt-on product that can go on any sprayer. In real time, it can tell you what percentage coverage you’re getting.

It has cameras that detect droplets as small as 150 microns on leaves and makes real-time adjustments to spray parameters such as pressure, rate, boom height or speed to maximize application efficiency while you’re driving as fast as 15 miles per hour. It then narrows the decision tree from hundreds of spray parameter combinations to a handful of optimal settings based on the tank mix, nozzles, crop and weather.

AFN: But why do you need to constantly vary parameters on an acre of soybeans? Would there be much variability?

VJ: Yes! There’s so much variability. This tool doesn’t just measure, it predicts what the steady state of your droplet is. So it adapts to temperature, humidity, and things of that nature. And then in real time, it’s adapting your speed, boom height, and pressure.

So let’s say in one area of the field, you have a certain leaf density here versus somewhere else where it’s slightly thinner. You don’t need to spray the same here as you do there. How many droplets are you getting on the plants here and then what’s the minimum amount you can spray for this part vs that part to still keep, for example, 20% coverage?

In the last few years there has been technology like spot spraying that helps you adapt to one of these input parameters, in this case, where are the weeds so I can turn the nozzle on or off [so you’re not spraying herbicides indiscriminately]. But this is only useful for one spray a year, basically that one herbicide [application] you’re doing.

RealCoverage works with any input from defoliants to foliar nutrition, fungicides, herbicides and insecticides.

We can already plug into the rate controller for some sprayers and automatically adjust the volumetric rate that’s getting sprayed on every acre using existing industry protocols. We provide recommendations for other adjustments to boom height, speed, nozzles, etc., and our customers implement the feedback as they spray in real time. However, our vision is to make spraying so easy that all growers would have to do is drive, and our systems will ensure they are getting the best outcomes on every spray.

Image credit: AgZen

AFN: Tell us more about AgZen’s second product: EnhanceCoverage…

VJ: EnhanceCoverage, which is coming out next year, makes every droplet stickier [by coating whatever has been sprayed with a ‘cloak’ of adjuvant] so you’re maximizing your efficacy.

Typically, companies are mixing the adjuvants [into whatever they’re spraying onto a crop] in the tank. But many of the droplets can’t get into the plant quickly enough before they bounce off the leaves.

By cloaking the droplets in the adjuvant, the leaf is seeing pure adjuvant, if you like, and the droplet is far more likely to stick. Our in-field data on EnhanceCoverage shows a 1.5-2x improvement beyond what RealCoverage unlocks. 

AFN: How are you cloaking the droplets with adjuvants?

VJ: We basically switch people’s nozzles and add a direct injection system, so every spray nozzle just gets a different tip and a different injection line so that the adjuvant is coming in in a separate line [when they are sprayed out together, the adjuvant coats the crop input]. But we’re not developing new adjuvants, our tech works with products companies are already using.

Every grower knows how much it costs to replace a nozzle tip. They do it all the time. And direct injection systems are also well-known, so the cost of implementing this is accessible for a grower.

We’ve already retrofitted many spraying machines with this unit, but we’re not launching it this year for a couple of reasons. One, we want it to be universal no matter what sprayer you have. And second, we want growers to get comfortable with using our first product, RealCoverage, first.

AFN: What’s involved for farmers and how tough of a sell is this?

VJ: RealCoverage is a true bolt-on product that can go on any sprayer – be it a self-propelled sprayer, a tow behind, a 3-point hitch unit, or even an airblast sprayer. It typically takes about three hours for our team to install a RealCoverage unit onto a large self-propelled sprayer. 

It’s also easy to operate. The interface runs on a tablet that growers can start using the same day their installation is completed and our installation teams also stay at each install location for up to five days. We ride along with operators as they are spraying, help them establish how effective their current spraying practices are, and show them how they can unlock efficiencies right off the bat before letting them explore optimization across their farms.

As for ROI, this is going to pay for itself within a year, but there are many different aspects to the payback. They are reducing chemical use and water usage [crop inputs are typically mixed with water before application], and reducing the time spent on the field as you can drive faster sometimes. But they are also getting better outcomes. They are saving 30-50% on their chemical costs and increasing yields.

For the first time we’re seeing where these droplets are going and how that correlates to an outcome. And then farmers can start asking deeper questions. For chemical A, how much coverage do I need for this outcome vs for chemical B?

AFN: How much progress have you made to date?

VJ: We’ve already signed leases for over 60,000 acres across the US from Georgia to Iowa to the Midwest this year, so we’re already generating revenue. You can buy them upfront or we can do lease-to-own contracts [whereby the equipment is leased with the option to buy at the end of the lease term].

In 2023, we performed fallow field burndowns, weed control in soybeans, defoliation in cotton, and disease control in a number of vegetable and fruit crops, including vineyards in California, Italy, and France. 

In 2024, our units are being leased predominantly for row crops like cotton, soybeans, and corn. We are also doing a few thousand acres of peanuts.

We’ve shown our system works in row crops at 15 miles an hour, but we also have customers who’ve saved up to 50% on fungicide sprays in vineyards. Trials with universities including Texas A&M, Cornell, and UMass have also validated that we can save 30-50% on foliar sprays.

AFN: What’s the longer-term impact of this greater precision and feedback loop on chemical use on farms?

VJ: Longer-term this can create value for everyone in ag. So an OEM can utilize this to make its machines better; chemical companies finally know how much product is actually on every leaf, so they can make their chemicals better; and regulators can learn from this [more precise data on the percentage of pesticides that reach their target can help regulators make more informed decisions].

AFN: How have you funded the business so far?

VJ: We’re backed by a fantastic venture partner, Material Impact, which funded our $3 million seed round in January 2022 and we are looking at a Series A later in the year, but not for technical de-risking. Last year we did 12 trials and pilots with large commercial customers and had our product validated by three leading agricultural universities and we already have a suite of patents, so it [raising money] is all about commercialization.

AFN: Was your IP developed at MIT and then licensed to AgZen?

VJ: Some of our initial background IP was [developed at MIT and licensed to AgZen], but a lot of our IP has been developed since we spun out. Every product is protected by its own patent range, some granted, some in applications.

We’ve pioneered this approach of feedback optimized spraying and what’s enabled us to do that is decades of expertise in understanding how droplets bounce, but also how droplets stay, spread out, and evaporate. And so all of that has given us a unique advantage.

Join the Newsletter

Get the latest news & research from AFN and AgFunder in your inbox.

Join the Newsletter
Get the latest news and research from AFN & AgFunder in your inbox.

Follow us:

AgFunder Research
Join Newsletter