Skip to main content

‘The Android of agriculture’: Monarch Tractor CEO Praveen Penmetsa on the future of farming

Monarch Tractor’s Praveen Penmetsa has a grand vision for agriculture, and it includes autonomous electric smart tractors powered by AI.

Share this story

We spend a lot of time here on Decoder talking about electric vehicles and the future of cars, and we’re usually talking about passenger vehicles or maybe cargo vans. But there’s another huge industry that can also reap the benefits of electrified transportation: agriculture. 

I co-hosted the Code Conference this week, where I had the opportunity to hang out onstage with Monarch Tractor CEO Praveen Penmetsa. Honestly, this was one of my favorite conversations of the entire event. 

We are utterly reliant on farming as a species, and farming is utterly reliant on tractors. If we don’t have tractors, we don’t have food. But electrifying farms is hard, and Praveen explained how he and Monarch are trying to tackle that challenge.  

Praveen and I managed to cover a lot of ground in a relatively short amount of time onstage. Farming is a very old business, and it’s hard to make big changes. You’ll hear Praveen say the average farmer only gets 40 shots at iterating their process — 40 seasons of growing. And Praveen sees a lot of opportunities for farms to make the transition to electric autonomous equipment without disrupting those 40 cycles of agriculture. And of course, there’s a lot of opportunity for his business to benefit from that transition. 

Monarch has three ways of making money, he told us. There’s the tractor hardware, of course; selling the actual vehicles. There are subscription fees, and Monarch charges a recurring fee for its software platforms, which is smart tractors as a service. And then, there’s licensing — Monarch has a deal with Case New Holland to use its platforms. The ambition is to compete in an open way with closed platforms like John Deere, and Praveen said his goal for the Monarch platform is to become the Android of agriculture.

Okay, here’s Praveen Penmetsa, the CEO of Monarch Tractor. Here we go.

This transcript has been lightly edited for length and clarity.

Hello, sir. Tell us a little bit about Monarch and what’s going on with this tractor

So our mission at Monarch Tractor is about how we make farming sustainable and farmers profitable. It might come as a surprise to most of us that less than half the farmers in the country actually make any money, less than 42 percent. And most of our farming is not farmed sustainable in terms of not just chemicals but also in terms of how we use our resources — everything from water to fertilizers. 

The way we decided to go about it was to build a small tractor targeting the fruits and vegetables farmers. That’s where most of our food comes from. And it’s electric, it’s autonomous, and it’s smart. And what this allows the farmers to do for the first time ever is to have a tractor that generates new revenue streams for them, using all the data that it collects while doing operations on the farm. So it’s a familiar form factor. 

So when I look at this tractor, I say, “Okay, we can sell it to a lot of people, still got a steering wheel. It’s got a battery in it.” But the thing is festooned in sensors, go look outside. It’s covered in cameras and sensors. It has machine learning embedded in it. You are headed toward autonomy. You have a software platform that can control multiple tractors at once. 

There’s a long way between fruit and vegetable farmers and someone who has 1980s things autonomously driving over a connect platform. How are you doing right now? It’s been just a little over a year. Where are we on the journey?

So we launched our tractor commercially in December. Now, we are the only electric autonomous vehicle that one of you can go out and buy. I’m very proud of that fact. 

You can do whatever you want on a farm. On a road, it’s a problem.

Exactly. The fact that we are out at commercial scale… the fact that we have Foxconn — the maker of our electronics devices, building the smart electric machine for farming for tomorrow, today out of Ohio — is a big deal. So we have more than 200 of these machines out on farms doing activities. 

Those are customers, they’re not test units. You’ve sold 200 of them?

They’re customers. Last month, we delivered 75-plus tractors to customers, trained them, and walked away. So they’re using electric autonomous machines now. 

And are you collecting data back off these tractors as the customers use them? 

That’s the great part is because the tractor sits at the center of every farm, we get to see every operation from day one, when the farming activities start, all the way until post-harvest when things stop. So we get data not only on what’s happening out in the farm with our cameras but the who, what, where, when, and how. Most importantly, the how — how are they farming? All of this comes in from our vision systems: 3D cameras as well as standard cameras. And then we use AI and a lot of data scientists to put this data into some structured data lakes that we can yank insights and reports out for the whole ecosystem. 

When I hear that sort of thing, my mind leaps to, “Okay, you’ve built a proprietary data system, the farmers are locked in, they’re going to end up in some sort of subscription contract with you, and the tractor is just the beginning of a long, potentially tortured relationship.” And I say that because that’s exactly how John Deere works, right? And lots of farmers are saying, “Okay, John Deere collects a lot of data. They’re going to sell us seeds.” All this happens. Is that your business model, or are you going to be more open? 

No, we are completely different from that world. That was the whole reason why I started Monarch Tractor was to change that. Right now, farmers can’t even repair their data or do diagnostics on their tractors without approval from the OEM. What we are doing at Monarch is we are licensing our technology out to all the tractor companies. One of our big partners, Case New Holland [CNH Industrial], is the world’s second-largest tractor company after John Deere. We have given our technology to them, so they sit on our data stack. 

So we are basically becoming the Android of agriculture, and the whole ecosystem can access our data lake through our APIs and provide reports. We are very happy to share everything that comes off of our tractor with everybody in the food ecosystem, not just the farmer. It could be you and me. Today, we live in a world where you and I know more about the DoorDash person that delivered our food than the person who actually grew our food, and that’s what has created this vicious cycle for us in our food ecosystem. 

When you think about that approach, the Android of agriculture, it’s a good line. Just to extend the analogy, at the high end, the iPhone does not seem all that threatened by Android, right? Are you making an “open always wins” kind of play here? Or are you saying there’s going to be a huge market, and we’re going to coexist with a more closed market? 

No, there is going to be a huge market. Today, we only have 3 million tractors sold every year, and we have 2 billion-plus acres of cropland. I’m not talking about grazing land and stuff — cropland. By comparison, 70-plus million cars are sold every year. So it tells you the disparity and also the opportunity for agriculture smart machines. 

So long story short, in the next 10 years, we’ll need a lot more of these machines if you’re going to stop using chemicals and do more operations on the farm. So this 3 million is going to grow. It’s already been growing for the last five, six years. This market is going to blow up. There’ll be new entrants into this market. We are seeing that. We are seeing people from the food ecosystem come in, not just the equipment ecosystem. We are seeing input companies like fertilizer companies are now building devices. So this whole market is going to blow up, and our whole idea is to allow this to happen by giving them the technology that allows them to build smart electric autonomous machines for farming. 

So, is that your revenue stream — the platforms and the technologies — or is it actually selling the tractors?

So, right now, we make money in three ways. We sell tractors, yes, we make money on the device side. We also sell subscriptions to our Connect and Automate platforms, which is the digital subscriptions that farmers pay us for. But the third revenue stream is our licensing revenue stream. Now, Case New Holland is our first one. They’ve already announced two tractors powered by Monarch technology. And now, they’re using our verbiage, which is an open ecosystem to differentiate and bring a new play to the ag market compared to what John Deere is doing. So long story short, three revenue streams: hardware, digital subscriptions; and / or your licensing. And we have unlocked all three already. So these are not hopes and dreams. They are recognized revenues sitting on our financial statements for all three. So the market is ready.

I grew up in Racine, Wisconsin, driving past the Case factory every single day of my life. You can see why I’m obsessed with tractors. Let’s talk about the actual physical challenges of an autonomous EV tractor, whether it’s yours or someone using your platform. It’s all the same problems as cars, right? Yesterday, the Rivian CEO was here. Somebody asked him, “What happens if this thing lights on fire?” That’s hard on a road. What happens if it happens in a field? How are you protecting the batteries? How are you sourcing the batteries? How’s that going to work?

So a couple of things there. Number one, ag is very challenging in terms of robustness and durability. The shock loads that a tractor sees on a day-to-day basis would set off the alarms on your car, set off the airbags, and would turn off your battery system right away for safety. We see that on a day-to-day basis. The amount of movement also that we see — whether it’s in terms of climbing hills, or our roof system when our GPS moves around two and a half feet sometimes. So how do you get a couple of inches of accuracy when the tractor platform is itself moving feet?

So, there are a number of technical challenges, number one. Number two is, on the safety side, we are able to leverage all the work that’s been done — both on the EV platform as well as on the autonomous platforms. So we get to stand on the shoulders of all the people that have done this work in urban mobility but take it in a different way. And the challenges we face in ag, other than the robustness side, are margins. Farming is a very low-margin business, which means we have to provide value for the farmer as a product from hour one.

If you tell farmers the payback is four years, you and I might be amazed by it, they’ll walk away. So payback for farmers has to be in months for them to actually buy this. So we had to make sure that the tractor saved money on diesel. Autonomy was a big unlock. Tractor drivers are very skilled operators; if they get a job wrong, it shows up in the yield six months later. So those are the kind of challenges that we’re solving today, and those are all product challenges that we ran into.

Is there a take-up from farmers on autonomy? Are they saying, “Fine, let the tractor drive itself”?

Right now, I’m amused to see some of the numbers from the automotive OEMs. But we actually have a 30 percent attach rate on our autonomy. I’ll tell you what that means. That means 3 out of 10 of our customers are paying an annual subscription for autonomy today worth $8,000-plus. 

And you said you hand the tractor over, you walk away, good luck. How do you make sure they’re safe? On roads, you have rules, you have laws, you’ve got a Department of Transportation. On a farm, what happens if there is a tractor on the loose?

So, that’s where the safety regimens kick in. Yes, everything is still geofenced. We have a whole deployment team that helps them on day one. This is very much like deploying enterprise software. Where we walk in there, and we say, “We’ll train your whole team. Let’s set it up. We’ll help you just take an outside geofenced map of your block. We’ll tell you what the first row, last row, middle row looks like, check your connectivity. And then you can start to use operations.” If in doubt, the whole thing comes to a stop. It’s supervised autonomy. It’s not tele-op; it’s supervised autonomy. So the supervisor can jump in and keep track of six to eight tractors at a time and say, “Oh, that thing is asking me for some information.”

This is your supervisor as part of the subscription, or this is the farmer?

Farmer-supervised. When I say supervisor, we have elevated the existing tractor driver. One thing that I was blown away by 10 years ago was just by looking at how Uber scaled in countries like India. I went there, and this person who came in from the village who didn’t speak the same language I did was able to drive a car, take me to my destination. The whole transaction happened, and he was able to make money. If they can do that, can we elevate an existing tractor driver to be a supervisor of a fleet of six to eight tractors with the right technology abstraction? Yes, we can, and that’s what we’re doing.

Farmers are obviously very concerned about durability, about repairability. These are the things that people are concerned about with EVs. How durable is this battery? How durable is this machine? How expensive is it to fix?

So, that’s a great question. Tractors are very often used for long periods of time for two reasons. Number one, there’s no new value that you get by getting a new tractor. It doesn’t give you something that the old tractor doesn’t, especially in the small tractors. So people tend to hang on because the new machines are actually less reliable, thanks to all the emissions equipment. The second thing is, essentially, they’re looking for a payback here. So, the value proposition comes from the tractor maintenance costs overwhelming you. That’s why they swap tractors. 

With our tractor, our battery comes with an eight-year warranty. We understress the battery right now. This is an energy play and not a power play, number one. Number two, there’s no DC fast-charging on farms because the electric infrastructure doesn’t exist. So we take advantage of that on the product side to increase the longevity of the pack. But also, more importantly, we have built this to be sustainable. If you look at our tractor, when you go out and check it out, you’ll see that the battery is swappable, which means tomorrow, when there are new chemistries, we’ll just swap the battery out. You don’t need to sell your whole tractor and buy a new one, number one.

Number two, the roof system, which is completely self-independent — if some of you walk out there, do pay attention. There are only two cables and four bolts that hold the whole roof down. The two cables are data and power. So that whole roof can be upgraded two, three years from now with a new sensor suite, new compute, and new connectivity. All of that is in the roof. So what this means is, farmers can hang onto the platform, upgrade the roof if they want to, upgrade the battery if they want to, and get even better lifetime from our tractor than what a traditional tractor does today.

We have questions beginning to line up here. I’ve got a couple more for you. You mentioned DC fast charging. Charging is a pain on cars, on roads with infrastructure. How are farmers managing this? What’s the range? What’s the longevity of the tractor?

So, that’s one of the big challenges that is actually preventing the energy transition of farming: electric infrastructure like you’re pointing out. What the tractor is doing is, for the first time, if a farmer buys 10 of our tractors, that’s a megawatt-hour of energy. If they deploy solar on the farm, which we are now enabling them to do as a package, they get even better payback both on the solar side as well as on the tractor side. The reason for that is, our tractor actually does triple duty. It’s a tractor; it acts as a mobile genset. So for the first time, the farmer can take a tractor to the middle of the field, plug stuff in, and power stuff off of their tractor, just like what the Ford Lightning does.

More importantly, while it’s standing at your barn connected to your solar, it acts as energy storage, making you money as a farmer. So in California, there’s a program where farmers can make money by putting this energy back into the grid. And also by demand response programs, which utility companies love in rural areas. They might not like it if you do it in downtown San Francisco. But in rural areas, where there’s limited supply, they like that. So this triple play is a great product view for us to allow farmers to get more utilization and more revenue streams from a tractor. So that’s how we are addressing the limitations of infrastructure. We also have a swap cart that has its own charger. So you can actually take it in the middle of the field, swap the battery on our tractor, and keep going 24/7.

Last thing you mentioned, this is not just happening in the United States, particularly on the energy side, you’re expanding much larger.

Yeah. Our goal from day one has always been that this is the device. Every device on our farm needs to be zero emissions, needs to be some form of automation with the lack of labor that we have in farms, and needs to be smart. So you and I know how our food was grown and also we can allow farmers to make more money. With that goal in mind, we built a tractor that will scale to every farmer. Our goal was a farmer in Africa or South Asia can eventually afford one. 

So I’m super excited to announce that we have just signed a partnership deal with the largest clean-energy infrastructure company in India, who is going to deploy our tractors in India at scale as a service so that small farmers who have two or three acres can just rent a tractor when needed from the infrastructure deployment and then use it and return it. And that, for me, is going back full journey to taking this tractor back to countries like India.

That’s amazing. You heard it here first. 

Audience Q&A

Nilay Patel: Alright. Questions. 

Daniel Vestergaard: Thank you for this. This is so fascinating. I’m Daniel from DR [Danmarks Radio]. A bit more elaboration on this. Getting a new device like this into the hands of farmers because they are so dependent on interconnectivity to their old equipment. And sometimes they also share the tractor between farmers with all this equipment. And this tends to be very conservative because they know what they get. How have you broken those barriers?

Praveen Penmetsa: That’s a great question. And that comes to our big differentiator as Monarch Tractor compared to our competitors. You made a great point, which is, we cannot change farm operations. People don’t understand that farmers get 40 iterations; forty tries in their lifetime at farming. That’s 40 seasons. If you ask them to say, “Hey, can you try something new?” That’s one down, right? 

So, what we have done with our tractor that is very different to everybody else is, we tell our farmers, “You can use the same implements that you have on your farm today. You can farm the same way. All I’m asking you to do is to swap out your tractor for ours. That’s it.” There’s no operational change: same implements, same attachments, everything else remains the same, same driver.

So, that has been the key to the transition. The reason other people don’t do it is that it’s crazy hard. We have now had to make sure that our hydraulics, our gearboxes, all the implement attachments, everything, can be bi-wire controlled and can be controlled from the cloud, and we get full data on each one of those things. That’s the deep technology part of this simple-looking tractor.

DR: When are you rolling out in Europe?

PP: So, we actually have two tractors that have hit the ground in Europe this week. The first one is being piloted in Italy by the end of next week. We also have a tractor in New Zealand now. And that’s the great thing about tractors is you can go anywhere, unlike a car. There are no regulations in compliance. If you guys are trying to get your device into market and you want to skirt regulations and compliance, call it a tractor.

Jay Peters: This is Jay Peters with The Verge. How far away do you think we are from some kind of federal right-to-repair legislation passing for agricultural equipment?

PP: So, there are two things there, just to unpack that a little bit. The reason there’s a right-to-repair issue today — and I have a lot of sympathy for all the diesel companies that build tractors — is that’s all emissions regulations-driven. You cannot let somebody monkey with your emissions equipment that is not certified to do so. So the tractor companies are struggling with that part. But the key here is, once we transition to electric smart devices, where the data can be open and it’s not emissions-driven or compliance-driven, the data pipe should be open.

So that’s the dichotomy is, you’re asking me to answer questions on will the government allow farmers to monkey with emissions equipment on the farm? They won’t. Not going to happen. But if you go to electric zero emissions, and it’s not emissions-related anymore, farmers should have every right to modify their tractor, come up with their own attachments, train it, and use it, which is what our tractor does. They can make their own custom implements, attach it to our tractor, run some training logic. We train the platform and start operations.

NP: Alright, very quickly. One more.

Audience member: Two different points. I was curious about your road map on the battery technology. Again, we see it on the renewables side of just extremely volatile, high degradation, extreme variance in heat, farming is in heat. So just curious around that. But also curious just around your technology that you’re seeing on how you can help local farmers. What data are you already seeing that’s a success story that’s making lower labor or lower water issues or lower pesticide use?

PP: Thanks a lot. So, the first one — I’ll make it short — is the battery side of it. There’s a lot of work going on on batteries. But our whole goal is, how do we give maximum energy to our farmers with maximum lifecycle? Which is what you talked about is, how long will this battery last? And for us, we don’t need any advanced chemistries there. The existing chemistries, it’s a matter of getting the cost down to where the farmers can afford it and enough energy. Our tractor has more than 100 kilowatt-hours of battery on it, and we sell that whole thing for less than $90,000 today. It’s pretty crazy. I think we can do more because we want to sell our tractor for $10,000 to $15,000 in India. That’s the challenge on the battery side.

On the safety side, we’re able to leverage a lot of the good work that’s happening on automotive and energy storage. So, we try to productionize that instead of starting new chemistry work. So that’s that side of it. You brought up a great question. I’ll give you two small, short snippets of how data’s been amazing. We noticed that when we deployed off our tractors to farmers, they started using it for all this mowing stuff. And we noticed that one of the farmers was getting nine-plus hours, and the other farmer was getting four or five hours.

Historically, they would not have noticed it. It would have shown up as diesel consumption. Nobody cares about diesel consumption on a tractor, right? But what we noticed was like, “Hey, why is there an energy delta?” Our cameras that recognized the mower said, “Hey, we can see it. The mower settings are off. Yes, they’re still mowing, but they’re using a lot more energy for a simple operation.” So we were able to, using AI, using vision cameras, say that your operation can be more energy-efficient for the first time if you do these small adjustments, and for the first time, your operator now gets feedback.

Historically, nobody would have noticed it. Maybe at the end of the month, they would be like, “Why is our diesel bill a little bit higher?” That’s a big change. The second big thing is, we gave our tractor to our farmers, and we said, “Hey, what do you want to use it for? You should probably mow with this.” And they were like, “No, I’m going to attach this weeding machine that’s going to cut grass between these vineyards.” We are like, “Why?” And they were like, “Well, we always wanted to use this machine, but we could not afford the labor cost of running this five times in a season. So we just spray chemicals right now. Once we spray it, it’s done for the season. And the chemicals are expensive.”

So the best example for us was, from a data standpoint, they are now able to see where they want to spray. They spray only in those places if there’s a problem. In every other place, they use our mowing machine, the under-the-line weeder, to basically cut the grass instead of spraying expensive, dangerous chemicals. Huge savings for them on the money side. They’re able to use the data to figure out where and how to spray. They’re able to use the data to run their operations for more efficiency. And this is happening in month three of our deployment. And that’s the power of AI and vision coming off of the cameras off the tractor.

NP: I told you people: tractors. You didn’t believe me. I saw the doubt in your eyes. Praveen, thank you so much. That was incredible.

PP: Thanks a lot.

NP: Go look at the tractor outside. It is so cool. 

Decoder with Nilay Patel /

A podcast about big ideas and other problems.