In this episode of Hardware to Save a Planet, Dylan is joined by Shely Aronov, Founder and CEO of InnerPlant. This Agritech company turns crops into living sensors for a triple win of improving biodiversity, impacting climate change, and improving agricultural yields.
Join us as we cover multiple touchpoints around how addressing agricultural inefficiencies positively impacts climate change. Listen to the podcast to learn how satellite imagery, hyperspectral cameras, and collecting genomic data can help detect plant health and reduce chemical usage for a cleaner and greener environment. Learn how InnerPlant is overcoming the challenge of balancing agricultural yield with social impact and environmental sustainability.
Shely Aronov is an experienced entrepreneur who thrives on solving complex and meaningful problems. She believes no other industry is more critical to our society than agriculture. Shely’s mission at InnerPlant is to develop tools that will help feed the world sustainably while driving a positive environmental impact.
To learn more about sustainable agriculture that uses signals from plants to cut the use of chemicals while increasing yields, check the key takeaways of this episode or the transcript below.
Key highlights
- 09:24 – 10:48 – An overview of InnerPlant’s technology and process – Plants are susceptible to environmental changes and have a highly evolved way of reacting to these changes. For instance, a lack of Nitrogen in the soil causes the roots to go deeper. These changes are triggered by a change in the plant’s DNA. InnerPlant has developed a technology that adds a line of code to the plant’s DNA. Every time an environmental change stresses the plant, the plant produces a protein that creates an optical fluorescent signal that can be monitored remotely, even from satellites in space. Also, when the plant produces its seeds, the seeds’ DNA will also carry that line of code, which makes the change self-propagating.
- 12:39 – 15:18 – What is the environmental and agricultural impact of the technology? Based on statistics, 30-50% of fertilizer is over-applied, while 50% of pesticides are misapplied, meaning they’re applied at the wrong time. The combined impact cuts the agricultural yield by up to 40%. It also affects biodiversity as the over-application of fertilizers kills the soil’s ability to sequester environmental carbon. By enabling the plants to define when they need fertilizers and pesticides, we’re protecting biodiversity, addressing climate change, and raising agricultural yields—a triple win.
- 18:05 – 19:20 – The InnerPlant business model – The company is building an ecosystem of partners to deliver the solution to the farmers. The simplest and most scalable solution is to sell seeds that have the code embedded in them. The farmers can then work with the partners downstream, who would monitor the signals from the plants and update the farmers on the required follow-up actions. Initially, the company is selling the seeds, but once the ecosystem is in place and mature, they want to own and monetize the signals and the data.
- 22:26 – 24:09 – An under-the-hood look at how the process works – Shely explains the process using the example of the soybean seed. The seed carries a code to trigger anti-fungal fluorescence. If the plant is attacked by fungus, the signal is triggered and captured by satellites in space. The satellites can capture the signal at a one-acre-per-pixel resolution. The monitoring station then advises the farmer on the acre that needs attention. John Deere sprayers are retrofitted with sensors that can monitor the plants within that acre, emitting the signal and spraying only those plants.
- 27:25 – 30:06 – A peek at the solution’s hardware – As a first step, the fluorescence must be detected, and surface-based detection would require additional light sources and cameras. However, when detecting from space, the company used electrophysiology techniques to develop a sensor that could detect the fluorescence in sunlight. The cameras on the satellite are programmed with custom algorithms to detect the signals, which are then relayed to a ground station for evaluation and further action.
Transcript
Dylan: Hardware to Save a Planet explores the technical innovations that are giving us hope in the fight against climate change. Each episode focuses on a specific climate challenge and explores an emerging physical technology solution, with the person bringing it into reality. I’m your host, Dylan Garrett. Hello and welcome to Hardware to Save a Planet. I’m here with Shely Aronov, the CEO and Founder of InnerPlant. The broad topic we’re discussing today is agriculture, which is an important one in the context of climate change. Agriculture is responsible for something like 25% of global greenhouse gas emissions. Inefficiencies in agriculture also have economic and other environmental impacts. As much as 30% of pesticides are wasted each year, costing an extra $250 billion and impacting our air, soil and water. Despite that, we still lose as much as 40% of our crop yields every year to pathogens. Shely and InnerPlant are tackling those inefficiencies with a very cool approach. They’re turning plants into living sensors to communicate their specific needs. I’ll leave it at that and let Shely share the details. To introduce her quickly, I’ll just say that it’s clear to me she is a natural leader and has that entrepreneurial spirit I’m always so envious of. She’s also passionate about improving our food systems. Prior to founding InnerPlant in 2018, she had founded a company making delicious and healthy hummus. And I’m very excited to hear how that led to founding InnerPlant. Shely, it’s really an honor to do this with you. Thanks for being on the show.
Shely: Thank you Dylan, that was a really nice intro.
Dylan: Well, I’ve been excited to talk with you. The other thing I didn’t say is I just learned you’re also a podcast host.
Shely: I am.
Dylan: The CropTastic Podcast. So that’s cool. I think you might be my first guest who’s also a host. Makes me a little nervous. I feel like it ups the ante a little bit.
Shely: Yeah, perfect.
Dylan: Can you tell me a little bit about CropTastic?
Shely: Yeah, it’s our podcast that we do for InnerPlant messaging. It’s about the future of agriculture. And the idea is to cover things that are not the mainstream opinion. So nothing that is the future is going to be an extension of today, rather more unique ideas that can get us to where we need to go. And then we cover it from the perspective of founders, farmers, and then industry, executives and thought leaders, scientists.
Dylan: So yeah, I said you started a Hummus Company. First of all, how did that happen? And then how did that lead to InnerPlant?
Shely: Very connected. First of all, great being here today. I’ll give you the quick rundown of my history. I’m from Israel originally, studying Industrial Engineering, started in Consulting. My career, I guess a lot of people have. And then I got an MBA from Stanford, which brought me to the US. I had a couple of opportunities around that. I started my first company, it was in Marine Construction. We were building a sewage system for the city of Cartagena, Colombia. And that was a super cool experience that I was probably not ready for at 26, but I did it anyway. And shaped me a lot, probably more than the next company. With the Hummus Company, it was more, I wanted to do something. At the time, I just wanted to take a little bit of money and then start something and make it self-sustaining. And I thought that that was my dream after graduating business school. It was something that I could grasp. I didn’t want to get into the VC thing. And I knew that we could make better hummus than Sabra could. And just like being a good recipe from Israel and making it a better way. By the end of that journey, I actually decided that’s not what I want to do at all. I don’t want a lifestyle business. I want to change the world. I want to make a big impact. I want to go big and still want it to be in food. But the solution was, let’s go into agriculture because the impact is in the field more than it’s on the shelf. And then technology heavy. Bringing something that is really unique and can make our customers successful and happier, rather than something that’s incremental. And at the end of the day, better hummus is incremental.
Dylan: And when you looked at agriculture, I’m curious how you looked at the whole space and how that ended up landing on what you’re doing with InnerPlant.
Shely: Yeah, so I’m going to admit I had a very unique lens. I came across the concept of biosensors right when I decided to sell this company. So literally I made the decision in April and then by June I came across this concept from a professor and I thought, wow, this is super intuitive. And if there’s a way to make this, there might be room for it. But I didn’t know anything about Ag. I’m a complete outsider to this industry. So I studied and I talked to a lot of people, starting with farmers. So number one, just talking to other farmers. And what’s interesting is if you go to a farmer and you’re like, hey, I have a way to help plants tell you what they need. Is that valuable? They’re like, yeah, plants have all the information. We just don’t know how to extract it from them, right? We look at them, we hope, and it doesn’t work. So that was the beginning. I would say from there, for me, it was all industry research. What are the technologies that have been successful in the past in Ag. ? And interestingly enough, there’s only been a couple in the last 30, 40 years, genetically engineered seeds and GPS driving tractors. So we wanted to go after a technology that has successfully been adopted. We wanted to choose the same reasons that it was adopted. So we really wanted to understand why. And then I wanted to scale. I have an aversion to anything that is small scale. We started with tomatoes, there were some ideas about grapes, all that is great. I love produce, but at the end of the day, there’s 350,000 acres of tomatoes in the US, there’s 85 million acres of soybeans, and one beats the other. So scale was the main focus for us from day one.
Dylan: That’s a cool story because it sounds like you were an outsider to agriculture and the core technology as I roughly understand it involves like biology and genetic engineering and stuff, which is also at least scrolling through your background seems like not something that you…
Shely: Not my thing.
Dylan: It sounds like you had a lot of some aspects of the business to come up to speed on. What was that process like of talking to farmers? I’m curious as an outsider, just maybe a stupid question, but just like the tactics of that. How did you get those connections and those conversations to happen?
Shely: So here’s a little secret about farmers. They love to talk. They admit it themselves. It’s not hard to get a farmer to want to talk to you. And if you’re going to go out and travel and see their fields even more so, they’d be happy to take you around, show you the ground, show you what they’re doing. They’re more than happy to open up everything and share. So it’s been easy. And I think it’s just necessary because we have one job to make our customers happy, right? If we don’t do that, then as a business, we’re going to fail.
Dylan: You said the core technology you heard about first from a professor, is that person involved in the company?
Shely: He was involved originally. There’s actually two professors. One is in molecular biology. The other one does nanoelectronics because the first concept we came across was very different from what we’re building. It was building little sensor clips that would sit on the leaf. And when applied to a substrate, we’d create a chemical reaction with a gene that we’re producing in the plant to create an electrochemical signal. And as you can imagine, that is not easy to commercialize at all. So we spent some time at the university. They’re amazing. It’s actually in Tel Aviv University and we spent some time learning it and we’re fascinated by it. It was a very unique concept, which is the reason they could publish about it. But the reality is that biosensors have been around for a very long time. And when we look at it from a commercial point of view, the question to answer is, what’s the most mature, easy way to do this? And the answer is fluorescence, which is what we ended up doing. So we understood very quickly that if we want to work on scale, we have to work remotely. And as remote as satellite is very likely necessary for millions of acres. Nevermind, it’s never going to be a sensor on the leaf. There’s a few ways to do even imagery-based solutions. You can do pigmentation. You can make the plant change its color. You can do bioluminescence. You can make it glow or you can use fluorescence. And fluorescence was best for many, many, many reasons. Some clear, like you don’t want your plants to look sick or glow. And some of it you can do different problems using different fluorescent signals, right? So very rapidly we moved to a different solution and to answer that question then at that point we also progressed from the lab because they’re working on different products.
Dylan: And just to explain what you just said, the reason fluorescence doesn’t have some of those other problems is because it’s not in the visible spectrum. Is that what you’re saying?
Shely: Yeah, fluorescence is a new light. So it’s not visible. And in order to see it, you have to excite it with another light. When you’re creating another, you’re basically creating little light bulbs in the plants. And that’s what’s unique about it. You can’t see it by eye. You can’t just see it even with equipment. You need to run algorithms. You need to know what you’re looking for. And I think that is actually a really good thing when you’re modifying your food.
There are two challenges to developing this technology. One is you have to make the plants create a signal, and the other one is you have to collect the signal, and both are equally challenging.
— Shely Aronov
Dylan: I’m a mechanical engineer. I have no particular biology expertise. Is there any chance you could explain what you’re doing to the plant, how you’re modifying the plant and how that works in a way that you think I could understand here?
Shely: For sure. It’s actually easier for me than explaining how we detect, but we’ll get into that too.
Dylan: Okay.
Shely: Hard things for us, by the way. One is you have to make the plants create the signal and the other one is you have to collect the signal and both are equally challenging, interesting. So the way it works is based on nature. Plants have a very evolved way to react to their environment because they’re immobile. They’ve learned over a millennia how to protect themselves from different things that are attacking them. So for example, if a plant is eaten by bugs, it’s going to start producing a protein in its leaves that makes it taste bad to repel those insects. If it doesn’t have enough nitrogen, it will literally mobilize its roots deeper into the ground so you can extract nitrogen better. And all these things are early, they start within 6 or 12 hours of the initial stress, and they’re different because the plant reacts differently depending on what’s happening to it. But it’s all invisible to us because these are gene expressions, these changes happen on the DNA level so we can’t see it. So what we do is we teach the plant, quote unquote, with a line of code, when you’re reacting to the stress, start to produce a new protein and that protein creates the optical signal fluorescence. So it’s a benign protein that we teach the plant to create and that protein is connected to the thing that’s attacking it.
Dylan: Oh, interesting. So you’re actually able to, the plant will provide a different signal depending on the type of stress it’s experiencing. You can know whether it’s nitrogen deficient or pathogen stress.
Shely: Exactly. Yeah, and that’s necessary because you need early and you need specific to be able to have it actionable at the end of it.
Dylan: Okay, you said insert a line of code. Tell me more about that.
Shely: I mean, that’s gene editing. So that’s the part that even if I try to explain to you is hard to do, but that’s a difficult one. It’s just that the easiest way to think about it is it’s a line of code. We basically insert something into the plant’s DNA that teaches it how to produce this new protein. And we connect that to the existing gene expression because you want something to turn on that light bulb and then to create the new signal. And the beauty about it is that it is just there for the next generation. So you have to make the first plant that has this line of code, but then the problem is we’ll just have that line of code in them. And it’s just in the future of those plants when farmers plant seeds, where those are going to be seeds that can communicate.
Dylan: So CRISPR gets a lot of buzz lately as like the Copy Paste Tool for Genetics. Are you using CRISPR?
Shely: We’re not. We’re looking into it. I think it’s interesting. I’m of two minds on this. One is there’s a lot of possibilities. It’s a really cool technology, but the other one is that it’s at a very early stage. And we’re still figuring out how to use it, creating tools, and being able to be efficient. Whereas genetic engineering, which is a similar tool, it’s just been around 40, 50, actually 60 years, I believe since 1960 is when we found it. We are much better at not only producing this kind of technology, but also at understanding what the outcome will be because we’ve done it for so long.
Dylan: You said the other hard part of what you’re doing is detecting signals. I do want to get into that, but I was hoping first we could talk a little bit. I was hoping you could help me understand more about what problem we’re solving. Why is it helpful to have plants that can communicate these stresses and how to help me put that in the context of these global problems we’re experiencing?
Shely: You started with some of the stats. It’s amazing how far we’ve gone in industrializing agriculture, yet how many inefficiencies we still have in our system. 30 to 50% of our fertilizers are estimated to be over-applied, and over 50% of pesticides are misapplied, meaning we don’t know the right timing, so we apply them. And it’s unclear there’s any positive value for farmers from those applications. And we still lose a significant amount of our crop to pathogens and pests. And that’s actually based on, the 40% is based on the UN report that came out last year, that 40% of global production is lost to pests and diseases every year, which is just remarkable. Cause it means that we could almost double the food that we make without farming one more inch of land, right? And all of this is critical because we can’t afford to be farming more land because that means that we deforested natural native land that’s already sequestering carbon to do something else. We can’t afford to reduce the oats cause we cannot afford to farm more land. And we can’t afford to just misapply pesticides and fertilizers because those create loss of biodiversity in our systems and they essentially kill the soils that the soil cannot sequester carbon. All of those reasons are basically the reasons that we’re creating in our plan. Knowledge and the right data will provide that knowledge, removing the need for the inefficiencies because the inefficiencies come because we don’t know. Farming is really, really hard. It’s not like running a manufacturing facility that you designed. It is like running nature with weather conditions and climate events and unknowns and you don’t really have the ability. And we do this at a scale of thousands of acres, if not tens of thousands of acres. The only way to do this today is to essentially over apply and manage risk in advance. And in the future with our system, you won’t have to do that because the plants will tell you exactly what they need. So we can really nurture and protect at the individual plant level, help farmers actually maximize their yields, catch the, you can think about as catching the pandemic on patient zero and then preventing the additional losses. But then also never use any unnecessary chemistry, just use the minimum amount that you need to get the work done and then no more than that. So it gives you all of those, the solutions to the system that exists today, still large scale, but planting and harvesting on the individual plant level.
Dylan: You’re providing, at some level, it’s about better data to then make decisions about applying fertilizer and pesticides. What is their data source today? How are they making decisions? How are farmers making decisions today about what to use and when?
Shely: Let’s start at the VC perspective. About 10, 15 years ago, there was the boom of the precision ad. So the solution was imagery. Where you’re just going to take pictures of the field and we’re going to figure out what’s going on. I can still tell you up to today, the best imagery that we use is NDVI, which is an index of plant growth or biomass. So you look at essentially changes in the plant’s green level. And the reality is that the problem with that information is the late stage and it’s general stress. With those, basically it’s telling you, you have a fever. You don’t want to find out you have a fever. You want to find out when you got sick and you can do something about it. Now, what farmers actually do is they listen to their crop advisors. And a lot of them, if you think about fungicide specifically, fungicide is used as an insurance policy. Farmers that know their area will spray at a specific time in the season, same every year based on the crop advisor’s recommendation, because they believe that given the certain conditions that are looking like they’re going to happen that year, there’s going to be value to them. That is how we farm. That is the data that we have today. And it’s a lot of guessing given that we lack so much of the critical information to make a knowledgeable decision.
Dylan: Got it. So you might be applying the wrong thing in response to stress. You might be applying it when it’s not needed. Is it also about, I have so many thousands of acres of a crop and I have a stress over here, but I’m applying it to the whole crop.
Shely: Yes, if you find something, if you walk through the fields and you see lesions on the leaf, you’re going to apply everywhere because that fungus is everywhere. You really don’t want to catch it at that point. That’s the other problem. You caught it too late. But yeah, I’ll give you an example. One of our farmers told me this and I love this example. He went out to his field. He saw that the crop was looking stressed. And then he saw a lot of weeds in the area. So he assumed the weeds were the problem. He applied some pesticides. The weeds were gone. And he has no idea what actually caused the stress. We treat the things we see. I like to joke that we’re in no warfare with weeds because they’re visible, so we can see them and we want to get rid of them. But the reality is that maybe the crop was deficient in nitrogen. Maybe there was a fungus in the place. Maybe it was just not that day because we didn’t get enough sunlight. I don’t know, right? But that’s how it works. And what you can see is the weeds, but everything else is still there and we have no idea.
We’re creating knowledge in our plants, and the right data will provide that knowledge, which removes the inefficiencies.
— Shely Aronov
Dylan: Yeah, it’s a much better diagnostic tool. Understand what to use and to catch it sooner. Very cool. So then what are you selling? Are you selling something directly to farmers? Are you selling the seeds then?
Shely: Our long-term plan is not to sell anything directly to farmers, but actually just to approve partners, and we’re building an ecosystem of partners in order to deliver the best solution for farmers. So this is really one of our understandings about Ag. Farmers don’t want another tool, and they don’t want another supplier necessarily. What they want is a better solution for them. So we decided, barely on that, the best way to get this to farmers is to embed it in the seeds they already buy. So when they buy the seeds, they just get this additional technology to then work with partners to be able to collect the satellite imagery and provide that to the farmers at the same cost of the seed. And then also to work with John Deere, who’s announced as a partner, in order to collect the data from their equipment and then be able to act on it. Because the equipment is already in the field, the equipment is already working its way with the products, so it makes the most sense for them to collect that information and then provide that to the farmers. We will, however, sell direct seeds in the early stages, but really long-term, we don’t want to be a seed company. We definitely are not going to be in the tractor or satellite business. We are creating the signals and we’re analyzing the signals.
Dylan: So you would sell your technology to a seed company who had sold seeds to the farmers. And are the farmers selecting seeds with interplant technology and paying a premium for that? How does that work?
Shely: That’s how it works.
Dylan: That’s how it works. I guess is there just a direct financial motivation for them to do it? There’s a payoff for those farmers who are selecting the more expensive seeds because they know they’ll end up using fewer resources, they’ll have higher yields.
Shely: Yeah, there’s always the benefits of, okay, you can reduce your stresses, your fungicide applications, you can reduce your losses to fungus pressure, you can increase yields by just relieving the fungal pressure, not even just, not what you lost in crop, but just enhancing the ability of the crop to be more productive. all these benefits. I actually think that eventually farmers will want this technology because once they farm in the light, they won’t want to go back to farming in the dark. Again, it’s all about risk management. And I’m a true believer that at the end of the day, if you provide something to farmers that will significantly reduce the risk of farming and enable them to scale, then that will be something that is a very sticky technology. And all the other benefits will not just come. Everything that I can tell you today is probably only 25% of what farmers will actually do because there’s so many things that they could do that I can’t even think about. And that’s the beauty is we want to enable them to do endless experiments and just get continuously better in their job.
Dylan: I feel like there’s this risk that we’re running out of land and enough land and resources to produce food for our growing population on the planet. And you’re talking about how this could potentially allow us to grow almost twice the amount of the food with the same resources. Is that a real risk? Is that something that we’re heading towards? And that’s one of the reasons we need to be really focused on these kinds of solutions.
Shely: That’s an interesting question. I don’t actually know if we’re running out of land. I can tell you the following things though. A lot of scientists believe that at the rate that we’re farming, industrial farming at this scale, we will not have arable land or like healthy soil that we can farm in 30 years. So I do think that if humanity has proven one way to go extinct, well, is by burning your soils. This has driven the extinction of so many other people in the past. So we definitely want to avoid that. And I would say the other thing, deforestation is a problem, right? There’s really no incentive for us to deforest land. It’s not just the biodiversity and the ecological natural systems that we want to maintain. And I know people are working on a lot of solutions for this, but the reality is that the only way that we know to eliminate carbon from the atmosphere is trees and plants. And we don’t want to take away systems that exist to do that and then try to find creative ways to fix that problem elsewhere. And this is where science meets the other part of it, which is great. Science will take us so far, but we also have to make sure that we can maintain a lot of the just natural systems that we have.
Dylan: I want to make sure I understand the whole solution here. So it may be an easy way to do this or a good way to do this would be to walk through the process. So let’s say a farmer is using Inner Plant Seeds provided by one of these Seed Companies on their land and something starts fluorescing. That signal is picked up by a satellite. It’s transmitted down. You know, what happens from there?
Shely: I thought we were going to go through the whole cycle. Okay, let’s go through the first product concept as an example, because our first product concept is a soybean that signals fungal pressure. So in the near future, farmers will be planting seeds that can communicate fungal pressure. Once the season starts and the plants are growing, we’ll start collecting sunlight imagery at a one acre pixel. Once we have that data, we’ll be able to identify the one acre resolution and infected acre in the field. So we can then tell farmers, this acre has an infestation of fungus. And then that information is also shared with the different farmer management systems. There’s quite a lot of them, but all of those can then benefit from that data. It will be shared with the crop advisor that can select the best product for that region and the best timing to get out to the field. And then also with the John Deere equipment. Now the John Deere equipment that’s called See & Spray has the capabilities to then scan every individual plant and be able to detect that signal in the field. So the equipment will go out, equip with the right product to the location that we’ve identified from space, and then scan with cameras and sensors, individual plants, verify the disease and only apply on the infected areas. And that’s how we close the loop. Actually, there’s one more point. The signals go away when the problem is resolved. So a week later, we can collect imagery from satellites and validate that the action actually produced a result, which is eliminating the disease. And nitrogen is different, but also cool.
Dylan: I see. So that would be a different product. You would have soy plants that are detecting nitrogen deficiency is a different product from soy plants that are detecting fungal deficiency.
Shely: So over time, our plan is to continue to layer them. So the next product concept is actually going to be soybeans that can detect or signal fungal pressure and insect pressure at the same time. They are going to layer the nitrogen and so on. And corn, so corn nitrogen is a really big problem. That’s where most of the nitrogen is used. A very interesting fact that gives me a ton of hope is that the emissions that happen in agriculture, the end to our emissions, which are a huge part of the agricultural problem, don’t come from the use or the production of the nitrogen. They come from the runoff. So those happen because we apply nitrogen on plants that don’t have a deficiency. They’re not able to absorb more. And then that runs off into water sources and creates the end to pollution, which is 300 times more warming than CO2. And the whole reason I’m telling you the story is that we don’t have to eliminate nitrogen completely. We just have to eliminate the over application of nitrogen. And that’s very much aligned with what farmers want to do, because they don’t want to eliminate nitrogen and then suffer yields. But they definitely don’t apply nitrogen on plants that can’t use them because there’s no economical value. We can all win together, which is, as you can see, a very big part of the InnerPlant man’s rise. How do we make something that’s good for the world and good for the customers for various different reasons?
Dylan: I didn’t realize a lot of the emissions are due to just too much nitrogen being applied, not being able to use it in the plants, and then it’s runoff.
Shely: Exactly, that’s the most of them. There’s some from the production, there’s some still do, but most of it comes from the runoff.
We will, however, sell direct seeds in the early stages, but long-term, we don’t want to be a seed company. We definitely are not going to be in the tractor or satellite business. We are in the business of creating, analyzing, and monetizing the signals.
– Shely Aronov
Dylan: And then of that whole process we just talked about, it sounds like there’s some amount of interpreting what the satellite is seeing and turning that into insights and stuff. Is that part something that InnerPlant is working on? Will somebody else do that?
Shely: We are always going to be the ones interpreting, is there a signal? Then we want to work with the partners because again, this is what they do really well and we are not. We’re never going to be as great at recommending what is the right product. For that, what you want to have is a really deep understanding of that area, a really deep understanding of the product portfolio that’s available, the different active ingredients, how they work and so on. So the best thing is really to feed this data into the different platforms and then enable them to be able to recommend better solutions.
Dylan: So yeah, you go up to, we’ve identified there’s a fungal issue in this acre or this grouping of acres or something based on the satellite data.
Shely: I think I probably haven’t highlighted this enough. So one of the big exciting parts of our technology is we identify within 48 hours, where it takes at least a week and it could be several weeks before you can see any visible symptoms. But for us, being very early is key, right? So 48 hours post-infection.
Dylan: That’s the time it takes for the plant to be able to signal, I’ve got this infection issue.
Shely: Yeah, it’s interesting that it starts within six hours usually, but it takes about another 18 hours for it to produce enough protein so we can see it from space. So if you happen to be in the field, you can see it probably with the equipment, but that’s also tough if you’re going to be in the field, it’s very random. But from satellites, 48 hours.
Dylan: Yeah, that’s pretty cool that you can see this from space.
Shely: That’s really the big innovation.
Dylan: So that gets into the hardware. So what’s happening on the satellite?
Shely: This is the other side of the equation. We are looking at fluorescence outside. And the key part of it was, okay, how do we detect fluorescence outside without an additional light source? Because if you need a light source, you’re not going to be able to do it from satellites. So usually with fluorescence, you excite it with a light source, and then you will pick up the signal for a camera or a spectrometer. And you can do that. And we certainly use that for, let’s say, an in-field piece of equipment. So from a tractor or from a drone, you can do that. But once you’re in space, that’s not going to work. What we found is a way to do this by using the sun as the only excitation light. This is really cool stuff. So this came from a domain called Ecophysiology. And it actually is another mature science. Just like I said, biosensors have been around for a long time. So has been the detection of fluorescence outside. But that fluorescence is a natural fluorescence that the plant has. It’s called chlorophyll fluorescence. It comes in the infrared, but it is there. And we, the scientists, have developed a way to see this. And we’ve been, as humans, collecting that data for at least a dozen years from satellites. What we did is we then implemented a very similar mechanism. We hired someone who was one of the innovators in that space who works with us, Ari Kornfeld. And then we utilized the same equipment, similar algorithms, but we adapted them to look at our fluorescence. And the big difference is just, it shows up in the visible spectra. The cool thing about this, is the equipment itself, if there’s filters, we use a hyperspectral comma or a spectrometer because we need higher resolution. But I would say the more exciting part is the mechanism of how you do this retrieval. And I can walk you through how that works. So basically, the reason that we can see this outside using just sunlight is because sunlight is distinct and looks very different spectrally than your signal of fluorescence. All of the sunlight, or if you capture, if you look upwards and you capture all of the sun’s spectra, you will have this very distinct spectral resolution because those gaps are constant. So what we do is we compare the downwelling light, all of the sunlight, to the upwelling light in one measurement, the light that’s coming back from the sun. All of the light that’s coming back from the plant is either going to be originating in the sun and will have the very unique spectral signature or the new light that we created, which is snow. And because they’re so different, we can then do something called N-member analysis, a regression model to separate the two and ask the question, is there a new light in here? And that’s what’s enabling us to see this very tiny signal. We only do about 2 to 3 percent of upwelling light.
Dylan: I see. So you’re essentially subtracting the light from the sun, from the light that’s coming off the plant. And if there’s anything left over, that means that plant is generating a new signal.
Shely: And we purposely look in areas where you have deep gaps, because where you have deep signature, because we know that that’s an area that’s easier to subtract.
Dylan: That’s where you’ll have a higher signal to noise ratio.
Shely: Exactly. So we look at things like red and green and different areas.
Dylan: Okay, and it is, you said it is in the visible spectrum?
Shely: Yeah, but you can’t see it because it’s just a tiny new light, so you can’t see it.
Dylan: I’m just thinking about some of the practical challenges of seeing something from space, like from these satellites is there, what about cloud cover and sort of stuff like that? Is that an issue?
Shely: I do think that clouds are challenging, there’s no doubt. And I’m sure there’s going to be some limiting factor on getting images every day, but you don’t need an image every day. You need several images a week. The other thing is that the lights are just getting better and constant lights are amazing. When we started InnerPlant in 2018, there was no Hyperspectral in Space. There were probably two by NOAA, NASA. Now there’s like 30 different entities. There’s probably two dozen startups and there’s so many government agencies with Hyperspectral capabilities and you have SAR and you have just instruments constantly getting better. I feel pretty optimistic that in five years, even the cloud issue will be resolved. They’re just constantly getting better and it’s happening really fast.
Dylan: So Satellogic is launching Satellites that have cameras specifically designed or selected or spec’d to see interplanet signals. Yes. And you have enough coverage with these Satellites that, I mean, at least the end goal is to have frequent enough imaging of these fields. Because the whole point is to be able to capture the data as early as possible. So what, daily images of the fields or something like that?
Shely: Daily images, yeah. We want to build a constellation. But we’re starting with a proof of concept, which is October. That one satellite can take an image every five days. Realistically, once you get to, let’s say, 10 satellites, you can do daily images. They’re getting pretty cheap to make and launch. It’s amazing. Everything you’ve opened in that Space.
Dylan: How much are we talking?
Shely: I don’t know if I’m allowed to say it’s not my data. Well, I mean, I do think this is known at this point that it costs about a few thousand dollars per kilogram to launch. And the new cubes that are not just that logic planet, all these companies are launching are in the hundred kilogram scale, right? So it would have been like in the past, tens of millions of dollars probably. And now we’re talking about whatever, $200,000, $300,000, it’s crazy.
Dylan: That’s awesome. And then at the ground level, so once you know it’s in this area, you’re talking about applying specifically to the plants that need them, to the individual plants. And that’s happening with hardware that’s mounted to these tractors.
Shely: Yeah, so See & Spray is already in existence. See and Spray is technology that several companies develop at Deere LED, I would say, and others like a handful of other companies. And what it means, it’s a sprayer that’s equipped with cameras and sensors that is taking pictures of the field and then doing image processing to figure out what’s a weed and what’s a plant. And then the idea is to identify the weeds and only spray the weeds with herbicide. And they’re showing that you can reduce 60, 70% of herbicide use by doing that. Those same cameras and the same sensors are going to be adapted to be able to see our signals. And it takes a small amount of adaptation, and then they can go through the field and do something similar. Obviously, we’re not looking to compare and look at weeds. We’re trying to make the signal visible and then make the invisible visible. So once you can see the signal, you know, there’s whatever, green, and it’s not green. It’s invisible and it’s a new light, but it will be in what would be referred as the green spectral, that is fungal pressure. If it’s red, it’s insect pressure and so on.
Dylan: Okay, so those systems are actually doing on board image processing.
Shely: Yes.
Dylan: As they’re driving through the fields.
Shely: Yeah, it’s pretty insane.
Dylan: And then directed spray for, yeah.
Shely: Have you seen one of those? You program, you should.
Dylan: I did look up some pictures. It’s pretty cool if you just love it.
Shely: It’s pretty cool. I just saw one a few weeks ago and I was like, since saying.
Dylan: Yeah, no, it looks really cool.
Shely: Cause we’re going like 15 miles an hour and spraying this like tiny spots, workable.
Dylan: Yeah, it’s super cool. And will there be new InnerPlant sensors on those, cameras on those, or you’re saying you can use the existing imaging hardware and it’s more about how to process the images?
Shely: We believe it’s going to be the existing. I can’t really share and it’s too early to say, but it doesn’t require too much.
Dylan: We’ve touched on this a little bit, but what kind of scale of impact, this is obviously a global challenge, making agriculture more efficient. What scale of impact do you hope InnerPlant will have in the future? And what does the company need to look like to get there?
Shely: Yeah, I want to see this trade done. 200 million acres of soybeans, 150 million acres of corn. Those are the acres of these crops. Just to give you the idea, I want to see this in every field. And then the idea is to really get to as close to patient zero every time, eliminate any need for unnecessary chemistry, use as many biological and more natural products as possible, and timing is critical for those products to be better. Give back all the yield possible for farmers. There’s so many benefits. And my mind is like, how do we scale Ag. or how do we continue to farm at this super large industrial scale as if we were farming our backyard? And we’re going to be there. We’re so close. But this is all critical because people don’t want to farm. Farming is really, really hard. So this idea that we’re just going to reduce yields and have more people in the field, it doesn’t work that way. We want like we’re now at 1.9 percent of the global population that’s farming. We need to find technologies that can help us work better, but also work with nature. And all of this is like broadcasting chemistry that needs to go away. That was the best we had for a while, but we can do so much better. We need to eliminate that. And then all of the other effects will also follow. So resistant pests, resistant chemistry that doesn’t work anymore. The degradation of soil, all that can go away if we just work in a more holistic system.
Dylan: Interesting. Yeah. So that gets into some of the benefits I hadn’t thought about. But so I think what you said there is by timing these things better. we can actually use more natural chemicals.
Shely: Yeah, when I was making the hummus, this is a big learning from that area. I was making an all natural product and very quickly I learned what that means is that you can use some preservatives, but they’re bad, like they just don’t work that well. Whereas if you want to use non natural preservatives, they’re amazing and they would last forever. But if you want natural, you’re basically using very weak products. It’s the same, right? If you want something natural versus a synthetic chemical, you’re going to get something that’s more, that’s weaker and that is more variable. And the key to that is better detection, better diagnostics and finding diseases at very early stages so that you can still get the right results with a non synthetic. You have to bring it that way. So yeah, it opens up a lot of opportunities for things, products that are more environmentally friendly to then become either more effective, more predictable or more cost effective and really all of those at the same time if we want massive adoption.
Dylan: The other thing you said, I hadn’t thought about this before, but the percentage of the global population farming today is 1.9%. What did that used to be? I’m guessing it’s reduced a lot.
Shely: Yeah, so we just cut those numbers. In 1900, 41%.
Dylan: Wow.
Shely: Yeah. And wait, think about it. In 1900, we fed 2 billion people. We now feed 8 billion people. So one thing I take a lot of, I can’t really deal with solutions that reduce yield. And that’s my big problem with organic solutions. It’s great, but it will actually reduce the access of food to a lot of people. And the problem is that those people are not the people that consume that food. So we have to find solutions that are good and are going to work for the entire population of the globe and not create any kind of social harm. And that’s where technology is necessary and increasing yields is always necessary. But we can’t continue to increase yields at the cost of the soils because then we’re essentially taken away from the future generations. This is like a very complicated thing. circle, circle, circles, but hey, we have to do everything. We call it the evergreen revolution. We need a way to feed enough people today, but not at the expense of feeding people in the future.
Dylan: Yeah, it’s a tough problem. What do you think are going to be the biggest hurdles or challenges in achieving that scale you have your site set on?
The signals go away when the problem is resolved. A week later, we can collect satellite imagery and validate that the action produced the result, eliminating the disease.
— Shely Aronov
Shely: I think the biggest hurdle is just the complexity of our solution and the moving parts. All of these things that we talked about, and there’s many more, there’s regulatory pathways if actually, by the way, I don’t know that people know this, but with genetically engineered foods, there’s actually so much regulatory approval because they want to make sure everything’s safe.
Dylan: I was wondering about that.
Shely: Yeah, and we’re very much on board. I love the fact that it’s so heavily regulated, but you have to walk through those. You have to collect the data. You have to do all the research, right? There’s many, many hurdles, and all of those are really about getting the pieces together. But the beauty is that we’re at the tipping point, at least for us as a company, where we have the technology working. We have some of the key partners on board. We’re about to announce some big new partners as well. We have a Big Seed Company. We have the Deere’s Partnership. We have the Satellite Companies. And it’s getting to the point where it’s just happening, right? And then once it happens, because it’s better, it’s inevitable. And I think by the time it gets to the farmers, that’s the easy part, because we’ve designed this whole complicated structure so that for the farmers, it’s easy. One thing that was super important to me is that when they get it, I don’t want them to work on anything. Like no tools, no trying to figure out what to do with this, no changes to the operation, no additional work, nothing. Like for them, it’s seamless. And it took a lot of bringing a lot of pieces together to be able to deliver something seamless at the end of the day.
Dylan: I love that. Yeah. Back to designing for your customer.
Shely: Yeah.
Dylan: We talked about it at the beginning. That’s, yeah, that’s really key. Awesome. My three final questions. How optimistic or pessimistic are you about the future of the planet and why?
Shely: Obviously I’m optimistic because it’d be hard to be a pessimistic entrepreneur. I think that technologies can solve a lot of problems, but I worry that even now that we understand that climate change is real and there’s a climate crisis, we don’t fully understand how bad it is. There’s very few people that I know that understand that we’re already too late. And the reality is that none of us living will probably experience the consequences of our actions right now. So I don’t know if we’ll wake up fast enough. And I don’t know if people want to wake up because everything else seems to always be more urgent, like AI. I don’t know. Now AI is urgent, the climate is about time. And I don’t worry that if we end up with a time crunch, maybe humans can come up with a solution that can solve the problem, but that’s brute force. So that will come at the cost of a lot of natural systems. And we’ll do it at the expense of we have to. Right now, we’re trying to do it without all of that harm because we have time. And if we don’t take it seriously enough, we’ll get to the harmful part. And also a book recommendation on that. It’s called The Wizard and the Prophet. And it really talks about some of those themes. It’s really fascinating.
Dylan: Oh, nice. I love book recommendations. Okay, Wizard and the Prophet. Thank you.
Shely: Yeah, yeah, basically it’s about the scientists versus the naturalists. Simply, none of them are right.
Dylan: Nobody’s right. That’s the one thing we’re sure of.
Shely: We’re both friends in Iran, but we’re both in Thailand. We don’t know. It’s suffering.
Dylan: Who is one other person or company doing something to impact climate change that’s inspiring you?
Shely: There’s two that come to mind immediately to me. One is Blue River, Jorge Heraud and Ben Chostner. So Blue River is the company that created Seed and Spray. It was then bought by John Deere in 2017. And we actually work really close with the Blue River team and they are remarkable. I think they’ve done more to move agriculture than any startup in Ag. has done. And we haven’t even seen the beginning of it. It’s going to be even more amazing over the next 10 years. The other one is Nick Halla, who is a friend of mine and he’s on my board. He was the first employee at Impossible Foods. I will say Impossible Foods probably did more for climate than any other startup I know. They’ve elevated the need for climate solutions and they’ve also elevated this with consumers directly, which I think was just remarkable. And even doing crazy things like commercializing a GMO that you consume and making it something that consumers are okay with. I think it’s just tremendous. He’s now, by the way, he left about a year ago and he’s now working on a foundry for climate solutions. So the more exciting news to come, but still working on essentially how do you remove as much CO2, methane and nitrous oxide from our system to get us back on track to even the minor goals.
Dylan: Yeah, cool. Check it out. Those are delicious burgers too.
Shely: It has to be good and good for the environment.
Dylan: Yeah, exactly. No, that’s a good call out. What advice do you have for someone not working in the climate today who wants to do something to help?
Shely: The same advice I have for all entrepreneurs, which is if you try, you might fail, but if you don’t try, you definitely fail. So might as well try. And the reality is that there is funding now in this market, so it’s a good time to get started and we have, I don’t think we’ve unlocked almost any of our goals yet. So we need as many people working on this as fast as possible and in every direction possible so we might actually find a solution that works and that solution is probably going to be a hundred different things at the same time. So I think it’s actually super exciting and interesting to be in the climate. But I have another piece of advice that goes back to our mantra, which is remember that even if you’re a climate company, there is no such thing as a climate consumer, right? Unless you’re selling to consumers directly. You’re working in an industry, whether it’s agriculture, construction or energy, and your customers are going to be in the industry. So remember that you’re building something that makes that customer successful and is also good for the world. You can’t create something that is going to create a burden for your customer and make them less effective because then you’re not going to get adoption and without adoption you’re not going to get impact. So again, try to understand that we all want to do good for the world to find solutions that are good for your customers and for the world. It’s just good for the world. It’s probably not going to happen.
Dylan: I love that. And it seems like that is clear in your strategy. Yeah. That’s driving you. Awesome. Shely, that was really fun. I think what you’re doing is so cool and really impactful. So I’m really happy to have gotten to learn more about it.
Shely: Thank you.
Dylan: Hardware to Save a Planet is brought to you by Synapse. To find out more about us and how we develop hardware solutions for the world’s most ambitious companies, head to synapse.com. And then make sure to search for Hardware to Save a Planet in Apple Podcasts, Spotify, and Google Podcasts, or anywhere you like to listen. Make sure to click subscribe so you don’t miss any future episodes. On behalf of the team here at Synapse, thanks for listening.
News
Podcast
