Resources

• Oishii – Vertical strawberry farming company focused on premium flavor and smart agriculture
• How Oishii Uses Bees, Robots and Solar to Sustainably Grow Strawberries Indoors
• Oishii opens solar-powered facility to grow '20x' more strawberries to meet market demand
• Tortuga AgTech – Robotics startup focused on automating harvest in fruit crops
• Oishii acquired Tortuga AgTech’s IP, assets, and engineering team

Host and Guest

Host: Mike Toffel, Professor, Harvard Business School (LinkedIn)

Guest: Eric Adamson, Robotics Executive at Oishii; Former CEO, Tortuga AgTech (LinkedIn)

Transcript

Editor's Note: The following was prepared by a machine algorithm, and may not perfectly reflect the audio file of the interview.

Mike Toffel:

Eric, thank you so much for joining us here on Climate Rising.

Eric Adamson:

Thanks for having me.

Mike Toffel:

So, we're here to talk to you as one of the HBS alumni out there in the world who's working in an area that is a business and climate change space. And so, we'll talk all about the ag sector here that you're involved with and automation. But why don't we begin just by giving us a little bit of a background as to how did you end up at a company and co-founding a company and then now working at a company that is all about automation, robotics in controlled agricultural settings. It's not something that people roll out of bed when they're 22 and think, that's what I'm going to do. So how did you get there?

Eric Adamson:

Yeah, good points. It's a very niche and interesting thing to end up in. When I graduated from undergrad, I was at Stanford in the Bay Area. I didn't really know what I wanted to do, but I knew I cared about the environmental impact that human society has and the relationship with the environment we have. And I had a general feeling that it wasn't going as well as it could go back to 2006. I stumbled into the world of energy, energy efficiency, electric power, oil and gas, mining, as a consultant coming out of undergraduate. And I focused on those industries to start.

But as many people know, there's a natural overlap between energy and food and water, the energy food, water nexus and so on. And all these things have big interfaces with the environment. So, I started being more aware of agriculture. We all eat food every day if we're lucky, but also agriculture is an environmentally intensive industry. It turns out my family has a lot of connections to Ag, as many of us do. If you go back far enough 50 % of people were working on farms, but on both of my parents' sides, my grandparents were smallholder farmers in Canada and in Australia. So, as I went to HBS, I knew I was focused on energy, but also just environmentally intensive industry. And I became more aware of agriculture and the ways I might be interested in moving my career into an agriculture-focused career.

After HBS at McKinsey, that's what I did. In the last 18 months of my career at McKinsey, we're solely focused on agriculture. I moved to the Denver office, which was new at the time, and had a fairly strong agricultural focus. Some of the biggest global partners were based there and I was able to explore a lot of the large agribusiness side of agriculture. That's where I met my co-founder Tim. Tim is very different in his career progression. He was an engineer, and got his PhD in mechanical engineering at Berkeley, but he ended up in McKinsey and really wanted to build something. His family was also an agricultural family. His wife is an agronomist, and we got to talk and realized that this was a passion area for both of us and we really wanted to jump in and get our hands dirty. And that's kind of how we got to Tortuga.

Mike Toffel:

Got it. And so where did the idea of thinking about building robotics to assess and pick strawberries, which is what I understand Tortuga did, like where did strawberries come from and where did the idea of automation?

Eric Adamson:

Yeah, great question. We both knew we wanted to apply technology and maybe just smarter, operations thinking to agriculture. It’s a well-trotted statement that agriculture is one of the least technologically adapted industries there is. So, we kind of had this idea, well, there's a lot of technologies, in particular, Tim had seen technologies at biotech startups, and he had his whole PhD on really advanced automation, and he thought there's technology we could apply to agriculture. I wanted to apply it in a way that made the farms more sustainable. He really cared about the farm's affordability and access to healthy food for people. And then we basically spent our nights and weekends for about six months doing a classic kind of McKinsey value chain study and looking at the entire industry and figuring out like where might we be excited to play.

And our thesis at the early going was there's a lot to do with genetics, but we're not geneticists and we don’t probably can raise capital to become geneticists. And, and so that's not really our area, but there's a lot of interesting things happening with genetics and with soil science. And there's a lot of things happening in retail, but we don't know anything about that. But what we happened to kind of know about was operations in intensive industries. And it turns out the entire ag value chain is really centered around the farm. And there's profit taken and all kinds of dynamics, but if the farm doesn't work, the whole system collapses, and the farm is the most environmentally intensive place too. So, if you can make changes there, we thought you could create value. We thought you could create resilient farms, more sustainable farms. And that's why we focused on the agriculture farm layer.

Then we went, okay, well where in that might technology play the best and we look for the value pools that we thought technology could apply to and automation happens to be that in fruits and vegetables, 50%, 60%, sometimes up to 75 % of the farm cost is labor, manual labor, repetitive labor, very difficult labor. And when we got to talk to these farms, not a lot of people want to do the job. There’s a perpetual labor shortage around the world. So, if you could apply automation technology to these fruits and vegetables and then we honed in on fruits and we honed in again on strawberries. You could create a really big change in the industry. You could capture value that currently was kind of going away to waste or going away to like not harvesting every crop you could or going away to poor quality on the supermarket shelf therefore less sold so, you know press on the right pressure point at the middle of the fulcrum of the whole industry and the middle of that is harvest and you can really transform the industry. And so, it took us six months to dial that in and then we talked to a bunch of farmers and dialed in even further. But that's kind of how we came to berries, high value, lot of the value captured in or kind of wrapped up in harvest, solve that one problem, unlock a lot of value, change the industry. And then we just went and started building robots and trying to get them into the field.

Mike Toffel:

Yeah, I mean, that’s the thing about labor at the farm, some might think, well, there's the planting of the seeds, there's the maintenance of the watering and so on, and then there's the harvesting and then the turning over the soils and getting ready for the next year's crops. And I think we all know that the planting part is maybe not completely automated, but it is huge scale tractors that allow one individual to do much more than people could do in the past. Right. So that's seen a huge economy of scale in a sense using farm tractors, irrigation systems, you sometimes see them as somewhat automated. And so, what's taken so long, like why only now if those technologies have been over the past decades of experiencing technological change to shift from people to machines, only recently and even now it's still just unfolding in front of our eyes on the harvesting side.

Eric Adamson:

Yeah, I think there's two important components. The first component which I'll dive into is just the agricultural market structure and the cycle. And the second component is technology. But the first component is classically agriculture is known as a long-time scale industry. And that's because it's driven by the weather and by a biological process that is repeated millions of times in one tiny plot. And so on order to innovate in agriculture, you need a lot of different cycles, and you need to see those cycles come through and the classic lean methodology is just slowed down because you have to wait for the season and the season comes and goes and then you have to wait again. And that's true all agricultural technology. So that’s classic. And even the tractor, which is a no brainer now, it was like 40 years from the first tractor used on a farm until John Deere came out with the tractor that actually became the one that everyone started buying and even that took 10 to 20 years to really start getting out into the world because it's a capital problem it's an operations problem. You can't just do that overnight. And you must control the things that are happening on the farm.

So that's one big component which applies to any Ag tech or ag problem and then the second component is the technology which is just as big. Robotics is very new. We wrote a blog post once about this - everybody thinks robots are everywhere; everyone thinks it's all been automated and there's cool videos everywhere of robots doing backflips and dancing and so on so forth But if you actually look at it, there are not a lot of companies that have scaled robotics to an economically effective industry scale.

You've got Amazon and a bunch of other warehouse robotics. They're moving stuff around That's kind of a 2d problem. But even Amazon has not really cracked scale manipulation picking things in 3d and moving products around moving boxes around. That's still a frontier area. Even more frontier is doing that in a biological system that is fragile and perishable So you have berries that are only ripe for two days and if you don't pick them they rot.

And when you do pick them, they need to get out of that field in two, three hours or they're off, you need to pick them at such a high level of effectiveness, such a low mistake level that it's really difficult to industrialize that technology. So, we were applying frontier robotics, frontier AI to this problem.

The same problem applies in all robotics. It's very difficult to go from an awesome demo or even your first trial to an operational scale unit that works and makes money and makes value for the customer. So that's what's taking so long. And it's still happening, right? We're still in the middle of it. I think you now have this big humanoid wave where everyone's saying, we're just going to create jack of all trades, amazing robots that'll do anything. And my answer is, they might be able to do it for a video.

They probably even have a video soon where a humanoid is picking a strawberry, it's going to take them a huge amount of time, effort, and capital to focus on an industrial scale effective solution to that problem. And that's what we were focused on for nine years.

Mike Toffel:

Yeah, as I think about it some more, the robotics applications that I've seen in factories and warehouses are mostly very controlled environments, very repetitive tasks. So, whether you're building an engine, for example, in an auto plant or doing assembly in auto plant, these are areas where I've seen robotics, it's all about precision doing the same thing. Or maybe you're making three different models, but then there's just three different settings. Very different from a plant saying like, where did the berry grow? And on what day should you pick it? And so, I imagine you're integrating this 3D manipulation with some camera vision to determine ripeness and when the right moment is the right moment. Of course, you must distinguish what's a berry from a leaf from a stem. And then once you find the berry, you must decide, is it ready for picking today, tomorrow, the next day, some prediction piece? And then you let it loose to figure out the optimal picking. Is that right?

Eric Adamson:

Yeah, roughly, I think it's a very deep problem, even though you can describe it simply. In reality, the nuances of how berries present and how to get to 97, 98 % accuracy is hard. One analogy I've used in the past is like we just had our first son. He might kick his first soccer ball at age two, but he's going to have to learn all over again at age seven. And he's going to have to learn all over again at age 11 and decide where and how to kick it. And you'd have to learn again in high school and again to become a professional. And the depth of that knowledge, sure, is just kicking a ball, but like there's a whole lot more that goes behind it. And the jump is exponential each time. And so, for strawberry picking, it's similar, many people or projects or companies have done in the lab picking a hundred berries in a trial. But you know we picked a hundred thousand berries a day, and we're picking millions of berries a year. How do you do that at 97 % accuracy as the light changes as the plant itself changes day to day? Pest and disease problems come through. Do you have enough information to make those intelligent decisions? What fundamentally was the thing we were pushing the most? Once we had a low-cost robotic system, we had to develop ourselves. So custom robotics hardware is very hard. And we used some of the frontier controls, software controls of how to move that robot around in a custom way. Then the frontier was the decision making, which is AI. So not only is it right, but how do I approach this to minimize damage to something else? Is there something in the environment that I didn't expect to see? What sequence should I pick these berries in to make it easiest on me? So that it's really truly a specialized AI that is doing the job; a very specific job, but a very nuanced job.

Mike Toffel:

So, what does it look like when this robot is working? For those who haven't been inside of an indoor agriculture or a vertical farm, which is where you're primarily working, indoor more than outdoor.

Eric Adamson:

Yes, we at Tortuga were focused on outdoor farms with some structure. So, you have basically a metal gutter and fake soil and the strawberry plants are planted and they grow, and they hang over the sides and they're in free space. So, you're picking berries that are hanging in space. That's more sustainable, more resilient. It's also easier for technology. So, a perfect place to start. We're now applying that same technology to a vertical farm environment which is even more controlled, so it is even more conducive to automation and structuring things in the right way.

But previously at Tortuga we had a four-wheeled, basically like small golf cart, if you will, and it had spaces in the middle to put berries, two arms that were operating independently and working in each other's space, which is quite unique in robotics, and they had cameras on the end of them and we had other cameras to navigate through the farm, which is a non-trivial problem too.

And the robot would look around for berries drive through find berries that wanted to pick check again figure out how I wanted to pick them pick them check to see whether it picked them well, put them in the right compartment depending on the size and the quality and do that over and over again and we actually had crews of 25 robots that were supervised by two to three people. Those robots were running and could run 16 to 20 hours a day and the people would make sure that there weren't problems turning it off and turning it back on again. If there were, and managed berries off the robots, they check them quickly for quality, put them into the customer's operational chain.

So, once you have this deep technological problem solved, it turns out you have an operational problem to solve because you can't just say, great, it's all robots now. The robots aren't perfect. There's some nuance in how they work. So, you need people there and then you might need a crew of people to come through afterwards and clean up what's left, clean up the berries that aren't ready for market. So, you actually have to solve like an operations problem of let's move these two crews through a rotation with 30 crews of 25 humans. And how do we make that work well operationally with the customer system? So that's where we were at the end of last summer. Then at Oishii now, we've got a rack of robot arms and they're picking 24 hours a day. So very different and even more conducive to robotics, but still kind of using the same specialized AI and similar hardware to what we were doing in the outdoor controlled farms.

Mike Toffel:

Now we've been talking about this as an operational innovation, as you mentioned earlier, to replace labor, which was difficult to get access to in some places, with capital. But of course, we're here on Climate Rising also because there's a climate aspect to this. So, I can imagine that there's an adaptation or resilience story here and a mitigation or decarbonization story here, probably some of each. So, can you talk us through the climate implications of the company?

Eric Adamson:

Yeah, absolutely. At the very high level, agriculture is one of the most, if not the most, environmentally intensive industries there are. Even just carbon alone, I think 30 % carbon emissions are related to agriculture or the supply chain. 80 % of freshwater is used for growing food. All kinds of impacts on nitrogen and other chemicals are putting into the environment to control the environment for the plants. Even a soil-based farm is a controlled environment for that crop in its own way, but that has impact on the broader ecosystem.

Our thesis was that certain farms with just enough control and structure are more resilient. So, the farms we were working with - you've got these gutters up out of the soil. You're using less water. You're using less chemicals. You usually know fungicides, which for things like fruit are very important because everything wants to eat delicious sugary strawberries before the humans can do it. Lots of spray. But if you can control the environment in the right way, you can mitigate that significantly. Our robots also had UV light on them, which is a new emerging area where we could run shifts at night with UV light shining on the plants and significantly reduce the amount of fungicide spray for things like mildew, which are the number one yield killer for strawberry farms. So that's good for the environment. That's good for the bottom line. It's also good for the operations because now you have less time waiting for that spray to clear because that spray cannot go straight to the supermarket. So, all that is really helpful for mitigating the impact of agricultural intensity and the climate shifts.

So, people have heard about the floods in California maybe, which killed a lot of yield in the last few years for strawberry. The farms we were on had bad muddy days, but they didn't lose any yield because they were up out of the soil. So, the flooding just happened on the mud below the plants. There's climate control, a little bit of temperature and humidity mitigation to allow the plants to be buffered against extreme heat or cold. And that's helpful as well for adaptation or mitigation. And then all these things apply on the micro level for efficiency. And as you said already, efficiency is usually good for sustainability. Fewer mistakes, less waste means a much more efficient food system, and better traceability for crops. If you know exactly which strawberry you picked at which time and you know how good that was, you have a photo of it or even a video of it being picked and you have that traceability all through the supply chain, you can reduce waste down the chain. We started doing a next day forecast with one of our customers last summers that helped them operationally plan for the next day, both sizing their crews correctly, but also getting their sales right. The last day adjustment of the just-in-time process, being able to say, we're not going to have, or we have too much. Can you take more? Which reduces waste.

So, all these things combine into a system that is more efficient, as you said. Also, a system that is more resilient to the changing climate and the changing geopolitics, changing labor flows, changing costs and availability of chemicals that are needed for farming. That's kind of why we wanted to do a harvest system. If you do the biggest job at the center of the farm, you can add to these other benefits that are also close to the center of the farm. If you pick one of those easier jobs and try to then expand into harvest, it's probably not going to work because harvest is the hardest job, the biggest job. So, try to do the big thing first and then add these easier things as you go. And that thesis, I think, proved out for us last summer especially.

Mike Toffel:

Got it. Well, one of the things that's interesting is you said that there's a range of controlled environments everywhere from maybe the least controlled, which is just an open farm – and then you can add tunnels or tarps and things like that to semi-control it. But it still has access to outdoor air flow, and all the way up to an indoor ag vertical farm or non-vertical farm, but a glass house that is very climate controlled, light controlled, temperature controlled, and so on.

I'm wondering, as you mentioned, that at Tortuga you started with the semi-controlled environment, right, the outdoor but somewhat protected. I would have imagined that perhaps it might have been easier to start in the fully controlled setting and then easier way out to less controlled setting. But you started in the middle. Can you say a little bit about that?

Eric Adamson:

Yeah, good question. We were balancing our desire for impact as well as the necessities of fundraising venture capital. And there are probably better crops to start with for automation than strawberries but not for raising capital. Strawberry is the Goldilocks of a big enough market that has a big enough labor problem. There are bigger markets with a smaller labor problem or smaller markets with a bigger labor problem But we were kind of Goldilocks saying this hey, there's a there's a you know, two three four billion dollars’ worth of farms that have one two three billion dollars’ worth of labor that exists today that need a solution and they're in these tunnels or they're in a glass house and you know from certain perspectives, those are also even more environmentally sustainable than say a fully controlled environment that might need to source renewable energy in order to minimize their carbon footprint. Heat, HVAC, massive challenge in a lot of systems, including a glass house, but especially in a fully controlled environment.

So, we kind of focused on like, well, obviously our tech would apply even better in a vertical farm, but in 2016 and even today, there are not many vertical farms that exist. Oishii is the world leader in strawberry vertical farming, both in scale and quality. And we're still scaling up, but you know, it wouldn't have been a smart thing for Tortuga to focus on just one or just two customers of that scale to build our technology initially. So, we raised at the end, $52 million over five or six rounds. And I think a big part of our story was it's not just one crop. We also made table grapes, for example, and that's a fully outdoor vineyard. And it's not just one type of farm. We're a flexible robot that can go indoors and outdoors, that can do multiple things, multiple jobs, multiple fruits and vegetables.

Mike Toffel:

So that's all about expanding the total addressable market.

Eric Adamson:

Yeah, yeah. And I think there's inherently a lot of painful trade-offs to address in trying to hit the right sweet spot on fundraising and the right sweet spot on what you can execute on realistically with a small team. You know, biting off one of these problems is hard enough. Saying you're going to do three or four of them at once is a lot to ask of the team. So that's one of the core challenges I believe you might have asked me about later. Maybe we get into it later, but the trade-offs on getting that right is truly, truly difficult. And it's one of the hardest parts of being a founder at any startup, I'm sure, but certainly at a climate startup, how to make those trade-offs every day and try to get 90 % of them right and hope that that's enough.

Mike Toffel:

Yeah. Well, let's talk about that now seems as good a time as any. People have lots of different stakeholders that they're trying to engage with in the early days of a startup especially. You've got your potential customers. You've got your colleagues. You've got your investors. You've got your kitchen cabinet maybe, people who you have other relationships with who are just informal advisors. And inevitably, you're facing conflicting advice, for example.

So, in this case, you're talking about the founding team, the investors pushing you in particular directions. Maybe just some reflections on navigating that process, staying true to your mission, but also not being stubborn. And of course, there's a need for capital. So, you can't ignore their advice. But maybe taking all their advice is infeasible, too.

Eric Adamson:

Yeah, absolutely. Yeah, I think we had an excellent relationship with our investors. We're very fortunate. I think some of that we deserve a little bit of credit for as well is having a really, strong mission and a strong strategic vision of what that meant. Basically, climate startups have an incredible leg up in this world, right? Because you have a purpose that is beyond simply raising capital or growing. You have a mission that matters to people. And it doesn't matter necessarily to investors, but it matters an incredible amount to your employees.

Early employees take on a huge burden and they need to believe in what you're doing, and it needs to be bigger than just their career, although you also need to have a good career path for them. The investors, if you have that clear mission and then you have a clear vision and you go find the investors that truly believe in it, they'll still push you, but we were able to really stay focused for the most part on what we wanted to do because inherent to our success there would be a more resilient food system, a more sustainable food system. It wasn't that we had to chase these things to prove it out. It's like look if this is a huge opportunity, if it works, it will be big everyone says that: the only question is not the market or go to market or customer sales or anything like that, the question is can you make it work? Can you execute? And that was very helpful for us.

Starting with a mission, having a clear vision allows you to navigate trade-offs with everybody. We did have to explain or have conversations with our employees around what the investors are asking for. But we also did the opposite. Our first board meeting with our first lead, is an excellent investor of ours. He came in and he kind of challenged the core technological thesis. He's like, you guys are building custom hardware. That's really, hard and it takes time. Why don't you just take it off the shelf robot and prove that the customers will pay for it and then move forward. And there was like a knockdown drag out fight. But it was an effective fight. You need to take the hard path and that's what makes you different.

And in doing so, we had a lot fewer questions to answer for the next 10 board meetings. It wasn't like, are you guys doing the right thing? It was like, well, how do we do it better? So yeah, I think the takeaway on all that is you've got to have a really clear vision and you have to know where it can flex, but you also have to know where it shouldn't flex in order to give yourself the ability to stay the course.

Mike Toffel:

So, you and your co-founders started Tortuga in the UK and spread to California, and were building up a team, had raised a bunch of money. And then you were acquired recently by Uishi. So, talk to us a little bit about that decision.

Eric Adamson:

Yeah, absolutely. We started partnering with Oishii in I think October of last year. And we had a strategy shift from a capital perspective to start focusing on partnerships with strategics, with folks who cared about the mission and the vision. A lot of that was born to a very difficult fundraising environment. Venture capital in general is still struggling, I think, other than applied AI in very specific ways. But agriculture had a big boom and bust from the time we founded, and we saw the highs of it in 2021. And we saw the lows of it really in the last 12 months. So, for us, we were thinking, all right, well, we want technology to succeed. We might need to turn to other sources of capital for growth or acquisition. And our conversations were with Oishii, also with a handful of other very large equipment companies in the agriculture space, who all saw the importance of what we were doing on different timeframes. I think in the end, the match with Oishii was an excellent one. We have very similar missions. So, at a high level, we want to improve the agriculture system and bring better products at a lower price point to more people. I think Oishii has done an incredible job of using what they call smart farms to create the perfect environment for the Japanese varietals that they're growing and do that year-round. And they started with $50 strawberries sold only at Michelin star restaurants and now have created this incredible journey of reduction in price and availability at lower price points. Now there's packs from $8 to $10 in your Whole Foods markets in Chicago and New York and down the Eastern Seaboard. So that was our vision too. It was like, how do we make more fruit and vegetables that are better for you and better quality, more available at a lower price point. Technology spoke for itself. We did a really quick early trial where we shipped a robot out and just put it in their farm environment and hacked their farm environment to move to the robot and then did some limited testing and it just worked really well out of the box. There were obviously problems we had to tweak and tune but it was obvious a fit technologically. It's going to create value for Oishii no matter what, whether it's a partnership or an acquisition. And then at the end of the day, it came down to a decision of, for Oishii, we really want this technology. It's going to help us build on our existing technological platform and accelerate our progress in automation to keep pushing down this price curve, literally see hopefully millions of dollars in value in the next 24 months.

 I think the stated goal is a 50 % cost reduction on labor for harvest in the next 6 to 12 months, which we're quite optimistic about. And from our point of view, it's, well, this partner feels good. We have similar mission, similar values. The technology is integrating well. The teams seem to be integrating well. Let's go for it. There was some pressure on the negative side of there's not as much capital as we need. I think probably both of us would have preferred to have 12 months and keep exploring this, but the time is right. It made sense for both of us to pull the trigger. And so, I would say to your point, for me personally, for Tim, there are bitter parts because being a founder is an amazing experience despite how hard it is, but we're quite pleased with the transaction and the integration so far and have been nothing but impressed with the Oishii team to this point. So yeah, very big, long journey, almost 10 years, but in the end, quite proud of the exit.

Mike Toffel:

Got it. So Tortuga was a technology provider. And Oishii, its scope of activities sounds much broader. They operate these large, environment controlled strawberry farms. So, they're doing production. They're also doing distribution. And are they going all the way into retail or are they retailing through other stores like Whole Foods and so on?

Eric Adamson:

Yeah, they have. So funnily enough, when Tim and I started Tortuga for the first six months and what we raised our angel capital on was a fully integrated farm. It makes perfect sense from a business school mindset. Like, well, if you have the technology, why not structure the farm around it? The practicalities of doing the operations, the technology, the sales, the distribution are incredibly difficult. And we decided early to shift only to technology. I think Oishii to their credit and probably to their consternation too, they did everything, and they kept doing everything. And they were fortunate enough and talented enough to be able to raise enough capital to keep pursuing that strategy. They raised, I think, almost $200 million to this point. So, four times more than we ever did.

I think they need that money to be able to do everything, but they do in fact produce, they operate farms, they distribute, and then they do retail. They do not have points of sale, so they sell through Whole Foods and through other high-end retailers on the East Coast. They do some of their own distribution in the New York area to where they started with the high, high-end berries that are going to restaurants and some other high-end areas. So, there is some direct, but for the most part distributed through the existing channels. Nonetheless, a huge amount to bite off and they've done an impressive job at executing on all those fronts. We’re thankfully only focused on the automation and the farm operations side of things, which I think is still a huge, huge component to being able to scale this up to around the nation.

Mike Toffel:

Let's look forward and think about what the opportunities are and maybe the challenges. Anything from public policy changes? We're seeing a change in the U.S. on climate policy, for example, and in other markets similarly. There are changes in technology. You're on the cutting edge here of AI. Consumer preferences are changing in some ways. And then, of course, at least now, there’s quite a variable tariff environment. So, tell us a little bit about what are the areas that you're focused on - what are opportunities that make you smile and maybe challenges that keep you up at night?

Eric Adamson:

Yeah, I think what interested me in the beginning of my career and still today about energy and agriculture is that they're so big and global that every producer feels all these things in their daily business. You know, I remember in COVID times, our farm customers were fighting to figure out how to get people across the border. And then they were dealing with the spikes in fertilizer costs. You know, dealing with tariff environments, all these things apply to your individual business all the time. And that's very interesting. It's also very difficult. From our point of view at Tortuga, we had tailwinds the whole time. So, we founded the company in April 2016, and Brexit election happened just two months later. That was very good for our thesis.

When Trump was elected, was very good for our thesis. The first time when Biden was elected, it was very good for our thesis. In an environment where there's less immigration and less labor availability, robots make sense. In an environment where there's labor availability, but you want to give those people labor protections, you want to provide a living wage and higher living standards, robotics can help because it can mitigate some of the cost improvement. This could also give you the ability to give those people better jobs on the farm. There's always too much work to do on the farm. So, all our farm customers were looking to keep and grow their people while using robots for the harvest. How do we grow rather than be constrained by labor growth? I think the tariff environment is quite challenging from a manufacturing point of view. And that is a huge consideration for robotics. For us right now, it's not the premier thing to be concerned about. I think the main thing in robotics from a macro point of view is still the difficulty of applying any of these robotics’ technologies all the way to the end industrial problem that is valuable for the customer and for the company. And that is the nut to crack that robotics as an industry is still struggling.

So, I think the macro challenge is less about any of the politics, which can cut you both ways, but you survive. And the macro thesis is so strong that it continues. It's more about the capital. It's a capital problem, a structural capital problem that the vallwy of death in robotics is bigger and deeper than people thought. You need more capital.

Mike Toffel:

Just unpack value of death for those who aren't familiar with the phrase.

Eric Adamson:

I mean, this is a classic concept, particularly in capital intensive industries where you can build a pilot for a small amount of money, but to get to commercial scale where it's clearly working, it takes a lot of money and it's hard to get the risk capital matched up with the opportunity. So, in venture, this often presents itself as a series B or series A wall that a lot of companies hit where they proved enough at the seed stage, but they can't prove the scale or for a deep tech company like us you know, we got to series B and like we had millions of dollars of revenue coming but it wasn't enough - people wanted to see 10 and the valley of death is, we need your capital to get to 10, but you need to see 10 to give us the capital. So that's still a problem.

To make something work economically means orders of magnitude jump in difficulty or in effectiveness of the technology. Like I said, you can make a demo that's 80 % accurate right away. You can tune it up with new technologies and transformer-based models have done this, picking strawberries at 90%. But if the customer doesn't want you on the farm unless you get to 96, that's like 20 30 50 times more effort to get from 90 to 96.

Mike Toffel:

So let me turn to advice. So, one of the reasons we launched this podcast in the first place was to give listeners an opportunity to learn more about a whole host of industries and job functions and technologies. If those who have listened to this thing, yeah, this is something that I could really see myself wanting to learn more, whether it's the ag side or the robotics side or automation or deep tech, what are some resources that you would point them toward to learn more?

Eric Adamson:

I think there is a robust world of blogging and podcasts and things that talk about these business opportunities and so on. I think the best way is just to dig in and talk to people in the industry. Every conversation I have with a customer or now that our customer internally is Oishii operations, anytime I talk to somebody in operations I learned something. So, the biggest resource you can have is just curiosity.

We learned a huge amount in the early days from our plant vendors. We bought a bunch of plants to grow in Tim’s basement, and I spent like an hour and a half on the phone just asking them about the market structure and pricing and why things are the way they are You really need to just be as curious as possible. The resource is your own curiosity and your own energy to go talk to as many people as you can. You cannot do it from desk; desk research only gets you so far.

To that end, a couple of the pieces of advice that I received early days from other founders, which I would give again, one of my friends, who's also an Ag Tech CEO, said two things. He said, one, no co-founder gets any equity until they are full-time on the endeavor. I think that is a wonderful piece of advice. And that ties to my piece of advice, which is, the idea that you have needs to feel perfect from a philosophical point of view, a business point of view, a career point of view. Because once you jump in, it's immediately going to be broken everywhere and everyone's going to be telling you how it's a terrible idea and it's never going to work. And if you don't feel that it's perfect when you start, you will not make it. And therefore, anyone who's doing it with you, and you should hopefully have a co-founder. Although that's changing these days with some AI tools, generally I think having a builder and a seller is the right thing.

Mike Toffel:

Well, Eric, it's been a real pleasure getting to know your journey and to learn more about how robotics and automation in agriculture are evolving and the climate aspects of it. So, thank you so much for spending time with us here on Climate Rising.

Eric Adamson:

Thanks for having me, Mike.