Ep 2: “Automation in Action: Harnessing Automation for a Sustainable Food Future”
How do we feed a growing world population in a world where natural resources – especially water – are under strain? How do we build a more sustainable future?
In this episode, we speak to Lisa Prassack, Agri-Food Innovation Expert and Data Strategy Consultant for Prassack Advisors, and Mike Meadows, Regional Farm Manager of Stemilt Agriculture Services.
We hear how automation can enable farmers to produce more food while using less water, fuel, and fertilizer, and help them understand what is going on with their crops in real time, so they can make decisions based on rich data.
Key takeaways from this episode:
- The unique challenges farmers like Mike face, and how automation technology can help to build stronger, more sustainable businesses.
- How automation technology for agriculture has evolved over the past 70 years.
- How farmers can use automation technology to improve work/life balance for their staff.
- Why a “boots on the ground” philosophy is essential for technologists building agricultural automation systems.
To learn more about THL’s cross-sector strategy to uncover opportunities in emerging technologies, visit THL.com/automation.
Mike Meadows [00:00:03] The biggest challenge on these large firms, like I’ve been fortunate to be a part of over the last few years is there’s hundreds of valves with many people involved, whether there’s just always things that play into how you grow a crop. Back in the day, we went out with a show or whatever we dug around and that’s basically how we manage irrigation. Then you fast forward to a thousand acres and you can’t do it. It’s too many acres to do it that way. And so over the course of the last 17 years, I’ve tried several different technologies that were going to be the answer.
Jim Carlisle [00:00:38] That’s Mike Meadows, regional farm manager of STEMILT AG SERVICES, a family owned farming operation based in Washington State. I’m Jim Carlisle and this is Automation in Action. I’m here with my colleague Sarika Ramakrishnan, who had the opportunity to sit down with Mike to learn about the challenges that growers face and how automation technology is helping them produce more food while using fewer resources. Sarika You also spoke with Lisa Prassack, agrifood innovation expert and data strategy consultant for Prassack Advisors. What stood out to you most from those conversations?
Sarika Ramakrishnan [00:01:13] Agriculture accounts directly and indirectly for a significant portion of the U.S. economy, and we don’t always think about that when we pick up a bag of apples at the supermarket. It was really illuminating to hear from both Lisa and Mike about the material advances automation is making for those who actually have boots on the ground. And the problem they’re solving is critical. How do we feed the world’s growing population but also address the reality of constrained natural resources, especially water, food security and sustainability? Does it need to be an either or we can have both? Technology is the key.
Jim Carlisle [00:01:53] You know, I actually do think about automation when I’m picking up apples at the supermarket. Anyway, let’s hear from one of those folks. Can we start with Mike?
Sarika Ramakrishnan [00:02:02] Let’s get into it. Hi, Mike. Welcome to Automation in Action, and thank you for joining us today. Can you tell us a little bit about yourself and your career in agriculture?
Mike Meadows [00:02:16] When I was very young, we had a family farm and we were actually dry land farmers for about the first 15 years that I was farming. And we farmed cherries, peaches, pears and apples. And I was about 20 years old, we implemented our first irrigation and we drove a well at irrigation to our production in our farm and which doubled fruit size and production. So that was my first introduction to irrigation. 1993 when I was about 28 years old, I lost my father, got a crash course in Business 101. Basically what we had to do was try to move more to direct farming. We’d done a lot of that. We had a lot of peaches that we direct marketed, but we weren’t going to be able to survive taking everything straight to a packing house. So we started direct marketing pears, cherries, peaches, apples. We planted pumpkins. We did some vegetables, and we did really well with that for several years. I had two kids that were both very young and we were my wife and I were working some pretty crazy hours and so I had an opportunity to go to work for Bill Zirkle at Zirkle Fruit about 2005. Manage a little over a thousand acres of tree fruits. So I decided to take that option and worked for them for about 17 years, which is where I was able to start implementing automation. And then this year I had an opportunity to go to work for STEMILT AG SEVICES, another large fruit trade company, and they’re big player in the tree fruit industry, really excited to be a part of the company. They packed for apples and cherries for Costco, Whole Foods, Wal-Mart, all the big players, their family owned business, which is one of the reasons that I wanted to work for them, really excited to be a part of the company. They packed for apples and cherries for Costco, Whole Foods, Wal-Mart, all the big players, their family owned business, which is one of the reasons that I wanted to work for them.
Sarika Ramakrishnan [00:04:24] How has the technology landscape changed over the course of your career?
Mike Meadows [00:04:29] There’s been a real push inside our industry to try to automate a lot of things, some in the irrigation has been a focus for me personally over the years, trying to find a way to get that automated. The biggest challenge on these large firms, like I’ve been fortunate to be a part of over the last few years, is there’s hundreds of valves with many people involved, whether there is always things that play into how you grow a crop. Back in the day we went out of the show or whatever you dug around and that’s basically how we manage irrigation. Then you fast forward to 1000 acres and you can’t do it. It’s too many acres to do it that way over the course of the last 17 years. So I’ve tried several different technologies there that were going to be the answer. Each one got a little bit better.
Sarika Ramakrishnan [00:05:19] And what is the problem exactly that you’re trying to help solve as it relates to irrigation? How does the technology really help you?
Mike Meadows [00:05:28] What I was trying to do with it was be more tactical on when we could apply irrigation in specific time frames versus when you have people involved, you don’t necessarily want to be turning on a valve at one o’clock in the morning. It’s just that’s really hard on people. So what I wanted to build with automation was target irrigation run timers when it would work the best for the crop that we were trying to grow at the time, at that day, at that week, that month, that part of the year, that was what I was trying to basically accomplish with automation for several years. And like I mentioned before, we had companies come to us and they just could not understand the scope of our operations. These are big farms, a lot of moving pieces, and we’d start out the journey with these companies and we’d start out small to make sure they could handle those per se, and seemed like they could always do it when it was on a small scale. But then when we started to scale up and started adding hundreds of acres, it got really difficult for them to be able to manage. I’d say it’s in the last five years that technology has got a lot better, it’s more competitive and something that you can actually try to implement on a large scale.
Sarika Ramakrishnan [00:06:45] I know you’re in the middle of a cherry harvest. We talked a little bit about that before we started the interview. So maybe using cherry trees as an example, before you had access to automation technology, how did you figure out exactly how much water a tree required?
Mike Meadows [00:07:01] Well, before we had good tech, there was a lot of actually pulling soil samples and trying to train people to help you with that on these large acres. Basically, every morning we had soil probes that irrigators would pull and we’d set schedules. But when you get into heat waves and we get closer to harvest, cherries can demand a lot of water. And so kind of how we we did it for years and then as first software came along, we were able to probe schedule. We were able to do what we call reeds with a neutron probe twice a week and we would actually drop an instrument down into a tube, take readings twice a week on over 200 locations, upload that data to a computer, and that would plot everything on a graph. So you could see if the schedule that you had written was adequate. The downside to all that was you only got to look at that twice a week. So it was still pretty slow and not the most accurate. When probe schedule moved to electronic monitoring in a few years ago, we were able to get that data roughly about every 30 minutes on just the soil. What I came to realize, and now that I have probes scheduled back and I’m using that on a daily basis, you can actually tend to over water with probes scheduled if you’re not careful.
Sarika Ramakrishnan [00:08:26] And what happens if you underwater the cherry trees or overwater them, for example?
Mike Meadows [00:08:32] You know, it’s not a good situation. You can lose crop size, cherry size can be reduced or it maybe just can’t finish the cherries very well. It’d be small. If you’re way over water, you can actually turn leaves yellow on a tree and you can actually kill a cherry tree by overwatering it. They don’t like to have what we call webbed feet, which is too much water.
Sarika Ramakrishnan [00:08:53] So it sounds like with some of the new technology, you can really micro target where and when you’re irrigating. Is that right?
Mike Meadows [00:09:01] Yeah, exactly. Even with probe schedule, I would have in some blocks I might have three of those set up inside of blocks. I could monitor three different locations so that I wouldn’t overwater one or under water and another. So you could be pretty specific in areas where you needed it.
Sarika Ramakrishnan [00:09:20] I know that you work with one of our portfolio companies Phytech, they’re an Israeli technology provider, and their solution helps increase yield while also decreasing water usage while improving the quality of the harvest. Would you be able to explain a little bit more about how that works?
Mike Meadows [00:09:38] So they have the probes in the soil and then they’ll have also something on the tree a little puncher then rammer, I think they call it. And then they also like right now they’ll be starting to put dendrometers on the actual apples. And so it’ll give you a measurement every day. If you’re under water and you can actually see, let’s just say an apple right now is at 40 millimeters and tomorrow is at 40 millimeters again, we get a new reading in the third day, it drops to 39 millimeters and the fourth day and so 37 millimeters, basically you’re under watering so that it’s flowing water out of the apples. With certain varieties like Honeycrisp, maybe we don’t want it to get too large, with this new technology, you can actually water and keep it keep an apple roughly at the same growth stage for several days if you want to, so that apple won’t finish too large. Instead of just continuing to water an apple until it gets too big and then it’s not a very good apple to sell. What Phytech does is will give you a a grid, a line where you can see each day where it plots on a growth curve. And if it’s plotting on the growth curve, you want, what the irrigation cycle you’re running great or if you’re getting too large and you can take off, if you will, and try to keep it right on the line and you’ve projected for the target size apple that you want to achieve at harvest.
Sarika Ramakrishnan [00:11:03] I know that by Phytech customers use 25 percent less water on average than the standard amount recommended by scientists at the University of California, Davis with increased yields. But I wonder if there’s another benefit. From your experience, do you think these types of technologies, not just Phytech but automation technology across the board are improving your crop yields?
Mike Meadows [00:11:27] I think it does two things. So if you can measure apples size and soil moisture and tree stress, those three parameters, if you can measure all three of those, we can run less water. Based on what Phytech is telling me because Phytech two can monitor the stress on the actual tree. We can go at least one additional day or two in some cases between irrigation sets using Phytech versus from schedule because you can actually see what the stress level was on tree versus what your soil is saying that you need it to replenish, if you will.
Sarika Ramakrishnan [00:12:05] So Mike, we’ve spoken about the impact of automation technology on the conservation of resources, in this case water and the impact on both yield and quality, how else does it affect the business of farming?
Mike Meadows [00:12:20] Though, a good example of how it can save you in other areas was power. A good example is the beach property that I mentioned earlier. So if I’m running half the irrigation down here, that’s half the power requirements. So that’s a pretty big player. We have seen as as high as 30 percent. But I think 15 percent savings in power is pretty achievable with automation. On the labor side, what has been pretty awesome about it is we moved from you know on a 900 acre piece of property. We moved from eight full time irrigators to four.
Sarika Ramakrishnan [00:12:55] Can you speak a bit about how adopting this technology has improved work life balance for your team?
Mike Meadows [00:13:01] I have been fortunate over the years to have people that want to do that work. The thing about automation and the one thing I didn’t really count on was we can build schedules for these workers. One time we were eight people on the weekends and we move out of four, and unless we were cooling, we could move that to two people just to monitor [00:13:23]Friday and Saturday. [0.4s] And so what actually happened at all that was we have to give people time off. Even through pretty busy times, crews schedules where people could have a little bit more of a life outside the farm, if you will, versus have to be sun programing about making sure it opens and closes.
Jim Carlisle [00:13:45] Sarika that’s a great story. We can see technology creating first and second order impacts and solving a number of critical problems all at once.
Sarika Ramakrishnan [00:13:55] Absolutely Jim. And the need is only going to increase. In 2022 the U.S. government had to sharply reduce water allocations to farmers throughout California’s Central Valley, a strip of land totaling over 20,000 square miles and accounting for over 17 billion in agricultural output annually. The Central Valley makes up 75 percent of California’s irrigated acreage and 17 percent of all irrigated land in the United States. But after years of minimal rainfall and a shrinking snowpack that supplies water each spring to the state system of reservoirs and canals, over two dozen counties, comprising over 50 percent of the state, found themselves in a state of extreme or exceptional drought. Luckily, California experienced a lot of rainfalls this year, so conditions have temporarily improved. But most experts believe that just buys us a little extra time. We’re seeing the same issue in the Colorado River Basin, where the federal government recently had to broker a deal to ration water usage from the seven states that rely on the Colorado River primarily for commercial farming. We know that technology is one solution to the longer term problem helps conserve resources like water and fuel, increase yield and produce better quality crops, not to mention improved work life balance for farmers. We’re about to hear even more of my conversation with Lisa speaks to a range of issues. Importantly, she was able to offer some historical context. Technology may be getting better and more sophisticated, but it’s been part of the agricultural landscape for several decades. Here’s what she had to say.
Lisa Prassack [00:15:40] So spend a lot of time with ranchers in their agronomists walking fields, understanding what problems they’re trying to solve, what tools can help them with that.
Sarika Ramakrishnan [00:15:49] Automation is such a buzzword. I’m curious to understand what you mean by automation in the spaces that you’re actually working in day to day and what that actually looks like.
Lisa Prassack [00:16:00] Yes. So that’s an outstanding question because when we talk about automation in agriculture, for example, since the mid 1990s we’ve been doing auto steering, auto guidance for tractors. So we put them on what’s called an AB line so that they go up and down on a field so that they maximize planting, minimize the amount of chemicals they need to use and optimize for harvest. So we’ve been looking at automation as an aspect in agriculture for quite some time. Automation in harvesting is very straightforward in commodity crops. So how we harvest corn, soybeans, cotton, peanut, etc. we have technologies in tractors that do that for us. But when it comes to tree fruit, we have challenges with bruising and how do we get all that off? So with automation there, it’s really a combination of human and technology that helps with that performance.
Sarika Ramakrishnan [00:16:58] To your point about some of the challenges automation is helping to solve, what are some of those kind of key challenges that growers are facing, particularly from a resource and sustainability standpoint?
Lisa Prassack [00:17:12] So good question. So when it comes to managing our people labor resource, one of the opportunities that we really have is how can we make sure that they’re getting to the right place at the right time to be able to do a certain activity. So how do we provide them with their work flow and enabling them to get from one field to the next to the next and do the activities that are required? So that comes down to software. We talk about sustainability. We’re really focused on reduction of fuel use in that process. We’re also overall looking at reduction in material that we’re applying. So can we use less fertilizer or is it really the fertilizer timing? When it comes to permanent crops, in particular, up to 70 percent of the performance of a crop is highly attributed to the water timing and the amount and when that’s delivered. So the fact that we’re able now to turn on and off the pumps that deliver water is really important because a lot of times the alerts that might come from our knowledge of the plants would be at two or three in the morning. And it’s really hard to send out staff at that time to get them to turn these pumps on and off. So there’s new tools and technologies that are modeling these crops that are allowing us to understand when and how much water needs to be delivered to optimize the crop that we will harvest.
Sarika Ramakrishnan [00:18:37] What would you say today is the level of penetration or adoption of both the software technology and some of the hardware software combination technology that you were describing?
Lisa Prassack [00:18:49] One of the challenges we have with technology adoption is how unique our grower environments are. Their size, the way they go about doing their business. And so we’ve been very delayed in terms of adoption of tools and technologies. And a lot of times what happens is the technology is designed without walking the field. So we have a lot of technologies that initially come out that really don’t understand our way of working in agriculture. In permanent crop, I’m lucky if I’ve got 3 to 5 percent of my grower is actually using these tools and technologies. So they’re getting it, they’re using it, but they’re really the early adopters. We’re not even approaching the pragmatic middle at this point.
Sarika Ramakrishnan [00:19:37] It feels like agriculture and farming has evolved so much over the last, I don’t know, several centuries because food has been central to life for the beginning of time. And there’s been so much innovation, layers of innovation. And this seems like just the newest layer. And I don’t know if that’s something you could kind of help walk through and what incremental value we’re getting from the data revolution and from sensors and when this next inflection point started in the context the last hundred years or 200 years.
Lisa Prassack [00:20:13] Well, if you’re okay, I would like to start with the 1950s.
Sarika Ramakrishnan [00:20:17] Yeah, that would be great.
Lisa Prassack [00:20:19] So in the 1950s, one farmer basically said about 15 people, and that’s when we started to adopt tractors to replace animal power on farms. And our global population was around three billion at that time. In the 1960s, one farmer would feed about 26 people. And in that scenario, seeding became much more efficient. You know, we’re moving then into, you know, four plus billion people on the planet. In the 1970s, four-wheel drive happened. So in the 1970s, tractors would see them more often and one farmer is now feeding about 50 people. In the 1980s, biotech improvements. So seed design and crop protection and being able to ward off disease and improve performance again, a major yield play. We have one farmer now feeding nearly 80 people. So then if we get into the 1990s, that’s when I get into this automation discussion where we have GPS guidance for tractors and for people keeping track of the work that they’re doing. So in the 1990s, that’s when we hit one farmer feeding about 100 people. So then in the 2010s we had genetics and data improvements. So we’re mearing how we design the genetics in the seed technology with data analysis to determine populations for planting based on the soil type and the historical yields. So we’re being really smart about what we’re planting, where to maximize production. And then we have one farmer feeding over 150 people. Today we have integration happening with actionable intelligence that’s driving performance and production. We have one farmer feeding over 200 people in the market today. The challenge being is that our genetics improvements are not keeping pace with the population growth. So we need to change some ways for doing things. We also need to be more judicious on our use of water and then again fertilizer. So there has been a conversation about when I’m working with agronomists and we’re all getting together and talking about what’s happening in the market and we have somebody come in, speak, inevitably they’ll say, if in doubt, fertilize, to improve your production. That can’t be our mindset anymore. It is how much water and fertilizer do we need to use to get the quality and the product that we need in order to feed and sustain the planet.
Sarika Ramakrishnan [00:23:11] I think laying it out that way and is so powerful in terms of how we’ve evolved so much, in terms of how many people one farmer can feed. And curious to get your take on the automation technology that’s in the pipe, what we have today and if that is adopted more widespread among farmers and ranch managers, how that number of how many people each farmer could feed would evolve in the future?
Lisa Prassack [00:23:43] It’s a really good, insightful question to ask about. How many farmers do we need in order to feed people? The aspect of it is comes back to what processes and skills can we replace with automation to enable our people to provide more insight and more value to our farms in our ranches? We are really limited by our harvest skill set, so anything we can do in automation, in harvest and seeing some very unique and novel harvest technologies happening in vegetables, and then also spray applications because we have to deal with things like fungicides, we have to deal with things like insects. They’re not going away. But what’s exciting is how we do automation control on [00:24:32]mating [0.0s] disruption of those pests so they don’t really take off. So we’re actually being able to identify age of an insect. And then based on that age, we have a biologic puffer that emits a pheromone that then disrupts them from being able to mate, which reduces their impact. So we’re really looking at solutions that integrate the automation along with the decision making. What we’re using those tools for is to enable our farmers and our employees to focus on higher value activity that they can do when on the farm. So fundamentally, we just don’t have enough farming resources or farming staff to produce at the level we need to where our population is going. So we need automation to go alongside with all of our people that are willing to work in the agricultural business to help them to improve their performance and success in the field.
Sarika Ramakrishnan [00:25:39] I think what stood out to me from the conversations with Mike and Lisa was really how far we have come over the last one or two centuries in farming automation and how much opportunity still remains today. Lisa mentioned that in the 1950s, one farmer fed around 15 people and today one farmer feeds over 200 people. It’s really exciting to think about how that number trends going forward as farmers try to feed a growing population throughout the world.
Jim Carlisle [00:26:14] That makes a ton of sense. And it’s interesting when you think about automation through the centuries, right? Maybe the first form of automation was the wheel. And, you know, we think about what an agrarian society looked like maybe hundreds of years ago as we as we read our history books. Agriculture may have been one of the first major adopters of the early forms of automation technology. Interestingly, Sarika, when I think about it, they may be behind the curve on adopting some of the most modern forms of automation in particular, you know, the use of sensors and data in, in agriculture to help make decisions.
Sarika Ramakrishnan [00:26:56] Yeah, I completely agree, Jim. I think to date we’ve made a lot of progress with automated tractors, for example, that farmers are able to control remotely without humans having to physically be present on the field. And these tractors are equipped with some of the more modern technologies like vision systems and GPS. And this is enabling farmers to plant seeds and weed and also have made a lot of progress in in irrigation to control when and how much water is used, as Lisa and Mike talked a lot about. I think the harvest segment has been a bit harder to automate as machines run the risk of damaging harvested crops. But to your point, it feels like there isn’t as much penetration of these technologies in farms today as there could be. So if one farmer feeds over 200 people today, it’s exciting to think about what that number can grow into over the next few decades. There is over two million farms in the U.S. alone, and that inherently makes, you know, the go to market distribution and implementation of automation technologies significantly more complex and time consuming.
Jim Carlisle [00:28:22] As we know from some of our other investments, we’re going to automate sales and marketing as well to make that more efficient. But the value prop is definitely there. I mean, agriculture is a space where there’s a lot of uncertainty in the actual production, right? When a manufacturing environment you control temperature and climate you don’t really have to deal with. But that may also be one of the strongest reasons to adopt automation is that variability. And as the cost of adoption comes down, as we’ve seen robotics in many industries become less expensive for customer adoption. And as the data and feedback mechanisms improve, I would expect that optimization under uncertainty in a much less certain environment will be a real advantage for commercial farmers. I’d want to shift maybe to one other topic before we wrap, which is we talked about automation being controlled really still by, you know, a single farmer now producing for 200 people versus 15. But there’s there’s still a human at the middle. What are the impacts on labor? This is a really clear cut example of the societal benefits I think that we see as a result of automation, irrigation savings, conservation of resources is is really important from an environmental perspective. And broadly speaking, I think is important from a society view as well.
Sarika Ramakrishnan [00:29:45] I totally agree, Jim, and I think even for the labor involved, for farmers, given kind of the labor shortage in the industry and the demand that farmers have to cater to automation really helps improve the day to day workflow and demands on farmers and improves work life balance as Lisa and Mike had mentioned.
Jim Carlisle [00:30:15] You’re absolutely right and the environmental benefits are real. I grew up spending a lot of time in in Colorado and in the western states. Some of the stories about what’s happening with the Colorado River and some of the drought issues that have been experienced in the West. Those types of things can be solved with with automation, in part where there’s better control over water usage and needs. As we think about managing finite resources for the better