This episode of Over the Edge features an interview between Matt Trifiro and Keith Bradley, Vice President Information Technology at Nature Fresh Farms. Keith discusses how Nature Fresh Farms is leveraging Edge technology to help automate and control their greenhouse environments and yield more produce. He dives into the many metrics they monitor and the information gained during the planting and growing process.
This episode of Over the Edge features an interview between Matt Trifiro and Keith Bradley, Vice President Information Technology at Nature Fresh Farms.
Keith is leveraging Edge technology in Nature Fresh Farm’s automated greenhouses. In these highly controlled environments, Keith’s team is able to monitor plant weight, growth, irrigation, nutrient absorption, and more. This allows them to automate systems that get plants what they need, exactly when they need it, to help increase their yield. In this conversation, Keith explains how this makes them resilient to changes in weather and climate, and outlines the many metrics they monitor during the planting and growing process. Keith also provides predictions on the increased focus on technology in agriculture, taking into account climate change, as well as land, water and food scarcity.
---------
“It's the exact opposite of farming how you usually think of it, where you've always got the farmers waiting on rain and hoping for more rain, or hoping for less rain. We basically control our environment top to bottom.”
“We have to control everything, from how much nutrient the plant gets, how much light it gets, how much CO2 it gets, the biocontainment of it. We control and monitor every aspect of that plant's life.”
“That's always been my goal is to eventually get to the point where now a grower isn't proactively doing something to the greenhouse, the AI is proactively telling them, this is what you should do.”
“I find the more that we can do analyzing data at the edge, the better we get. It is when we're waiting for information to get from the edge to the core, to the cloud, and you're waiting for it to get up there to analyze it, that we missed that opportunity.”
“We don't get affected by weather change. The average weather goes up by a couple degrees, it doesn't affect us the same way as it does a field farmer. Drought doesn't affect us the same way. All those things just don't affect us. In about three or four more years, I can see it starting to get more of a highlight, and more of a ‘we need this and we need to focus on this.’”
---------
(02:30) Keith’s start in technology
(03:41) Inside Nature Fresh Farm’s greenhouses
(07:19) Technology around the plants
(08:14) Automatic functions in the greenhouses
(10:30) Data they collect and where it goes
(12:59) Amount of data they collect
(14:00) Using data to improve business
(16:25) Measuring taste factors
(18:00) Empowering the growers to use data
(20:36) Digital transformation in agriculture
(21:47) Experiments in the greenhouses
(23:11) Tracking products after they leave the farm
(26:00) The edge and product consistency
(28:41) Creating a unique taste as a competitive advantage
(30:48) IT and grower collaboration
(32:00) Keith’s approach to edge and the cloud
(37:00) Keith’s technology wishlist
(40:00) Where Keith wants to see innovation in greenhouses
(42:00) How far are we from a fully automated greenhouse?
(43:54) Emerging importance of tech in agriculture
--------
Over the Edge is brought to you by Dell Technologies to unlock the potential of your infrastructure with edge solutions. From hardware and software to data and operations, across your entire multi-cloud environment, we’re here to help you simplify your edge so you can generate more value. Learn more by visiting DellTechnologies.com/SimplifyYourEdge for more information or click on the link in the show notes.
--------
To learn more about how Nature Fresh Farm’s is leveraging Edge technology, please visit:
Nature Fresh Farms / Dell Case Study: https://www.dell.com/en-us/dt/case-studies-customer-stories/index.htm#sort=%40desfirstpublishdate%20descending&numberOfResults=15&pdf-overlay=%2F%2Fwww.delltechnologies.com%2Fasset%2Fen-us%2Fsolutions%2Fbusiness-solutions%2Fcustomer-stories-case-studies%2Fnature-fresh-farms-cultivates-the-future-of-agriculture-at-the-edge.pdf
Nature Fresh Farms / Dell Customer Stories Video: https://www.dell.com/en-us/dt/case-studies-customer-stories/index.htm#sort=%40desfirstpublishdate%20descending&numberOfResults=15&video-overlay=6311527828112
Follow Matt on Twitter: https://twitter.com/mtrifiro
Connect with Keith on LinkedIn: https://www.linkedin.com/in/keith-bradley-76105049/
Narrator: [00:00:00] Hello and welcome to Over the Edge. This episode features an interview between Mat Shapiro and Keith Bradley, vice President Information Technology at Nature Fresh Farms. Keith is leading the company in leveraging edge technology in automated greenhouses. In these highly controlled environments, Keith's team is able to monitor plant weight growth, irrigation.
Nutrient absorption and more. This allows them to automate systems that get plants what they need exactly when they need it to help increase their yield. Keith also provides predictions on the increased focus on technology and agriculture, taking into account climate change. As well as land, water and food scarcity.
But before we get into it, here's a brief word from our sponsors.
Narrator 2: Over the Edge is brought to you by Dell Technologies to unlock the potential of your infrastructure with edge solutions, from hardware and software to data and operations across your entire multi-cloud [00:01:00] environment. We're here to help you simplify your edge so that you can generate more value.
Learn more by visiting Dell technologies.com/simplify your edge for more information, or click on the link in the show notes.
Matt Trifiro: Two years ago when I started the Over the Edge podcast, it was all about edge computing. That's all anybody could talk about. But since then I've realized the edge is part of a much larger revolution.
That's why I'm pretty proud to be one of the founding leaders of a nonprofit organization called the Open Grid Alliance for oga. The OGA is all about incorporating the best of edge technologies across the entire spectrum of connectivity. From the centralized data center to the end user devices, the open grid will span the globe.
And will improve performance and economics of new services like private 5G and smart retail. If you want to be part of the open grid movement, I suggest you start@opengridalliance.org where you can download the original open grid manifesto and learn about the organization's recent projects and activities, including the launch of its first innovation zone in Las Vegas, Nevada.
Narrator: And
now [00:02:00] please enjoy this interview between Matra Firo and Keith Bradley, vice President Information Technology at Nature Fresh Farms.
Matt Trifiro: Hey Keith. How are you doing today? Not too bad. You? I'm doing great. You are a technologist. When did the bug hit? How did you get into technology?
Keith Bradley: I always had the technology bug, and I'd like to say I really got my start right in high school.
I started just playing with things and the computer person there at the time said, I don't really know how to run my network. I don't know what's going on. Do you want to start helping out? And it kind of grew from there to where I started doing everything. I started being their network person in grade 12 and OAC at the time, in grade 13 and that.
So I was the person that they would call in and say, Hey, can you fix that? Can you fix that? Can you do this? Even going right down to the root of being the at back in the day, the word perfect 5.1. You were there, right? I was the guru. How do you make it do [00:03:00] that? Yep. And it just grew from there. Went to University of Windsor.
Did computer science, did stuff like that, and from there it just kept growing out.
Matt Trifiro: Yeah. And so you work for a farm now?
Keith Bradley: Yep. We are a farm. We grow, uh, bell peppers, tomatoes, cucumbers, and starting this July, the good old strawberries.
Matt Trifiro: Okay. Now these are the farms with rows of crops that I see as I drive north of where I'm living, north of Dallas.
These are in greenhouses, correct? Yeah.
Keith Bradley: So. Think of a very, very large 32 acre building.
Matt Trifiro: Okay, wait, 30, 32 acres. An acres, like a football field. That's a big bigger, that's a big building. That's like five Costcos or something. Yep. Yep. Okay. Okay. Big building. Huge building.
Keith Bradley: Yep. It's all glass. The sides are glass.
The top is glass, and we basically turn that into a controlled environment for growing, basically whatever we want. [00:04:00] But we specialize in peppers, tomatoes, and cucumbers, and it allows us to control the lifecycle of that plant from the days when it's 80 degrees out like today to the days when it's zero degrees Fahrenheit outside.
So we can control it and we can keep that plant growing. We control everything from how much light it gets to the temperature, to the moisture, to everything. So it's the exact opposite of farming how you usually think of it, where you always got the farmers waiting on rain and hoping for more rain, or hoping for less rain.
We basically control our environment top
Matt Trifiro: to bottom. So let's go inside the buildings. We're inside this building. And what's the temperature? Typically, no matter what it is outside,
Keith Bradley: So typically you're looking, uh, 60 to 70 degrees Fahrenheit no matter what. Okay. Probably approaching 80 to 90 in the summer when we really can't keep it l lower than that because of the outside temperature, but that's where we wanna keep it, is in the seventies.
Seventies and eighties is that optimum range.
Okay. [00:05:00]
Matt Trifiro: And what does it look like inside? If I was standing inside one of these things, what would I be seeing? So,
Keith Bradley: very typical of what you're seeing in here, row in there, on a typical farm, you see rows. Now you're just gonna see rows, but they're actually hanging in the air.
They're actually being, there's no dirt.
Matt Trifiro: There's no
Keith Bradley: dirt on the ground. There's dirt on the ground, but it's covered by plastic.
Matt Trifiro: Okay, so I'm walking on plastic and then suspended from the ceiling are like hanging pots
Keith Bradley: or, yep. So it's actually like a long gutter. We actually call them gutters. It's like when they build you gutters at your house for your e troughs, they do the exact same thing.
We build one long gutter going all the way down, and on top of that is a small, probably about two inches by six inches slab we call it. And then a little smaller pot, a four by four inch plant that sits there and grows in that.
Matt Trifiro: Okay. Four by four inches. That's small. You can get a lot of those in 32 acres.
Keith Bradley: Yep. To give you an idea, so we have 200 acres in total for greenhouse right [00:06:00] now, and we're growing about 1.6 to 1.8 million plants. At any given time. And
Matt Trifiro: the plants that produce food that I imagine you sell to grocery stores and restaurants and other retail trades, did those plants in four by four containers produce.
I've grown tomatoes and cucumbers before and they give be, those plants get pretty big. But you the little root system in this four by four. Yep.
Keith Bradley: Yep. So it grows up in the little bottom slop. So thes do get bigger. Okay. But for you, like a tomato vine, how long would your tomato vines be on the farm, do you think?
Matt Trifiro: Well, okay, so I think I've probably gotten a tomato vine up to six feet if I'd stretched
Keith Bradley: it out. Yep. So ours could get to about 50,
Matt Trifiro: 50 feet. And how many tomatoes would one of those plants produce in a season?
Keith Bradley: So you're looking at one tomato plant. I'm gonna guesstimate and I'm just, I'm not a grower, so I don't know a hundred percent, but I'm gonna say this, that we grow enough to feed a small family, that's for sure.
After each plant, like for the whole [00:07:00] year,
Matt Trifiro: what's the technology that's there? Around the plant and coming to the plant. Is it tethered back? Is it wireless? What is it? So
Keith Bradley: everything's pretty well wired. We run, I'd say 80% of our iot technology or edge technology is all wired. We don't rely on wireless because of latency issues.
But basically think of this because the plant is growing inside of a building. We have to control everything from how much nutrient the plant gets, how much light it gets, how much CO2 it gets. The biocon containment of it, like we control and monitor every aspect of that plant's life. How many fruits it has on it.
Maybe it has five fruits. That's too many. We'll cut one off. So it focuses on the other side. How do you
Matt Trifiro: measure how many fruits are on
Keith Bradley: it? We actually have people that go up and down and keep track of it and we are working with AI technology at the end that will do all that counting for us. We're constantly striving to do things like that every day.
Matt Trifiro: So there are humans that [00:08:00] pick, probably they're humans that count is watering Automatic is the nutrient delivery automatic. So
Keith Bradley: all that is basically automatic, fully controlled by our growing system. It basically monitors how much nutrient goes into the plant and how much comes outta the plant. So we know how much the plant actually accepts and how much it doesn't accept and it grows and then we recycle that water right back into it.
So we'll use that same water over and over cuz our feeding system is actually a closed system.
Matt Trifiro: Okay, so lemme get this right. So is it like a drip irrigation, a little tube? Yeah. Little tube. Each plant has their own tube. Yep. And each plant gets their own dose of water and nutrient mixture. Yep. Yep. And then there's a sensor that can measure how much of that water and nutrient mixture the plant uses.
Keith Bradley: Yep. Well, and the area uses basically, maybe not the individual plant, but how much that area is using. So it'll drip out the bottom through the trough. Yeah. And then we'll measure that. So we have a, it's called spoon on the bottom. And then another, another sensor for the nutrient level. So [00:09:00] we call it spoon.
So it basically drips every milliliter Ill drip. Okay. Okay.
Matt Trifiro: Wow. And that's how we measure it. Right. And do you have one of these like NASA dashboards where you all look into like, okay, plant 675 needs a Yeah,
Keith Bradley: we don't quale that far, but we have maps, heat maps of our greenhouse, like what's growing, how they're growing, what they're doing, and stuff like that.
Even as far as biocon containment, what is Biocon containment? So a typical farmer would spray for pesticides, right? Mm-hmm. So we try to refrain from using any pesticides and instead use biologicals to kill the bugs we don't want in there. So for example, we'll release a lady bug to kill something, okay?
But we'll do the Jurassic Park thing I call it, and we'll only have males or only females of that type of bug. So they die off after a few weeks. Gotcha. And all that is controlled and monitored through sensors.
Matt Trifiro: The [00:10:00] insect population
Keith Bradley: of, yeah, we'll keep track of all that. It all goes through data and all keeps track.
And you
Matt Trifiro: said that you gotta control the co2. Do you have like tanks of CO2 that you'd open up when you need more co2? So CO2
Keith Bradley: has been provided by one or two ways. You either do liquid co2 Okay. Or we have very large natural gas generators. So what's the natural burn off? When we're heating our greenhouse is co2, so we actually funnel that CO2 into the greenhouse and give it back to the plant.
Okay. So that way we're taking advantage of every little bit of energy that we can. So
Matt Trifiro: describe to me all of the data that you're collecting and where it goes. We
Keith Bradley: collect everything from, like I said, the weight of the plant. We have sensors to record how much each plant weighs. Not every one, but like in sections.
Mm-hmm. How much light they're getting, how much nutrient they're getting, how much co2, how much nutrients they gave up, what is the outside temperature for them too? How is that affecting them? And then the number of fruits [00:11:00] on them, how many fruits are actually growing on them. And how many are getting ready to be going from a flower to a fruit.
So we keep track of all those things with everything from a human to an actual iot sensor, with most of 'em being all iot, cuz we have a lot of things going on and there wouldn't be enough people to do it
Matt Trifiro: all. Yeah. Well you mentioned 32 acres and then you mentioned. 200, yeah. Will officially be
Keith Bradley: 250 acres come this fall.
Matt Trifiro: Yeah, that's a lot of plants. A lot of four inch pots to monitor. So mostly wires. So all these sensors from these things go probably follow the same route as the drip irrigation. Go back to a what?
Keith Bradley: We have an unstructured data system that actually hoses all of our data. It's a large repository data. That either hits it or our VX rep cluster that stores all of our SQL servers.
So it's a vSphere cluster and that manages all of our data. So it all hits. One core location that is local to each greenhouse. These
Matt Trifiro: 200 acres of greenhouses, they're all
Keith Bradley: over the place, [00:12:00] spread apart. We have facilities here in Lemington, Ontario and in Delta, Ohio. We keep all that locally. We process what we can, we parcels what we don't need.
Then it gets moved to our main cluster in Canada, which then stores the data and we start doing our analytics of what's going on. Did you just
Matt Trifiro: send it over the Plano Internet?
Keith Bradley: Yep. My internet does it. We actually use an SD WAN solution to get communicate between all of our facilities and it makes it easy to traverse that data and a nice, so how
Matt Trifiro: many data points are you collecting centrally?
Every,
Keith Bradley: so data points, I tried to get how many data points? Cause I think it gets confusing. So what I did is I actually did a study to convert it and I said, so I collect about 12 mgs of data per week per plant.
Matt Trifiro: Per growing thing. Not per greenhouse.
Keith Bradley: Yeah, per plant. If I include all the stuff after it's picked, cuz we not only keep track of it at the first edge where they're growing, but we keep track of it on the packing [00:13:00] line.
So we keep track of not just the lifecycle of the plant, but the life cycle of your pepper, your cucumber, your tomato, right to the point where it leaves our facility. We want to know everything that's happened to that pepper.
Matt Trifiro: That's super interesting. I can imagine this. Have a very clean picture of this.
Yeah. Can you walk me through how it's identified a problem or helped you improve the growing? Give me some examples of how you actually use this data. To improve your business.
Keith Bradley: So think of it this way, as we can control the environment they grow in, we can find the perfect rhythm for a plant, I guess you can say.
Going back to finding that perfect rhythm for life in general, we actually look for not only how much food the plant wants, but when it wants to be fed. So that way we know this is the point in time you like to be fed at say 9:00 AM, 10:00 AM 11:00 AM this is when you wanna be fed, and we take all that data.
We know how much was picked from each row, and we can analyze that to say, oh, you know what? We're not quite getting as much yield as last year because we [00:14:00] stopped watering at 10 o'clock. We had an extra water last year, so we add an extra water back to do it. So we constantly look back at last year as to how do we improve it, what do we change?
Then a lot of times too, even our growers will compare because the weather kind of moves. Across from Ohio to Canada. We'll look at, say the weather front just came through in Ohio. This is how it affected the plant and react in Canada a little better to how it is cuz we can see that that quickly. By collecting the iot data and analyzing it, we've turned our growers from being reactive walking out in the greenhouse.
Ooh, the plants are wilting. They're not doing great to being proactive. We know tomorrow that it's going to a little chillier out, so we're gonna do this, this, and this to make sure they don't feel it. Or help them weather the storm better. Yeah,
Matt Trifiro: that's interesting. So you can control the health of the plant.
It's yield. It's interesting you said sometimes you take a fruit off cuz there's too many. Why would you do that For [00:15:00] size?
Keith Bradley: Just like when you know, you think about it, you're growing an apple tree back in the day and sometimes you take one of the flowers off because you only want it to grow two apples.
So there'd be two larger apples instead of three smaller ones.
Matt Trifiro: I see. So every, every plant has like a, a limited amount of Yep. That it can put in the Yeah. Energy, right? Yeah. That it can put into the fruit. Wow. Yeah.
Keith Bradley: Okay. Yeah, and we're constantly looking at everything like that. How do we improve it? Even to the point, like, we have artificial lighting, so we will light up.
The whole thing. Do they need more light today? Do they need less? You know, should we close the curtain so they can't see the light? Should we do a lot? We constantly look at every variable in that plant's life and how to change it and how to make it feel better.
Matt Trifiro: Do you measure things that impact taste?
Keith Bradley: Yep. We actually have a whole r and d section that, that is all their job is to look at how do we make the best pepper? We will be testing, say, 200 different varieties of a yellow pepper. And trying [00:16:00] to like, this is a different seed. This is a different company, this is a different thing. Which one has a better taste?
Which one lasts longer on the shelf and which one just gives more yield? And we are constantly testing that. Everything from our peppers to our tomatoes. We actually have tryouts to become part of the testing group because you have to have the right taste buds to do it right. And we're always striving to give the customer that.
Perfect bell pepper, that perfect tomato, and finding new varieties to challenge that.
Matt Trifiro: Yeah, that, you know, it's, it's interesting. So I, I went to cooking school at one point in my life and I learned what like real food tastes like real fresh food tastes like. Yeah. Living in California and eating that way and yeah.
In general, and I'm sure I haven't eaten one of your tomatoes in general, an outside grown tomato in the middle of August in California is like nothing I've ever been able to buy in a store. Yep. Well, I imagine that there theoretically isn't any environmental thing. That happens to them outside that you couldn't recreate?
Are you like Yeah. [00:17:00] Engineering these amazing tasting tomatoes in January?
Keith Bradley: Oh yeah. Yep. We plant year round, tomato crop is a year round crop with lights. We don't stop, so we never stop doing it. We do what's called inner planting, so as the plant ages out, it gets about a year old. We take it out and we put new plants right beside it.
And that way we're always producing tomatoes. The same with strawberries now, like we will be growing strawberries for our region in the middle of winter. We're planting our strawberries in July.
Matt Trifiro: Yeah. And I imagine in July prior to a system like yours, most of the strawberries probably came from Central America or Mexico.
Yeah, right. So you're competing with local farming by extending the year and perfecting the soil and that. Wow, that's really, really interesting. Yeah. Okay, so now all this technology, how do you visualize it? Like what tools do you use? Do you have data scientists that are doing
Keith Bradley: it? I always like to call 'em where they're grower scientists, grower scientist.
I like that phrase. I like it too because a lot of the people [00:18:00] that know the data and how to use the data are growers by trait. And so they were the people that worked on the farms that worked in here, and they understand it because there is an innate value that you just can't teach sometimes about a plant.
And so we've now more focused on making sure to train them of how to look at this data and how to analyze it more. And then we get the right team together to help them. Understand it and put it into the right functions and put it into the right sheets and ex SQL databases for us to pull it up. Have a nice visual reference of this and the future that we look at every day is eventually having an AI in there helping us grow it.
How to make it look better and not needing the grower to make the setting change, but more to predict this new. Right.
Matt Trifiro: You would think if you had, if you measured all the inputs and the outputs, you could apply machine learning to get Yeah. Teach it how to get more of the outputs that you want. Yep.
Keith Bradley: That's been always [00:19:00] my goal is to eventually get to the point where now a grower isn't proactively doing something to the greenhouse.
The AI is proactively telling 'em this is what you should do. Right. Based on you
Matt Trifiro: should increase the CO2 2% tonight
Keith Bradley: cause Yep. Because the weather's changing. Yep. And that's my goal as an IT person to get the growers to that point. We strive for it too because that way it allows us to not be in the greenhouse 24 7.
We can actually get home and actually relax cuz we know the AI can watch after it. You
Matt Trifiro: hear the phrase digital transformation, and I sometimes cringe. You know that, right? It's like one of those just like, ah, really? But in the case of farming, like it's really valid. And what's even more valid about measuring it is because you have the data you can, without a tremendous amount of incremental work, employ tools like ai.
To improve the growing, whereas if you'd never gone through the process of putting all those [00:20:00] sensors in, pulling all those wires, you'd never have that data. Most traditional farms don't even come close to that.
Keith Bradley: Well, and that's where too, like I said, we get into the proactive nature of what it is. When you go back, you know, and you have a controlled environment, without having sensors, without having automation, without having that, how do you control that large of an area?
You have to have people everywhere, flicking switches, doing all that. So when we first started, that was one of our goals of our, our owner, Pete acquiring that said, I want to automate this. I wanna make it so that one person can run a greenhouse. So on paper, inside of a computer, one person from his phone can make changes to the entire environment of our greenhouse.
It doesn't matter if he's sitting in Canada, in Ohio. Or even in Europe, he can look in and make a change to help dictate the future of that greenhouse.
Matt Trifiro: Wow. Okay. So you mentioned automation, right? Yep. [00:21:00] So you mentioned a few things that still require humans. Are you guys experimenting with robotic pickers or anything like
Keith Bradley: that?
Yep. We're experimenting with anything you can imagine. We did trials with everything from BIOS scouting to look for pests within the greenhouse, to looking at pollinating bees like we use bees to pollinate the
Matt Trifiro: flowers. Oh my gosh, I didn't even think about that. You have to bring the bees inside. Yeah.
Keith Bradley: So now automating that process to picking, to counting, to even just going down and counting the number in each row and stuff like that.
We've looked at automating every single process. The areas where we've really automated are maybe not picking in the rows and doing that work, but getting the bins that we pick in to the packing line. That's all driven by. You've seen 'em many times in shows where a little robot drives this bin around and takes it to the packing line and.
So the
Matt Trifiro: pickers have a little companion robot that they put the fruit in and then it takes it to the packing
Keith Bradley: line. Yep. It takes to the packing line, dumps it [00:22:00] out, and there you go. And that packing line will basically sort it or store it based on what variety is in there. If it doesn't have enough to make a lot of bins, it'll hold it.
If it has too many, it'll start. Okay, let's start packing this area. Yeah.
Matt Trifiro: Interesting. Do you envision a world where, You're literally tracking every fruit. I mean, a fruit has a chain of custody, right? It came from a particular plant that was grown in a particular way and got through the packing line in a particular way.
What's your thought on tracking the goods after they've left? We do have a
Keith Bradley: food traceability program, so we can actually tell you that within the box, say you have a box of peppers, we'll start with that one. You have a box of peppers. I can tell you from that box who picked it. And what row it came from and when it was picked.
Okay. At the box level? Yep. Yep. So we can actually tell you right from that box when it happened. And that was uh, part of our traceability just cuz we wanted to be able to know it. But it's funny when we started doing that and you talked about data and how it helps you [00:23:00] learn things. Once we did that, we did that more for the food safety and food compliance.
We started to be able to transfer that to the growing, cuz now we know. Well, how many kilos did that section, that meter squared grow? Because we now know where it all came from. That's when all that started to clue in, okay, well why does this meter square, the greenhouse do better than that section? Hmm.
Is it something environmental? Is it something not right? Are the feeding tubes not right? Right. And then like, and that's where we really focus on trying to improve every second. So as we look at that, well wait a minute. This little section is what's wrong? Oh, maybe the feeding tube's a little blocked.
Maybe this one didn't report right. And that's how we really, really go back. And traceability was a core function of what we like to do.
Matt Trifiro: Do you guys put those little stickers on the fruit? Yep. Hate those. Is there a secret to getting 'em
Keith Bradley: off? Not really. I haven't found it yet. I'm still struggling to find it.
So that's just what it's it. It's, it's, [00:24:00] if you ever have Kids zone, you bring 'em there. It's the funnest thing to watch and how they do it though, because the machine just kind of moves a little bit left and right and just Will it click
Matt Trifiro: these little, these sticker lines? Yeah. It just goes like straight.
Well I have to admit it, it does help with audit, with self checkout cuz you know you can scan all your vegetables. Yeah. You have to wake them up.
Keith Bradley: Yeah, it does help with that, but it's one of those things where it's like, yeah, no, I always try to find the one that's not quite stuck on all the way. Yeah, because that's easy to feel.
Well, you, you know, you when you
Matt Trifiro: think Right, that I never thought about that. I usually pick the Ripest group, but now I'm gonna pick, select the one that didn't get stuck correctly. So it, but it does seem that at some point you could modify the label in real time and give, literally give every fruit its own barcode its own way, own traceability.
Keith Bradley: It's coming because we're starting to have those sticker machines now that have a printer built into 'em. So that's starting. It's coming. It's coming out there. It's gonna be here sooner than you realize. I've seen them happening. I know they're out there. We haven't quite gotten to that style yet. We're just starting to look at how do we do that?
Because again, it's always [00:25:00] trying to figure out how does that improve us and how does that make us a better grower?
Matt Trifiro: Yeah. You know, I, I had a brewer on the show and he was explaining how much iot has changed his line of work, which is very similar, right? It's a place that doesn't normally have electronics, you're putting a lot of electronics in, and one of the things that he strove for is absolute consistency with the product, meaning he wants it so that when the customer, at the end of that manufacturing cold chain retail chain, Pops open that can, that it tastes exactly as it should taste coming off the line and all this IOT and measurement helps him do that.
So consistency's kind of interesting to me. I'm a marketer by trade and you know, you think about like, well McDonald's probably a good example. So everybody knows what a McDonald French fry tastes like. Everybody loves a McDonald french fries for the most part, and I've never seen anybody duplicate.
There are other good fries. But there's no other McDonald's fries, right? Yep. And apparently, like they have their own genetic strain of potato and they grown 'em to very strict things. So I'm [00:26:00] wondering if a potential for your business, or maybe it already exists, are manufacturers or packers or restaurants that are really looking for an absolutely consistent, you know, trying to make tomato sauce taste the same every batch.
Do you find that your product actually appeals to that? I think it really
Keith Bradley: does. I think we are able to produce a consistent plant, like you said, that consistency. It doesn't get affected. The same way that you don't have in a fields farm. You have the drought in the middle of the year, the typical summer, the spring being more moisture, the tail end being more moisture, the middle being hot, dry.
We don't experience that. We're able to maintain that consistent temperature from the day the plant starts producing to the day it, it's all done and it's time to get rid of it. So that's one thing we do get, and we do strive for that consistency every day. We want that. And one of the things that I'll admit that was the first time talking about taste is when we started doing strawberries, it was the first time that I [00:27:00] ever can only eat about five strawberries.
I. Because the taste was just so powerful, like a good, powerful, but it was one of those things and I thought, oh, is it just what we did? And I learned that that's what we're aiming for on every strawberry.
Matt Trifiro: That's pretty cool because I mentioned that, you know, from my personal experience, And it's limited.
The tomato that's grown on the vine in the backyard in Southern California growing up doesn't taste like anything you can get in the store, whether it's grown in a greenhouse or not. And I imagine like transportation and shelf life have to do with lava. But it does seem to me that the first wave of automated greenhouse growing was targeted at production.
Lowering the cost. Increasing the yield. Yield, yeah. But now you might be getting into a really interesting mention of competition, which is maybe you can figure out how to grow. A strawberry, it tastes better than any other strawberry and it becomes a branded strawberry. You know, like the sunkissed peaches and the branded stuff.
You're just like, oh, I'm gonna have the honey crisp apple. And because [00:28:00] Yeah, that's actually super cool. I, you know, I seem to remember I spent a lot of time in Northern California near the wine country, and you know, one of the things that you hear a lot. Is, well, you know, these grapes taste better because the plants were stressed.
Right? They're dry farmed, or, you know, that was a year when this or that. Do you guys purposely introduce stress into a plant to change the flavor profile? Have you done that?
Keith Bradley: No, I don't think we've ever really did that. We just concentrate on the variety. Yeah. You know, like I kinda look at a plant and I think our growers look at the same way as you want a happy life, you don't want to stress things out and, and maybe that's gonna change the pace to it.
Yeah. But I think when you stress something and you starve it, you have to then fix it. I wonder if
Matt Trifiro: anybody's tried to grow grapes in a greenhouse and make wine from them and try to build a whole end-to-end system to control that. That's. To
Keith Bradley: look into that. Yeah, that's a good one. Well, the greenhouse industry, we keep expanding.
Like I said, we're into the berries now. I know we're even doing blackberries and [00:29:00] raspberries, like all that stuff is coming into the greenhouse because we can grow year round. So, you know, it is gonna be cool to see. I, I think it's funny cuz like I said, I grew up in Lie Ontario. We've had greenhouses here since the day I was born.
Not as much as they have now, but we were only traditionally dead. Peppers and cucumbers, right? Then when lights started becoming available, they did tomatoes and now they're starting to get into strawberries, watermelon, cantaloupe, all the, all the berries. They're trying to get on every single little area that they can now, and I think that's because we've learned how to nurture that technology to to out loud them, to grow.
We've gotten better at it each day.
Matt Trifiro: You seem like a tinkerer. Yeah. Okay. Oh yeah. So how much of this was off the shelf? Hire a system integrator, you make it, and how much of this is just homegrown?
Keith Bradley: Our growing system is very common in greenhouses. The [00:30:00] homegrown stuff is how we learned that it, my team and the growers could work together to see the data.
I'd love to say it's my system. I'd love to say take credit for what we do, but. It's more the culture of nature, fresh, how everybody needs to contribute. And we always enforce that policy and we always wanted to do it. So yes, there is a lot of custom things made. There's a lot of customization, there's a lot of things done, but a lot of it came from, we need to try this or we need to try that.
And a lot of it to be truthful, came from failures. We tried to do this, it didn't work. But we learned we could do that and it does work. The core system might be outta the box and very generic for greenhouses, but how we analyze and control and collect our data is what I call is a nature fresh system.
And that's just how we do it. Like that's how Nature Fresh does it.
Matt Trifiro: And in this use case, each greenhouse is, let's call it the edge. [00:31:00] That's the place where it happens. And in contrast to, let's say, the centralized cloud, which means you need, you have VxRail cluster. Do you want to get out of the business of having to run servers in your greenhouses?
Or how do you think about the relationship with the edge and the cloud and centralized versus distributed?
Keith Bradley: I have an opinion. I, I love the cloud. I love it for disaster recovery. I love it for some features, but I have found that. Us hosting our data on-prem, we actually use very little cloud presence. Us posting the data on-prem gives us control of our data for the speed and access.
So the best example I'll use is we run ETLs on the growth data on what we're picking and stuff like that. When we were kind of using a cloud system a few years back and an older system, it would take. Six to eight hours for it to complete, to accumulate all the data. When we said we're gonna convert this, we're [00:32:00] gonna put it on-prem, it can change that ETL time from hours down to 20 to 30 minutes.
And it doesn't sound like a lot, but what it changed was now a grower can run an ETL and get an estimation down multiple times a day now. Yeah, yeah. And even for us as a company, We host our own S3 bucket on our power scale. So we don't need to go to Amazon to have that power, that S3 bucket, we got it right on-prem, we got full control of it.
It's all flashed. It's super fast. So when we talk to a vendor that says, Hey, we wanna try this robot out, we need to get an Amazon S3 bucket. Yep. Here it is done. Oh yeah. Yep. And then they see it because now you're not going. From the edge to the core to the cloud, you're more sometimes just staying right at the edge.
Right. And then we're only transferring from the edge to the core when we really need to. [00:33:00] Right. And we're trying to analyze it at that edge.
Matt Trifiro: Right. And so when you think of your system, because you mentioned on-prem, so in a sense, the servers that are running the greenhouse are on-prem, but you think of it as edge because you have your core, which is your centralized repository with your S3 bucket and all these things.
Yeah. And then you use the cloud for disaster recovery. Is that the right sequence? And so is all your data eventually makes it up to the cloud, or at least the data that isn't pruned at the
Keith Bradley: greenhouse. Yeah. A lot of the data makes it up to the cloud for our backup and disaster recovery. Maybe not a lot of the growing data that wouldn't be usable.
Mm. Right. Like if a natural disaster hit a greenhouse, it probably would never build. Yeah. Why? Why pay for that? But the analytics that we get from them that we need to get for it goes up to there To be backed up. Yeah.
Matt Trifiro: Cuz I was thinking it's not a failure of the cloud, that it can't process. It's not a fundamental reason that it can't process your, I mean, goo Google could run that workload internally in five seconds, right?
Yep. They have big enough servers and big enough, smart enough people to be able to figure out how to do that. And [00:34:00] so eventually there's gonna be services in the cloud that presumably you can't easily get. So AI may be a good one. May, you know, you may wanna use the AI at Azure or Amazon or Google and having to run that locally.
You might actually need a real data center. Have you experimented with the cloud AI services and how they may help you analyze your data
Keith Bradley: at all? Yep. We've played with 'em on and off. We've had success and failures. It never works out easy. The biggest success that we had was, was actually allowing us to more share the data between locations quicker, so, We have facilities in Mexico, we have facilities in Laredo.
It's easier for them to go up to the cloud and kind of reference it than for us to actually pull it out. Yeah. And then they kind of process that and then we pull it down and control it later on. But the cloud has given us when we need it, the agility to do things. For example, sometimes we need to spin up a couple more VMs to do something.
We need a small thing. [00:35:00] We have that connection. To start to run that, and that's where we do rely on the cloud for a few small things like that. I'd always like what Dell has told me, and I like the concept even at VMware now, that even to them, the cloud is your own cluster inside your server rack. So our VX rail cluster that we have on there is our first cloud, and that's where I really feel like that's what it is.
That's just our first cloud because really, Through the adventure of SD wan, all of our servers can talk to each other now very easily. So it's just a matter of where you wanna do it. I always have fear of the horror stories of cloud costing you way too much one month because you didn't realize what you did.
And so we really monitor what we do and that's why we, we've shied away from it, but we've always wanted more of that full control of our technology. I don't wanna kind of get stuck into the AWS or Google Cloud. This is how much resources you get. It's more shared. No, we wanna know exactly what it [00:36:00] is, and if I wanna change it, I wanna change it.
So now
Matt Trifiro: imagine your boss comes to you and says, you know what? We had such a great year last year. You could up open up the toy chest. You now have unlimited budget. What would you go out and wanna spend it on?
Keith Bradley: Oh, I already know I have, I have that wish every year. Yeah. I have the wishlist every year that's sitting there ready to go.
Right now we have very generic nodes that run, that do our workload very well. But I'd love to get into a couple AI specific nodes and to add that horsepower to 'em to be able to process even quicker what we do. And then the last thing I'd love to get into more is to create more edges. Where right now we have.
The edges that are in the greenhouse right now, we have one edge computer. I'd love to diversify that and take a 32 acre range and make it into two sixteens, and that way the analytics for that range are more local and quicker right there. Okay, so let's get the analytics for this area a lot [00:37:00] quicker for that grower that's watching that.
Then start pulling them all to the cloud or all to our internal cloud. We'll say and start looking at the bigger picture because. I find the more that we can do analyzing data at the edge, the better we get. It's when we're waiting for information to get from the edge to the core, to the cloud, and you're waiting for it to get up there to analyze it, that we missed that
Matt Trifiro: opportunity.
Why do you think that is? Is it a human psychology thing or, Nope.
Keith Bradley: It's just even a time thing. Cuz by the time. Sometimes data gets moved from the edge to the core to the cloud, and then gets analyzed. It's too late where I'd rather be analyzing things right at the start.
Matt Trifiro: Well, what's an example of something that would be too late?
Keith Bradley: So, Plants really like to be watered at very specific times. There actually is a sweet zone and it's not a same time every day.
Matt Trifiro: It's when they're thirsty and raise their hand.
Keith Bradley: Yep, yep. You gotta be right there. So if you're waiting for the computer to [00:38:00] analyze it, to go back to the cord, to go back to cloud, to get that information, then to send it back, you might miss that window.
Cause the window can only be three or four minutes. And so the more that we start to put at the edge, I think the quicker that we'll be able to take advantage of that and that'll help us do more analytics at that moment. That's pretty
Matt Trifiro: interesting. I hadn't actually heard that before. Yeah, but it makes sense.
It does make sense. It does make sense. Do you have any concerns of how much more difficult it becomes to orchestrate all the workloads when you start exponentially increasing the number of machines? Or is that not something you're worried about?
Keith Bradley: I don't worry about that as much. I know that you're making it more difficult.
But I always look at it as the more difficult it is on me, there's more automation I'm gonna go out and find to help me control that. How do we do it to make the right Kubernete cluster to
Matt Trifiro: do this? Nothing like a problem. To focus the mind.
Keith Bradley: Yeah. Like let's figure out how to do that. I would rather me saying, Hey, we're running [00:39:00] into a problem because we have too many VMs and we're analyzing too much data and we're getting an increase in yield.
I have to find a way to control it. Then be the opposite where, yeah, we don't have much going on, but we're not seeing a great increase in yield yet this year. Because every time we have an increase in yield is an increase of profits. It's more peppers and vegetables for the consumer, and that's what we need out there every day.
Matt Trifiro: Yeah. You need a visual overlay, augmented reality where it like paints the dollar sign value on every fruit as you walk through the, that's, that's a $1 tomato. Yep.
Keith Bradley: We, we try not to look at it that way, but
Matt Trifiro: it is No, I know. I know. Well, it is. It's a factory. It's a factory
Keith Bradley: that produces food. Yep. How much is this?
We have to heat the greenhouses up today. How much is that more gonna cost us? We have to use different packaging. How much is that gonna cost us and how do we get the yield increase to help compensate for
Matt Trifiro: that? Yeah. Yeah. Oh, I can totally say that. I can totally say that. All right, so let's look forward to the future a little bit.
Yeah. Given all the technology [00:40:00] that you've seen and have access to, there are probably some pretty big holes that need innovation. If you could pick, if you could set a team of researchers to, to innovate, where would you have them innovate in the sort of greenhouse world? For me,
Keith Bradley: personally, it's gonna be in the pollination of the plant.
I think everything starts from when you pollinate that plant. And right now you do that with bees? Yep, bees do all of our pollination, and again, I'm not a grower, but it's my gut of being in the industry for 10 years I've seen it, that we've learned that there's a specific time to irrigate a plant. So correlating that, is there a specific time that that flower should get pollinated and which flour of say three should be pollinated?
So instead of us coming and taking that extra flour or tomato off later on, It just doesn't get pollinated and the plant naturally drops it. I think that it would be the biggest groundbreaker and the coolest thing to start to see. Just [00:41:00] even the concept in my head of seeing the robots go down there and analyzing and puffing a little bit of pollen onto the plant.
It just makes me kind of like, oh, that'd be fun to watch and be a part of.
Matt Trifiro: Yeah. Wow. Okay. So pretty much my last question is when, when we have a fully automated greenhouse where the human is only occasionally needed because there's some exception that requires somebody with feet and arms where it's grown and packed and produced, and you just sitting in your room watching how it works to make sure nothing goes wrong.
Oh, that's a
Keith Bradley: hard one. There's a lot of facets that are very hard.
Matt Trifiro: Okay? Like what? What are some of the super hard
Keith Bradley: facets? For example, de leafing a plant. So taking a couple leaves off of the plant, we actually will get rid of some leaves. Things like that that happen. Why would you get rid of leaves
Matt Trifiro: to create more less unshadowed?
Airflow.
Keith Bradley: Airflow, airflow to the bottom, and then all that stuff, and just helping the plant focus on the fruit instead of the leaves. Okay? So it doesn't grow as much. And two, you gotta think about that 50 foot vine of tomatoes that's growing out there. [00:42:00] That 50 foot one, we have to put clips on it to have a string to hold it up.
Oh, yeah. And little functions like that are hard. I think that you'll see it, I'm gonna guess you'll see it in the 2030 range. Okay. It, it seems like a long way away. I don't think it is. It'll, it's gonna blink in a flash in eye and it's gonna be there, but I think it's gonna be a while because there's so much, and I think agriculture in general just isn't thought of as a technology field.
And I don't think that they, they're focusing on, every time you see a commercial about technology or automation, you always see a Tesla or a new plant getting this robotics to put the arms and this, that, and the other thing, or the iPhone's getting assembled by a robot. I think the agriculture's just kind of that little small enough market.
That there isn't as many people in it yet to focus on it. And I think it'll be a emerging one in the next couple years. And then we'll really start. So I'd say 2025, you'll start seeing a huge amount of people starting to go [00:43:00] into it. Cuz as food becomes more scarce.
Matt Trifiro: Yeah. Not everybody needs a Tesla, but
Keith Bradley: everybody needs food.
Yeah. And mega city start coming up and you need to produce more and more, like you said, the food closer to where you are.
Matt Trifiro: Yeah, so you'll minimize the transportation. Somebody put up a slide. I dunno if it's true or not, but they were talking about carbon footprints of agriculture and they were saying California and Florida every year.
Pretty much shit. The same amount of oranges back and forth.
Keith Bradley: Really that would be interesting to see. So we basically buy it from backboard to each other.
Matt Trifiro: Yeah. Yeah. I thought that was such an interesting thing. But yeah, I mean why ship? Especially something heavy like a watermelon or a an orange,
Keith Bradley: right?
Even semi-loads of peppers. Like why are we shipping it? Why don't we get it closer? And that's where I think as land gets scarcer, these huge fields get harder to do. And the weather changes cuz that's how they do. We don't get affected by weather change. Yeah, the average weather goes up by a couple degrees.
It doesn't affect us the [00:44:00] same way as it does a field. Farmer drought doesn't affect us the same way. All those things just don't affect us. So that's why I see in about three or four more years I can see it, uh, starting to get more of a highlight, more of a, we need this and we need to focus on this. It's changed a lot in the last few years just to focus on where your food comes from, and I think that started the process.
But now I think that once it starts getting a need for it, That's gonna really change.
Matt Trifiro: All right. Well, Keith, been an absolute pleasure having you on the show and learning all this, how the Edge computing is advancing the way we eat in agriculture. If people wanna find out more about Nature Fresh Farms, or if they wanna reach out to you electronically, what's the best way to do
Keith Bradley: that?
Um, you can always just reach out to us on our naresh.ca website and there's always info naresh.ca and come and hit us up there and we happy to, and you know, even just emailing me personally is fine too. I'm justKeith@naresh.ca. I love to talk to people. It's one of the things I do enjoy in life. I love getting a conversation going.
Anything from [00:45:00] technology to how you do this. It's something I enjoy, I have a passion for. I do a lot of engagements where I talk about what we do in depth and sometimes varies Minorly. But you know, I'm always happy to talk to anybody.
Matt Trifiro: That's great. That's great, Keith. Well, I was happy to talk to you today, so that I definitely agree.
Keith Bradley: Yep. Thank you for having me.
Narrator 2: That does it for this episode of Over the Edge. If you're enjoying the show, please leave a rating and a review and tell a friend. Over the Edge is made possible through the Jenna sponsorship of our partners at Dell Technologies. Simplify your edge so you can generate more value.
Learn more by visiting dell.com.