News

Agriculture Technology Podcast: Precision Ag in Cotton

Agriculture Technology Podcast: Precision Ag in the Cotton Industry

Posted on

In our latest episode, Jason Ward Assistant Professor at NC State University, joins to share a different take on precision ag - and more specifically, precision ag in the cotton industry.

Jason and his team at NC State discuss their recent research projects, how cotton growers are using technology in their fields today, and what they see on the horizon in the precision tech space.

Looking for more things cotton Jason points to the Cotton Cultivate site as an excellent one. (https://cottoncultivated.cottoninc.com/)

Tune in to Episode #92 here:


You can read through the entire episode’s transcript, here:

Tony: Hello everyone. This is Tony Kramer product specialist with RDO Equipment Company and you are listening to the Agriculture Technology podcast. Everyday there are phenomenal advancements being made in the field of agriculture technology. RDO Equipment Company is a leader in agriculture equipment and precision agriculture technology and is here with industry experts bringing the latest news and information from RDO and John Deere. Thanks for joining us in Agriculture Technology podcast.

Welcome back to another episode of the podcast. This is episode number 92 and today we are going to be talking about precision agriculture and also a little bit on precision-ag in the cotton industry. Before we dive into the show, please take a moment to subscribe to this podcast if you haven't already.

You can subscribe to the show on many different podcasting apps that we're streaming this to. It's on Apple's podcasting app. We've got it on Stitcher, Overcast, SoundCloud as well as many others. While you're out there, drop us a review. We'd really love to hear what you think about the show.

Lastly, make sure to follow RDO Equipment Company on Facebook, Twitter, Instagram and catch all of our latest videos on YouTube. You can also follow me on Twitter @rdotonyk. With that, let's get back to the show. I am really excited to welcome Jason Ward who is an assistant professor at North Carolina State University. Thanks for joining us on the show today Jason. To get started I'd really like to hear a little bit more about you and your background and how you got involved in what you're doing today.

Jason: Thanks so much for the opportunity to get to share a little bit of a different take on some precision ag. Actually, I grew up in Kentucky which is not a cotton producing state. It's interesting how I found myself working in precision ag in cotton production. I did my bachelors and masters there in Kentucky and I took a few years and worked for the USDA ARS in Auburn, Alabama for about four years and then found my way to Mississippi State University where I did my PhD work while I was working full time for extension there.

After that I joined AGCO for just over a year and then find myself here at NC State University leading a group I call the advance ag lab, and where we get to do some really fun things. Everything from field logistics to UAV work, including cotton precision ag.

Tony: You definitely have a very extensive background and career path when it comes to education and agriculture, specifically precision ag and what you're doing. I just want to dive into it right away. Let's just talk about precision ag to begin with, and then we'll talk a little bit about what you're doing in cotton production. Precision agriculture, what is it? What is the true definition of precision agriculture?

Jason: I think if you ask 100 practitioners, you would get 100 different answers for this question, but I really think that the way that my lab and my team defines precision ag is that it's not a technology. It's a technology enabled methodology and I think that's the key difference in how we approach it.

Precision Ag Podcast Cotton

















Traditionally, we used to think of precision ag as terminals in the tractor and the CAN bus, and the writing prescription, and doing as applied files. Don't get me wrong, those are all still very much precision agriculture, but I think that environment has grown and changed so much to say that, if that's the only precision ag that you're doing, you're actually missing out a much broader scale that's allowing you to make better decisions on finer and finer management areas so that you can really increase your input, manage your inputs. Get your output where it needs to be, and hopefully make good decisions that impact the bottom line positively and maybe can reduce some environmental impact.

Tony: That is really a unique perceptive and a different way of looking at because I too coming from the equipment industry here with RDO Equipment Company, there are a lot of people that look at precision agriculture as components, the hardware. That you're using precision agriculture if you have a John Deere display and receiver, but the way you explained it there as, it's a practice that you adopt, it really makes a guy think about it differently.

Jason: Yes, it's a way of thinking. How do I approach this particular management problem that I'm trying to go after? It's breaking down what data do I need to make this decision rather than shooting from the hip or doing it the way my daddy and his grandfather have done it. That takes quite a bit of challenge because you need to get comfortable with a completely different set of skills. We're asking that farmer or that consultant or service provider to do everything that previous generations of farmers have done in terms of being able to mechanically manage equipment, to manage people and logistics, but now we're asking you to be able to process data, and to draw actionable insights from.

It's a whole different tool set and the thing is we can't leave behind the ag. Even when we make it precision or digital ag, it still very much has to be the ag, it has to work for our production system. I think getting in that mindset is a bit of a challenge.

Tony: I would agree with that and I really do like the way that we have to shift what it is we're thinking about and how we're adopting this and utilizing precision ag from a holistic approach.

Precision ag, whether it's the hardware you're using or it's the mindset that we're in, down there at NC State, you guys are doing a lot of different research and everything. Why don't you talk to us a little bit about some of the research projects related to precision agriculture?

Jason: Sure, probably one that we're probably getting the most interest in right now is we're actually developing techniques to be able to detect physical crop damage after, let's say, a severe weather event or maybe after there's been some kind of extensive wildlife damage or even some kind of an insect issue, but really the weather piece it what drove it home for us.

Here in North Carolina, harvest season is hurricane season, so we find producers in a situation where after a year's worth of labor, diesel, inputs, seed and all of the things that went into making that crop, that a severe weather event comes along and can basically steal it out from underneath you. We really want to create tools that allow them to be able to assess the amount and severity of the damage and be able to do it quickly and we want that for producers.

We also want that for instruments agencies to be able to understand what's going on the line for them. We also want that for disaster declarations. The more accurately and the faster we can describe that crop damage, the faster additional resources may be able to come open at the state and federal level to help these communities start recovering faster.

Using in this case, UAVs to create these algorithms that can detect damage from color imagery is where we're starting, but we think we can eventually extend this out to even manned aircraft. I doubt we'll get to the point to be able to use satellite, but we really think the UAV and manned aircraft is most likely the fastest ability to detect that damage in that severe weather environment.

Tony: I like it. I think that last year speaking specifically from the region that we're in up here, the Red River Valley, Minnesota, North Dakota, we had some pretty late season severe hail damage. I think if you guys can nail down something for crop damage assessment, it would be one heck of an awesome tool to have in this industry.

Jason: Yes, we're really excited about it. I've got a postdoc who has just done some fantastic work with some deep learning using some artificial intelligence tools to see what we can extract from just simple RGB color imagery. We can go further with maybe some of these vegetative indexes or multispectral imagery, but we just thought let's get everything we can from a simple color picture first, and then figure out what our gaps are.

Tony: Beyond the crop damage, and what you guys are doing with that at NC State, are there any other research projects that you're working on today?

Jason: Yes, absolutely. So many of the crops in the South East or in Southern ag in general are traditionally a little bit underserved when it comes to technology. For example, I've done some work peanut growers. Right now, there's currently not a really good way to measure peanut's yield, so one of the most foundational precision ag tools, the yield map isn't available to those producers. I don't know that we'll ever find a perfect solution for them to map yield, but what we can do is map the machines that are pulling the combines. Maybe if there's something we can measure about the machine, or the combine settings or combine performance that can give us some insights about the yield of that crop, then maybe we can find a good proxy, just because that stream of peanuts are coming up out of the ground, they're dusty, there's rocks and sticks and everything else going up through when it's first harvested in the field so it can be a challenging environment.

We're just trying to see if there's things that we can measure about that machine that can make that process more efficient, and maybe use those field efficiencies a little bit better. Also, is there some proxy we can find that's going to equate to yield for these underserved crops?

Tony: When we think agriculture and we think precision AG, a lot of us, whether we're trying to or not, we get stuck on the typical corn and soybeans, maybe some wheat goes in there. There's a lot more to agriculture than just our typical corn and soybean crops. Even speaking here in the Red River Valley with the sugar beet crop that we utilize, we're starting to get the opportunity to run more of this precision agriculture or utilize this precision agriculture practice within these specialty crops.

To hear that NC State is working on improving the spectrum of crops that we're focused on in precision AG, it's really awesome to hear because, you know as well as I do, agriculture is so much bigger than just corn and soybeans.

Jason: Really, especially in North Carolina, we've got so much diversity in crop production. We have all the big players, got corn and wheat and soybeans, we've got cotton, we've got peanuts, and we've got a huge amount of vegetable crops, sweet potatoes. We are definitely a truly diverse state. That's one of the things that excited me about being able to work here at NC State, was trying to find the right tools and technologies that can help those producers in those specialty crops as well as the main line crops.

Tony: Now, keeping in line with those specialty crops and what you are doing, Jason, the cotton. Cotton production and as it relates to precision agriculture, let's dive into that a little bit deeper and just talk about what's out there and what you guys are working on to advance precision agriculture with cotton. The first question I got is, I'm not a cotton guy. Let's talk to a little bit about what precision AG technologies are out there today that can be utilized in cotton production.

Jason: Sure. This is a template right through the production cycle. If we look at planting, when cotton emerges, it is not a vigorous plant when it emerges from the ground. There's a lot of replant decision that has to be made. We're utilizing some tools now which necessarily haven't been used in other crops, but as far as moving some of this early season UAV imagery or other ways to detect emergence and vigor, are being applied right now to help look at that replant decision.

Right now, the standard definition is, if you see gaps greater than three feet on 50% of your field, then you need to replant. There's not a whole lot of people that are going out there and specifically measuring those gaps, they're not. There's a lot of windshield decisions being made. If you can see the gaps, if you're going at 50, versus can you see the gas if you're going to 30? There's some decisions being made in the way that's not super analytical.

We're trying to introduce some tools that can help make that more consistent and can hopefully pay some returns and preventing replants that didn't need to happen ,and making sure we find the replants that do need to happen.

Also, maybe if we look at Spring, cotton's an interesting crop, in that we actually have to apply, if needed, a plant growth regulator to it. If you look at a cotton picker, there's only a certain height of plant that can really go through that cotton picker. We need to keep it underneath that level, and one of the tools that we use for that is plant growth regulators.

We can look at ways to variable rate that PGR by using something that's very pretty common like a green seeker. We can actually connect green seeker reflectance to rank plant growth, and that gives us a pretty good correlation to aggressive plant heights to try to do some variable rate work in the plant growth regulators. My colleagues, Dr. Gary Roberson, here at NC State, he just had a PhD student finished up working on this project, Josh Brown. We're really excited to see the outputs that come from that.

Now, if we look in season assessment, mid season, we can always fall back to the UAV tools as that's becoming more and more popular. What we don't have are real consistent set of tool sets across all the cotton producing states to look at what can we detect early in the season that go along with our official variety trial programs to compare, not just yield, but are there other factors that we can start assessing about those varieties or hybrids that are being introduced into these official variety trials. What can we detect remotely, what are the right tools?

Working a lot to try to create those protocols that work very well across state with my colleagues all across the cotton bill.

If we look late season harvest, interesting environment. There's a few different machine types, there's stripper cotton, there's picker cotton. Although it looks like Deere's led the way so far with this wrapper picker that you see so much out there now with the yellow wrapper, the pink wraps, cotton modules. They still are doing a lot of processing of that data, they're collecting yields you can get from this cotton harvest.

One thing is we're looking to do is to bridge the gap between the field and the gin. After cotton is harvested, it gets wrapped, it gets the module gets made in the field, and then it gets transported to the gin. Gin is where seed cotton is converted into a bale of cotton, that is seed free, that's clean fiber that's been baled up before it goes to whatever end user it's going to.

Each one of those bales get sampled for quality. We can look at fiber quality, things like the diameter of that cotton fiber, the length, the uniformity of the length. All those things are important to the buyers who those properties affect during the spinning process and weaving process, the quality of that shirt or sheets or whatever it is that's being made from that cotton down the road.

What we're actually trying to do now is to collect that fiber quality data and track that bale all the way back through the ginning process, through the gin yard, all the way back to the field to a specific piece of ground. This bale came from this point, and these were the fiber qualities that came from this area.

Precision Ag podcast cotton

















Now we can try to make decisions about what can we manage to improve fiber quality throughout the entire growing season. We're pretty excited about the fiber quality maps. I think I had some colleagues in Australia beat me on getting a fiber quality map made first. Good for them. I'm glad they got out there. I think we'll be some of the first in the US to get these maps all the way through the process. It's a lot of data wrangling and trying identify the gaps between how a machine generates data versus the gin generates data and how we collect and go and analyze it.

Tony: It's obvious that harvest is really a key focus when it comes to the precision agriculture as far as cotton goes. There are a lot of opportunities, it sounds like in the other instances, planting and the replant decisions and what you're doing there. There's a lot of things that can be added or changed down the road. Aside from the fiber quality mapping, is there anything else new and exciting that's coming down the pipeline that you could share with us?

Jason: Yes, absolutely. There seems to be a lot of interest in what can we do at harvest with, imagine this, robotics. Cotton fiber and cotton yield is all very determined by what position on the plant that the cotton ball came from. If we can rethink harvest to now include a handful of small robots that go through a field, they can selectively harvest the cotton balls that are most mature and at the right timing to be able to pull the best fiber quality out of that field. [inaudible 00:20:19] From a single pass large harvester to a series of distributed robots moving across the field, selectively harvesting those cotton balls as a really exciting new initiative that some of my colleagues across other universities and economy corporate are really working on.

Tony: I know robotics is a huge push when it comes to the autonomous tractor or it's really cool to hear about the robotics as far as cotton goes. I had no idea that the quality of cotton and the fiber quality and whatnot, it all has to do with that plant and it comes down to your typical agronomics of corn and soybeans. Everything, it happens in cotton too. It's neat to hear about the robotics and how it can be utilized.

When we talk about all of the precision ag; the mindset, the practice the hardware, how is this stuff going to be utilized in cotton, whether it's the grower using some of this stuff, or it's the information that's capable of being viewed by the gin? When we think big picture, what do these technologies and data look like in the cotton industry?

Jason: Yes, I think what's really interesting is that we're really meeting the promise of what the technology was there to do. We've seen this happen in precision ag before, we saw in the first rounds of [inaudible 00:22:01] mapping or all these other tools were first coming online, a lot of people got excited. Then there's this gap as people realize it's a little bit harder than what we thought it would be. Now I feel like we're settling into this place where we're actually needing the promise of what precision ag was originally envisioned to do.

I think if we look across the board, the enabling technologies are the machinery data and the canvas and those systems that allow us to get that machine data. Then ubiquitous or semi ubiquitous in my area, cellular data coverage, so we can move data off these machines near real-time and collect it in a way that can be utilized very quickly. Also, I think creating these tools, whether it's a cloud-based or local machine based tools that allow us to get access to this data quicker and process it and push through to the meaningful relationships behind it.

Absolutely, if you look at our production system as integrated as cotton, there's all the way from seed through growing and harvest and have it all go to the gin. The gen being a whole another layer of data that goes on. It's very similar to some of the other markets perhaps like some of the vegetable markets where the shipper or packer plays an important role in the value chain. That last data gap really happens at the gin.

I know there's been incredible work in the software that run these gins now. What's amazing to me, a modern cotton gin, there's so much material that goes through it. It's really amazing how clean and safe it really is. Taking this raw cotton and they're creating a clean product that's ready to go into manufacturing. Closing that gap between the field and the end use really happens at the gin. Right now it can be a manual process, it can be somebody writing down a serial number on a module that's going in and then tying it to bale identifier tag.

It could be very manual, but creating the right tools that can read the RFID tags of the modules as they're coming in and create the right data pipeline to connect that to the bales, that's what's really going to close the gap for us to get this full identity preservation solution going on. So we can know this bale that was used to make these products came from this field. That's what really gets me excited.

Tony: Just hearing you talk about the tremendous opportunity that there is in the cotton production world of agriculture is just really exciting to hear it. Like we had talked earlier. We've always thought about precision agriculture, one, as a hardware, a component standpoint. When, again, you guys are looking at it from more of a practice standpoint, and you think precision agriculture and you always think corn and soybeans, but there's so much more to it when we start looking at these specialty crops and everything that you guys are doing down there at North Carolina State.

It's really cool to hear the different things, the way to look at things, and what you can do. With all of the research you've done, Jason, and the research studies that you're doing, is there any success story you'd like to share with our listeners about any of it?

Jason: Absolutely, but my success story sounds a little bit more like a failure when you first hear it. I was working with some of our local peanut producers, looking at field efficiencies at planting. Peanuts require a little bit of soft hand so you see some producers using some older style planters, not a whole lot of electric drive in that market right now, ground drive systems. This particular producer planted 800 acres with a six-row planter.

We were just looking, we tapped into their canvas to monitor their machine performance and how long it took them to do that job, look at their field efficiencies with that particular equipment complement. I looked at what I thought were some idle points in the field and realized that, "Oh, well, it looks like [inaudible 00:26:40] point of getting the refill, and there seems like there should be an easier way to do that."

I started talking about this data and this outcome, then I presented it, and I was telling the story and what I thought had happened here in this field. Then a producer says, "Yes, that's one of my fields and that's not what we do at all." He was very kind, he definitely could have [unintelligible 00:27:11] me over the coals.

I think that my success story there was to always make sure that we keep the data in the context for what that local producer is doing. Because it's very easy for us to think, "Well, we've got all the data and we can see and get all this insight and make all these decisions."

I think that we have to always remember and always be very aware that this happens within this larger context of Ag production. There's reasons why certain things happen a certain way. If we don't listen first before we start trying to talk about all this whiz-bang, the insight we can gather, then we kind of wind up making conclusions that maybe don't reflect reality.

Tony: I couldn't agree more when it comes to the in context, out of content. If we start looking at things out of context, it gets really messy in a quick hurry. That is a really cool success story to hear how you perceived it and how it was looked at.

If anybody wants to learn more about what you're doing at NC State, or maybe they're curious about some of the cotton production stuff, where can they go and who can they talk to, to learn more?

Jason: I'm at NC State University. My lab and my team, we're called the Advanced Ag Lab. You can definitely look us up on the Biological & Agricultural Engineering website at NC State University. I would always point you towards Cotton Incorporated. They've been a sponsor of my work and/or a producer-driven organization that really supports research and extension work. They dawn all through the value cycle of cotton. There's a huge number of resources via Cotton Incorporated that the producers can go to and anybody's interested about cotton production, Cotton Incorporated and their suite of sites really does a great job.

We're also in the process of standing up an extension portal. Part of my job is extension. I work with a lot of other great extension people across a range of commodities and technologies. We're in the process of standing up what's called the Digital Ag portal at NC State University. You'll be able to find that here when it gets launched within the next couple of weeks. At digitalag.ces.ncsu.edu. Of course, you can find me on Twitter @jkward

Tony: Wonderful, Jason. I just want to thank you. for sitting down with us today and talking about your perception of Precision Agriculture and also filling us in on so much of the information when it comes to cotton and the great opportunity that is out there. So, thanks again for doing this.

Jason: My pleasure. Thank you so much for the chance to talk about a crop that I think is really important and I really enjoy working with.

-----------------------------------------

Learn More:
Take a listen to all our episodes by visiting our website's Podcast page, or tune in wherever you listen to podcasts (SoundCloud, Stitcher, iTunes, and more). 

Have a story idea or a precision ag topic we should highlight? Connect with us on social media: Instagram | Facebook | YouTube | Twitter and connect with podcast host, Tony Kramer on Twitter at: @RDOTonyK.

Thanks for tuning in!