ydqa ep 21

John Dickow: [00:00:00] hello and welcome back to another episode of Your Drone Questions Answered. I'm John Dicko here to answer your drone questions, or at the very least, find the person who can answer your drone questions today. Our question is, when it comes to mapping and processing crop and plant health, it's a standard R G B camera on a drone sufficient, or do you need an N D V I sensor?

There's a little bit more to this question, a little bit more context to provide, and we'll get into that in a second. But I want to introduce our guest who's here to answer this question. That's Matt Donovan. He's the c e O of Agriculture Intelligence Inc. He's also the Chief Product Officer of Agro View or Agro view.ai.

Matt, thanks for joining me.

Matt Donovan: I'm glad to be here.

John Dickow: So first let's get to know you a little bit. What do you do? at Agriculture Intelligence Inc. Agri view.ai, tell me about your businesses and how you use

Matt Donovan: drones. I. as you said, I'm the, the c e O of Agriculture Intelligence and the Chief Product Officer of our flagship cloud-based platform, agri View, agri View is, an award-winning patented cloud platform, dedicated to the perennial tree crop, [00:01:00] space, invented by Dr.

Yiannis Amdis at the University of Florida. and I founded, the company, to bring this technology to market. To help, perennial tree crop growers like citrus, almonds, apples, avocados, and those kinds of things to help them with inventory counts, per tree health metrics, as well as full field, Digital leaf based nutrient analysis. my day-to-day is usually leading, the team of software engineers we have, as well as getting, direction from our chief science officer, Dr. Tis on the overall vision of how we can take raw information or raw data from various sensors. Refine that into an actionable piece of information, and then hopefully translate that information into actionable instructions, to infield equipment such as variable rate spray systems.

John Dickow: Okay, so I'm really glad we found you then to answer this question 'cause you sound like you, you fit the mold perfectly. so let me go ahead, let me dive into this question a little bit deeper. [00:02:00] This was a question that was submitted by a listener or a viewer. his name's Bradley. I'm just gonna read it out.

In mapping and processing crop and plant health for farmers and N D V I sensor is the most direct method of capturing the required data. However, In using web O D M I found there is a model for crop health that is produced from the standard R G B camera on my MVE air twos. Is this post-processing software adequate enough to produce a reliable plant health model for farmers, or will I need to purchase an N D V I sensor and drone just to map crop health?

Matt Donovan: I probably give the caveat that, there's all kinds of new innovations in the sensor space. but generally speaking, an R G B collection can produce indications of crop leaf density and greenness.

and it's sort of very dependent on the camera and the conditions of the data collection itself. But typically to do an N D V I, you would need a multi-spectral collection. That's sort of the simple answer, to the question. there can [00:03:00] be with, with a multi-spectral sensor, the sensor itself along with the, the processing system, the O d M engine for it, for example.

it can produce an N D V I. But N D V I, is valuable in some respects for certain crops. but it does have limitations. it really wasn't developed. For the purposes of what the modern precision agriculture, movement really is. it's a fairly old, it's nearly 50 years old, the N D V I, standard, if you will.

it mostly was, uh, Developed by NASA to try to help them understand where vegetation was and wasn't. it has been reliable and it's been an indication of some general, health, of differing crops. But, uh, in, especially in the perennial crop, the tree crop space or, the specialty crop, space like strawberries or blueberries, things of this nature, it's not quite as, precise as what, growers, small growers [00:04:00] or even commercial growers are looking for.

when they want to have someone come out, do a data collection, process that information into a, a visible map, and then apply, any type of, uh, model to produce a refined informatic that they can use.

John Dickow: Well, and so let's talk about maybe what is the best practice what would you really need to get, good overall?

Mapping and processing of, of crop health. maybe you can even talk about what you even use, or what you develop,

Matt Donovan: I guess to answer the specific question to the, person who wrote the question, they would need a multi-spectral, typically a, a five band multi-spectral sensor.

they would need a five band sensor that has the R g b, the red green, blue. they would also need the near infrared band and the Red Edge, there are sensors that are the five band variety on the market, and there's some, some very popular brands, uh, that would provide that.

to even attempt to do an N D V I, at least related to the, to the person who asks the question, uh, they would need to buy a five band sensor to do, a competent. [00:05:00] Representation of an N D V I. then more broadly speaking, and certainly, you know, not to sort of show our platform, but the, the avu.ai cloud platform is, is really trying to take a collection of different data sensors and then, assemble that information in a very specific way.

And then overlay that with the different models that would, um, be used to produce the kinds of things that the, the, the person who wrote the question might be interested in. And for what we do at Agra review, it's to really provide things like the inventory count. in tree crop agriculture, it's important to know, um, just like any business.

How many of these things do I have? If you were running a coffee shop and you didn't know how many, coffee cups that you had in your inventory, that would be a real problem. When you open the doors and people come streaming in and saying, Hey, I need my hot cup of Joe, and you don't have enough cups to serve your customers.

because of the scale of [00:06:00] agriculture really in any space, whether it's annuals like corn and rice, or whether it's in specifically, uh, related to, to tree crops and perennial crops, the inventory of what you have is really important. But drones are a really important step in tackling the scale.

Of these growing operations, like most people, like even people that are out in the field sometimes don't realize how big a growing operation can be. Even a modest growing operation of a thousand acres, it might take you. 20, 30 or 45 minutes to get to the other side of the acres because of the density of the growing area and the sort of limited space and the, the conditions of the farm road, around the growing space.

So drones and other aerial data collections give us a really good opportunity to collect data in mass, which is certainly what this, person that's asked the question is talking about. But then it's about what we do with it. The raw data is just raw data, and I think most people [00:07:00] that, have a drone, they pull the SD card off and, and they get all these folders that are in numbers.

So the Avu platform looks more towards a way to allow someone who's collecting data or someone who's interested in the result, perhaps like the grower, to utilize a drone surveyor like the writer, and To couple that data collection and then upload that into a system that can assemble it completely, provide rectified ortho mosaic picture, and then run models to do something like count the trees.

Or count the gaps where there are no trees, a missing tree, model. And then to run another model that evaluates each and every tree, like the Agri View platform itself produces a leaf density index, which is a an R G B measurement. But provides, uh, and this is kind of an interesting way to. Uh, back to the question, A leaf density index can be achieved with an R G B, but you need to have a model that actually understands how to detect that [00:08:00] greenness or how to detect that leaf density.

We also provide, things like the canopy area as well as the height of each and every single tree that's present. And with the height and the canopy, we can give a general volume calculation, uh, inside of our platform now with a multi-spectral. Collection, we can take that a step further. And another model that we do for citrus is actually providing a digital leaf nutrient analysis, which typically is, again, another process that is very challenging because the vastness of a growing area, and it requires people to go in the field with a bunch of trucks and a bunch of people to go pull leaves off.

And this is not soil analysis. This is actually a leaf tissue sample. Pull those tissues off. Bag them properly, tag them properly, and then send them off to a lab for analysis to then have some sort of timeframe, two weeks, three weeks, six weeks, however, [00:09:00] backed up the lab might be to then be given a refined informatic or insight into that area.

But with a Multispectral drone collection or a Multispectral data collection, we can examine the entire area. So you might have a 50 acre plot that instead of having, you know, dozens and dozens of samples that have been manually taken and sent to the lab, we can utilize a multispectral data collection, process it into our system, and we can examine all 50 acres.

So that 50 acres then can be represented into some standardized management zones. that indicates something specific to, for example, like a tree crop grower would be keenly interested in knowing where there's low or deficient levels of nitrogen or phosphorus or potassium, the sort of macros the protein.

Fat and carbohydrates of the tree world. that our system actually uses a multispectral data collection to provide a comprehensive, full field [00:10:00] understanding of that. The Agri View platform is now coming to market to assist those kinds of, uh, stakeholders, whether it be growers, crop insurers, other people that are, market strategists that are interested in how is the citrus crop of Florida?

That's the kind of things that, that of course, our platform is aimed to serve in the marketplace.

John Dickow: Well, that was actually gonna be my next question was, you know, we, a lot of our, our listeners, our viewers, and a lot of the people that actually we actually have on the program are sort of people just starting off.

Um, whether they've been in the industry for a while and they're just starting to use a drone as a tool, or they've been interested in drones and just starting a business with. Drones drone. When it comes to Aview ai, I mean, is it something that's pretty out of the box, something someone can easily access and kind of start using right away?

Or is it, something that, you know, requires a little bit of background knowledge and, and training to, to understand and really use to its fullest

Matt Donovan: potential? If someone's asking the question of like, how do I get into flying drones and surveys like, that's probably a whole nother podcast.

Yeah. I would succinctly say that data acquisition [00:11:00] via drones, has differing levels. I would say that in our types of applications, someone that's getting into this business would, would really need to understand. Wide scale data collections and then find the drones, which are gonna be more expensive, that , can provide a better, result, permission.

So if you want to think about it in acres or square kilometers, the drone starts to get into like some limitations there, but it can be very, very useful. It just depends on the drone that they invest in to buy as far as how the aview AI platform really works. Um, it would be more along the lines of we might wanna work with them as a data provider, which we can certainly, talk to them about, target areas.

Um, we would want to enroll them more as a, a vendor or supplier of our system, which certainly. If you're out there, go to Agri view.ai and, and find my email. by all means, depending on where you are and the crops that we might be able to serve, we would be able to work with them.

And then it would be [00:12:00] more about them making the data available to us so that then we could go to those crop growers and say, Hey, we have a, a reliable source, uh, a proven source of, of. Of data acquisition, and we can pair that to give you the kind of results that are gonna show you your inventory count your gap, count your perjury health, or if they have a multi-spectral, uh, payload, then we could even provide them this very advanced nutrient analysis.

It kind of works, in most, industries that there's a stream of result, the growers that we're working with, the crop insurers that we're working with, or, researchers or perhaps state officials who wanna understand the inventory of an area. they might come to us for that.

And then we would be able to engage, uh, people who are doing the data collection and the data acquisitions, as really a partner in that space. So, there would be a business opportunity for some enterprising, folks that are out there in the drone, space. It kind of ranges from like amateur weekend collector and then to [00:13:00] someone who's, doing it for work and then someone who's really doing it as an enterprise.

we would probably fit into the, the second and third tier. and then it would usually be calibrated against. Have you ever collected. Done data acquisitions for this type of crop. Have you ever been in this type of thing? And then most of the rest of it, as we all know, uh, some of the flight planning softwares and all of that kind of stuff, as long as you start at the right spot and heed the sort of the side lap and overlap and some of The general specifications we would put in place for a surveyor, it can be something, that would be an on-ramp towards, a potential business opportunity or an opportunity for them to make a little extra money. and then it's about whether the result was good. we have, the pilots that we've engaged, we typically, have our own specifications.

They do some test flights. we look at the data, the result. Usually improves once you've done it a few times. A little muscle memory if the drone does a lot of the work these days. And if you configure, have the right [00:14:00] drones, to do these kinds of flights, it can be something that can be scaled up.

Pretty well. If you're in the southeast, if you're in California, if you're in Texas or Arizona and you're anywhere near Citrus, I definitely, would want to talk to you.

John Dickow: Awesome. Awesome. Anyway, we're, we're running short on time here, but I wanna just throw one more question out there. 'cause you mentioned drones and, and there are certain specifications that are needed when it comes to the drone. what would kind of be the type of drone you would need in terms of baseline, specifications?

Or is it possible to retrofit a drone, with the needed sensors? are a lot of these drones coming with the sensors already that you would need?

Matt Donovan: We've seen everything From the D G A I quadcopter to the wingra fixed wing.

and then the payloads are, like what can fit on it. in the US with the, the drone ban and some other things. there's a, a select list of us drones. They typically, the advantage of it is they usually are, are a longer range drone.

The disadvantages, they're usually more pricey. I would probably defer a little bit as far as [00:15:00] what we could use. It more comes down to the result of the survey itself. if we have someone who has a. I would say a above weakened enthusiast drone. the DJs, the quadcopter, varieties, we're not a drone company, so I'd caveat that.

I would say it, it would, for us and any kind of commercial endeavor, you, you kind of need a drone that has some of those capabilities. Just because if, if they're trying to make a business of it, the smaller drones, the, the sort of the enthusiast drones, they're never gonna have the scaled capacity for collection.

we're kind of agnostic. if our grower has their own drone and if it's good enough and they're doing their own data you can upload your own data if you'd like. If you don't wanna do that, then we, we want to go find the people, who can, and, drones are gonna kind of range dependent upon the, the payload, it really, for us is about the result itself.

John Dickow: I appreciate that. I do just want to give you an opportunity to provide any final thoughts. going back to this [00:16:00] question of, using the right sensor or even if just a piece of advice for somebody who is interested in kind of getting this and might be new to using this type of technology for this type of purpose within agriculture,

Matt Donovan: we need to talk to, to growers and crop insurers and people that want to know about these tree markets. our system is able to produce, scaled very precise information. I. About the inventories, as I've mentioned, and about the health of these areas. and then it's really about growing the ecosystem of, quality, data, surveyors, that would be using drones, to do this.

it is a growing field. those opportunities, they're kind of emerging, but there's somewhat clear, like there's people that want to be out flying and they don't want anything to do with the stitching. They don't want anything. Kind of goes back to the question is Sure.

Your, your question or sort of was at least taking that next step. But there's all kinds and, it's a growing space for everybody. and more importantly, I think overall we talk about. Reducing carbon a lot. Like that's probably the other as a [00:17:00] closing thought is. in a lot of these spaces, the only way to try to get a tree count has been a manual count.

It's been a couple of guys or gals, in a truck. with clickers going up and down, kind of doing their own survey, and then keeping tabs on a clipboard there's been advancements to that, but we're still talking about having more trucks out in the field and any way we can take a step back and say, Hey, you know what, we can do this much faster.

Much more effectively, even more accurately. And all we had to do is be on the side of the field, you know, put a drone up, collect that information, no more need to put, an additional fuel, burning vehicle out there for the purposes of doing something that we now can do, through another means.

we push for that, as a core part of what we are as a company is to be a clean technology, but an actual realistic impact. Like we can't solve the world's problems like our company is at least realistic about With what we do as a company and with suppliers that are competent, you know, data collectors and [00:18:00] surveyors, this is an achievable goal in this space and, and every little bit helps.

John Dickow: It's a really cool innovation and really so glad that we had you on here, on this, uh, program. I feel like we could talk through, you know, two or three more episodes if we wanted to, and maybe we will, because we do get a lot of questions, when it comes to the agriculture space and the mapping space itself.

Thanks again, Matt Donovan. He's the c e o of Agriculture Intelligence Inc. The Chief Product Officer of aview.ai. Check out aview.ai. If you're interested in this kind of topic, if you're interested in this kind of question, I think there's a lot to discover there. and you can submit your own questions, uh, that we will find the answer to here on your drone questions answered.

You can submit those questions on Y dqa.io or if you're a part of the drone Launch Connect. community, which is Drone Launch Academy's private community. feel free to shoot us a message over there too. We'd be more than happy to take your question and, and find the answer to it.

But in the meantime, we'll see you in the sky.