top of page
  • Writer's picturePrecision Ag Reviews

Ep. 21: Automation and the Future of Swarm Farming with Dr. Ajay Sharda

When we ask people what technology they are most excited about in agriculture, one of the top answers we get is around automation. Thankfully, our guest for this episode – Dr. Ajay Sharda of Kansas State University – leads some of the most exciting research about using robots in the field. He is a thought leader in this area, and sheds light on the seemingly endless possibilities this technology offers as our industry continues to transform. In episode 21 of Precision Points, we unpack what it will take for automation technology to scale and where it can go once we get it in the field.

Below are insights on just a few of the questions Dr. Sharda and I discussed. Make sure to tune into our full conversation to dig even deeper!

What is ‘swarm farming’?

For years, equipment has gotten larger and larger to help improve efficiency and timing. When you take the operator out of the cab, however, the risk and liability of running such large machines autonomously is so high that Dr. Sharda believes we will see a transition to smaller, more compact robots. This is what we call swarm farming – a team of robots working effortlessly together throughout the field. The autonomous machines need to work toward a common goal, while staying safe from each other and keeping everything else in their environment safe, as well. For this to work, communication between devices will be key.

How will autonomous farming work logistically?

Experts predict that, with swarm farming, the operating functionality will be autonomous. However, there will still need to be humans involved to know where the robots are, and to be able to take them over and manipulate their actions if needed. Researchers are currently working on automating things like battery changes and product refills, which would rely on the robot being in constant communication with humans to let the base station know when something may need to be addressed.

The area that robots can cover varies on the machine being used. The robotic platform that Dr. Sharda is working with is built to work within the canopy. The current system they are developing is capable of doing six acres at a time; however that is based off of a 50% reduction in spray volume (the robot is designed to target areas that need addressed rather than blanket-applying chemistry). Dr. Sharda did say that, as the swarm of robots enter the field and begin to give feedback about their actions (e.g., spraying very often versus infrequent spot spraying), the human that is in the loop with the data can make the call to change plans and go in with a larger traditional rig if a blanket application is needed. The vision systems in the robots will give you eyes in the field near real-time.

What are the benefits of swarm farming?

Not only could robots help with labor issues, the smaller equipment should allow the grower to get into the field in a more timely fashion. The machines that are designed to work within the crop canopy may allow growers to apply value-added applications they may have passed on in the past or couldn't manage due to equipment size.

“We are targeting in-season management,” Dr. Sharda shared. “These are the things which are difficult to do, and these are the spaces which are difficult to access. And I think that can really push the benefits in terms of the cost saving, site-specific management aspect of it – the environmental component is on top of it – but also the valuable information which I get about my area and my crop.”

Not only can these systems reduce manpower requirements in the cab, they may also reduce the effort it takes to do scouting – and perhaps even do a better job than we do today. While robots are already in the field doing other activities, you may be able to get these devices to multi-task with scouting or even soil sampling.

“If you have systems that are going out and have vision systems to scout for insects, why would we not put additional cameras or other sensing systems to collect additional information?” Dr. Sharda asked. “While I'm doing some other tasks, I would like to harness the cost, expense and the effort to collect other information.”

Another benefit is that this equipment is being built to where it should scale to operations of all sizes. This means growers of all acreages could be running newer machines and receiving the data output from them. This makes autonomy a viable option for large-acre rural growers and smaller-acre urban and suburban farmers.

“Our intent from the very beginning was that this system should be modular and scalable. These are independent machines in themselves; they are self dependent on navigation and path planning decisions, in terms of making the application systems, which are smart enough to implement site-specific applications, and on the go while they are doing all things, collect every single piece of data,” Dr. Sharda said.

What are the challenges of autonomous farming?

There is a lot to be worked out before we are ready to scale autonomous tractors. One of the biggest challenges currently is managing the extensive amount of data these machines could supply.

“So that is one of our major goals and dreams – and I think a challenge as well – is how to maintain that data stream, data structures, all of those things; how do we do that?” Dr. Sharda asked. “The challenges with data sets from our machines, which is really organized at this time, is we still find that difficult, and fall short on timely managing those data sets within our system.”

The other challenge will be for farmers to make the mindshift change from a mechanical and hands-on approach to hands-off, but more intensive data processing.

“We are seeing this is history repeating over and over again,” said Dr. Sharda. “When you were farming with horses and animals, and then tractors came in. And I know, and I've seen a lot of folktales where people said, "These tractors are never going to overtake what we do and how we do it." And now, I don't think anybody can farm without a tractor in this time and age.”

Where to learn more about swarm farming?

Dr. Sharda says he would love to connect and have farmers be a part of the conversation about autonomy in farming. This is one of the many projects within what he calls the FARMS (Fusing Automation and Robotics for Agricultural Machine Systems) Lab at K-State. You can follow his work there and on Twitter.

What do you think the future of farming and automation looks like? Leave a comment below to let us know!



Host: Morgan Seger

Guest: Ajay Sharda

Morgan Seger: (00:22) Welcome back to Precision Points, an ag tech podcast from I'm your host, Morgan Seger, and in each episode, we strive to bring you unbiased ag tech information and ideas. And today's episode is a great one. I am joined by Ajay Sharda, an associate professor in biological and agricultural engineering at Kansas State. And where Ajay is focusing a lot of his work right now is on swarm farming.

Morgan Seger: (00:51) So, talking about autonomous vehicles in the field that work collectively together to complete operations on your farm. And I love this conversation, because he breaks it down in a really easy-to-digest manner. And we start from the beginning to not the end, but what this vision looks like, because this is definitely an evolving part of our industry right now.

Morgan Seger: (01:16) He spends a lot of time talking about how many decisions the machines are going to have to make compared to what a human operator would make and what this autonomous structure is going to look like, and even touches on how these devices may be able to multitask in the future and save time and energy for growers. And it's just a really exciting conversation for me. So, I think you're going to love it. We'll get right into the interview. Ajay, welcome to Precision Points.

Ajay Sharda: (01:47) Thank you, Morgan. Thank you so much.

Morgan Seger: (01:49) Well, we had the opportunity to catch up a little last week, and I am so excited about our conversation on swarm farming today. Before we dive into the technical goodness that you are going to share, could you kick us off by just sharing a little bit about your background and how you got to where you are here today?

Ajay Sharda: (02:06) Yes, I'm an agricultural engineer from my training back in India, and was fortunate enough to be around a lot of different technologies and a completely different type of agriculture there. My time back in Auburn working on different precision ag technologies there, working with John Fulton, who was my PhD mentor, learned a lot of different things and a very different perspective of what are the needs of the producers, what are they looking for from a technology standpoint.

Ajay Sharda: (02:43) So, my PhD time there – a little bit of a year at Washington State University – there was a lot of robotic automation work going on, and I'm finally here at K-State. So, everything which happened in my last few stops, whether it was precision ag from row crop standpoint, or automation robotics from tree food industry and how that industry was really moving towards the automation robotics, more so from a perspective of need, the labor, which was people not being available, and large acres high-value crops ... those to work with.

Ajay Sharda: (03:22) So, here we are in Manhattan, Kansas, at Kansas State University. It's been my home for the last seven plus, eight years. And it has been an exciting journey working with large ag equipment, but on the side it seems like robotics is going to be something everybody's going to ask for. There will be some needs which are driven by the type of work we are doing, and there'll be needs which will be driven by the social economic changes, which are coming in our landscape, in our rural communities where we will need some help to get some of these things done. So, that's a little short blurb about some of the motivation to do this work.

Morgan Seger: (04:11) Yeah. Well, and it's interesting that you're coming at this from a lot of different perspectives. Are you primarily working in row crops now?

Ajay Sharda: (04:20) Yes. Most of our work in Kansas is with row crops, corn, soybean, some sorghum is definitely one of our key crops. And obviously, things go around on wheat and canola, but yes, row crop is the focus.

Morgan Seger: (04:37) Got you. Well, at the end of each episode, I usually ask our guests, what's one thing they're most excited about, and most of the time it has something to do with automation or robotics. So that's part of why I'm so excited today to learn more about it. And for those of us who are just getting familiar with it, how would you define swarm farming? What does that mean?

Ajay Sharda: (05:01) The way I understand swarm farming is where we have multiple robots or autonomous systems working in sync with each other to achieve a common goal. That there are a lot of us who have been assigned a task to go out and start doing their thing. So, that's what I foresee swarm farming is, because thing is, from a robotic automation standpoint, we are moving from this step because smaller systems are, from a safety standpoint, from a standpoint of liability, which some of the large equipment has, so this is slightly forgiving and there is also a factor of cost and functionality on these things.

Ajay Sharda: (06:01) That's how I consider what swarm farming is. The way you would look at it, if you have a swarm farm, you get onto the farm you'll see multiple of those machines, very nicely working in a preset pattern, safe from each other, safe from the human interactions, and are not being able to hurt anyone.

Morgan Seger: (06:27) Got you. So then, do you expect these robots to be fully autonomous or is there going to be human manipulation in how they're operating to help them with that effortless pattern through the field?

Ajay Sharda: (06:39) Great question, Morgan. I think, from a perspective of swarm farming, these autonomous systems, what we foresee and what I foresee, should be autonomous from a perspective, or the operating functionality standpoint, there has to be humans in the loop in terms of being aware about where each system is and what the systems are doing. And to even overtake at some point where, these are machines so something will fail whether it is from a hardware standpoint or software standpoint.

Ajay Sharda: (07:16) Although there is a solid intent that these systems should be intelligent enough that they should foresee a potential issue, or say for example, a battery is going to die down and they needed to recharge, or even in the engine-powered systems, or gas-powered systems as well, that they should be able to come out to ask for help, or ask for recharge, refilling, and all those things.

Ajay Sharda: (07:48) But from all of those perspectives, our definite goal is that they should be completely autonomous. They should be able to know where they are going, they should be aware of their location, they should be aware about their functionality, and they should be in continuous communication with all other systems in the field. Let's maintain constant communication with the base station where you definitely have a human in the loop who is taking care of all those activities.

Morgan Seger: (08:17) Got you. Okay. Now I have a couple more questions off of what you said. So, the communication and the way it's traveling through the field, are you thinking we're going to use the type of GPS in that communication that we are already using on our larger equipment, or is this going to need something different?

Ajay Sharda: (08:36) Yes, there are lot of different things which people are trying in terms of different labs, and different approaches, but we are definitely going towards using the GPS, the technology which is out there, you sit in a tractor, or a planter or a sprayer, you set your AB line, you have your guidance pattern and the GPS works great. So, we have very good technology from a GPS standpoint. We are going to go with that, and I can share some of the things which we are planning.

Ajay Sharda: (09:10) We are definitely going to go for frequency, GNSS receiver, which have RTK correction or capability, they should be able to go out and anywhere in the field. You've also asked another thing about the communication part of it, which to me is one of the most important things in terms of how do we maintain communication in large spaces. Ag spaces are large, there are different terrains, there's an aspect of crop canopy in the geometry of the crop itself.

Ajay Sharda: (09:44) So that part too, from my perspective, is going to be very critical for our community to develop, so that these systems can talk and maintain a continuous communication about their whereabouts, continuous communication about the health of the system, what they are doing, what the needs are, and so on and so forth.

Morgan Seger: (10:09) That makes sense. And then as you think about those things that it may need, whether it's a recharge, or a refill, do you think that we'll get to where that is automatic as well? Where it says like, "Oh, I'm getting low so I'm going to head back to the station." Or is that a stretch?

Ajay Sharda: (10:26) No. Another great question. We were just talking about this the other day, so we are slowly getting into it. We talk about some of the navigation things. We talk about the communication aspect of it. Logistics will be, I think the next very important thing. So, what do we do in terms of maintaining these in the field? In one of our projects, Morgan, we have developed a battery pack, which is in the shape of a small, I don't want to say it's a missile, but it has a conical front end. And I can actually take it to the spot where my robot is waiting for the recharge to happen, I can pull my battery pack and I can insert my new battery pack using an unmanned aerial system.

Ajay Sharda: (11:16) So, that's one improvement, which can happen. And that's very fascinating. It's amazing to see how easy that is, in terms of concept. But I think that's very much doable with the current unmanned aerial systems and how good they are, how stable they are from all that part. Another important thing you asked about, if we are spraying we need product, water or chemicals. If we are seeding, we need seed. If people are looking at soil sampling and all that.

Ajay Sharda: (11:52) So, definitely our goal is, and we have done some work in terms of how to manage their navigation and plant planning aspect of it. Spare thing right now is that, they will keep a complete eye on how the product is being used in terms of the ... part of it, then they'll also have an idea about how long the passes are, and what is the rate of use of that product.

Ajay Sharda: (12:21) And those systems should be intelligent enough to say that, "Hey, no, before I enter my new pass, I need to go back and refill, because I know that I won't be able to make this pass complete, and I'll have to go a longer distance before I can come out and get the refill done." And the same thing, what people are, our approach is both ways from a charging standpoint, something mobile where we can go in and do something at the field location as well.

Ajay Sharda: (12:50) But what we are seeing is that the crop canopy sometimes could be a challenge in terms of locating, and implementing or completing that task safely and quickly, versus that robot can come out much faster and we can just swap the batteries, or whatever, refill if it is a gas powered and that system can be back in the field much more efficiently. Logistics is definitely one of the key items.

Ajay Sharda: (13:21) We have different concepts; we are thinking about fixed stations where we are filling that, or there are nursing robots as well. So, their job is to carry the crop input need. So, they will just come in and... But it's easier said than done. We are thinking that there'll be some human interaction at that point, but the ultimate goal is that they should not have any human interaction. They should be able to go to a spot, or nurse using these nursing robots who are just patrolling your headlands and they just come in and they do a handshake, and get what they need and go back to work.

Morgan Seger: (14:13) Sure, sure. Man, my wheels are spinning on what this might look like. So, what would be the field size or field length right now, if they were to be carrying a chemistry for spraying weeds or something like that, where it would be practical to have an application like this?

Ajay Sharda: (14:32) So you're asking about the size of operation, or how much area they can cover?

Morgan Seger: (14:38) Yeah. How much area they could cover.

Ajay Sharda: (14:41) So, there are a lot of things, we are talking about so many things and I think, even if we gather everybody who is engaged in machine systems in U.S. right now, and make a bigger group, we might fall short on number of people we want to tackle this problem, but I can talk a little bit more from a perspective of the USDA, the AFRI, our NIFA funded project. So, our robotic platform is designed to work within the canopy. So, it goes within the rows and in the intent that it will have a capability of vision to see where the insects are and what the intensity is, and do I need to spray to control it or not?

Ajay Sharda: (15:26) Given those conditions, and to top it off that I'm under the canopy, the effect of wind is too little, that the wind temperature humidity is much better, which is more suitable for application. So, site-specific knowledge-based decision, less the environmental conditions including wind and temperature. We anticipate that we can really bring the use of per acre of chemical dramatically. So, the current system which we are designing is considering that we may not apply for more than 50% of the area.

Ajay Sharda: (16:09) Also, we can go out and apply six acres at a time before they will have run out and do any refills, which is not a small area. And looking at time and anything, if it can do six acres, and who knows what it could be a little bit more, it could be less as well depending on the severity part of it, but looking at the vision part of it that we want to, vision means vision of the concept. And if we can scout timely, if we can scout on a spatial basis, we can catch those bugs, those early, whether it is insects or fungi or any other things, we can control them on smaller spaces rather than really going out and controlling on large spaces. So, that's the intent.

Morgan Seger: (17:04) Yeah. Yeah. I hadn't really thought about the way it may impact. So, it would still be if it's within the threshold for use rate, it's just not spraying if it's not present, which could have a lot of benefits for our soil health and the way consumers are viewing what we're doing and also for our wallet, right? By not spraying excess chemistry out. So, that's really interesting, but with each year being different, how is the machine going to know? We talked about this. So, how many decisions is the operator in the machine today the human actually making, and then how do we interpret those through a robot, I guess is the question?

Ajay Sharda: (17:46) I think they are all good questions. These are the healthy discussions or ideation things. But this is great. From a commercial machine standpoint, Morgan, and what I've learned from the entomologists as they go out, they have some areas, they scout for bugs, and it's all hand done. And they cannot go into some of the areas which are inside, they are along the boundaries somewhere where I can easily walk and all that.

Ajay Sharda: (18:24) So, based on that smaller area, which is less than 5%, or even smaller than that, they make a decision. Whether should I go out to spray or not? Now, there are two or three caveats. One, I may be on a site which does not represent the true population and severity of the pest tested in that field. That means I am delaying my trip in the field to control it, that can result in further damage before I come back to scout the second time, and I may lose zeal on that.

Ajay Sharda: (18:58) The second caveat is that that site may be all representing what the pest population is. So, I made a decision to go out and spray while my crop doesn't need a spray. So, now I am not applying a product, which I may not need to apply. So, now there's a perspective of cost associated with that in terms of product costs, fuel and equipment and all that. So, our thought process is that if these guys go in...

Ajay Sharda: (19:30) So, these guys have to be intelligent. Like, if we send 10 of these guys in different parts of the field, and I'm getting a lot of feedback from these guys that I've seen a site, I've seen a site and I want to spray, these are there over and over again, then I have to at some point, I should quickly make a decision that there is no way I should be continuing this path.

Ajay Sharda: (19:54) I need to bring these guys out. I've got enough locations which are sending me a lot of signal, that it is good number of population that I need to control. I should be sending my bigger raid to control it. Those guys go in and give me a feedback every now and then, that, "Hey, I saw a site, and it's sprayed." So, some of the things which, Morgan, we are anticipating is that we will have a user interface where you can pull in, like a lumber, and a lumber will show you where it has gone. And it'll show you spots of those sites, which have severity, which needs control. And then spots where it saw and say, "But they are not severe enough to control."

Ajay Sharda: (20:43) So, those knowledge pieces would be so much valuable, not only to know how the insect populations are spread out within our ag fields, but also to know how do they spread. So, we are dealing with aphids, and we were a little smart in selecting aphids, because aphids they may colonize and then they stay there and they feed. They don't move. But there are migratory aphids as well, which are going to just go from one location to another location and make another site.

Ajay Sharda: (21:22) So, I think with these guys in like you were asking, how do we know? So, year to year, or growth stage to growth stage. So, this is where these guys will come in and they'll give us special information, they'll give us the intensity and the frequency in locating those sites. And I think that knowledge itself will be very valuable, both for people who are working with insects. It's also giving us some idea on how to control them. Does it help?

Morgan Seger: (21:54) Yeah. Yep. That helps. And then, for applications that are maybe more preventative, say micronutrients in season, then you could task them out or just go with your larger normal rate.

Ajay Sharda: (22:09) Right, right, right. Exactly. And I think Morgan something which I'm pretty sure that everybody, and I don't shy away from sharing that as well is that, if you have systems like that, who are going out and have vision system to scout for insects, why would we not put additional cameras or other sensing systems to collect additional information? While I'm doing some other tasks, I would like to harness the cost, expense and the effort to collect other information.

Ajay Sharda: (22:42) Right now I can take pictures of the plants every now and then, and start looking at how my leaves look in terms of nutrition or something else. I don't know. It could be other things, it could be, I can even take some soil samples. Every now and then I stop, I take my probe out and I just gather some soil information and just send it back and keep moving. I don't want to collect any soil, I just want to gather some information.

Ajay Sharda: (23:10) So, you have smartly pushed me towards another area, which is the scouting part, which is really, really critical. And it could be an independent thing, or it could be a piggyback, but what we foresee, and I see that these systems while they are out there and they should be doing other errands, just some information, what do you see, what do you see on plants in terms of color of the leaf? So, absolutely.

Morgan Seger: (23:37) That's awesome. So, my wheels were going there, but I'm like, "Maybe it's too much to ask it to do more than one thing at a time, because we're already asking it to go do it by itself." So...

Ajay Sharda: (23:46) No, so I think we haven't touched a little bit when we were talking about communication, that will be very, very critical. Not only we want to know where our vehicle is, what are battery power, health of the system, health of our application system is, but we really want to offload data as quickly as possible so that we can start maintaining those data sets, edge computing, getting on, or getting to the edge, doing some pre-curation of the data on the edge or directly pushing to the cloud, because we would like to get that data into the hands of the people who can make sense out of it, who can create ... from that data set. So, we really don't want these systems to keep working for days and we get access to the data like a week from now.

Ajay Sharda: (24:34) You won't need that, you would like that data right away. So, the intent is that we would like that information in the hands of the right people as soon as possible. Possible in real time or near real. So, from that standpoint, I just need systems on it. I just need to really, time in terms of when I want to push what data and how? So, that is one of our major goals and dreams, and I think challenge as well, is that how to maintain that data stream, data structures, all of those things, how do we do that? And the challenges with data sets from our machines, which is really organized at this time, but we still find that difficulty, fall short on timely managing those data sets within our system.

Ajay Sharda: (25:38) I myself is a victim of that habit, because there are a lot of things going on, and sometimes you don't really, classify or get the data in the right spot. And then when you want it, you don't have it. So, I think that is definitely one of the biggest goals. And we can plan to meet a year from now and I can tell you that whether we are able to make any progress on that and all, because there are a lot of things going on, specifically in our lab to get this part of the challenge taken care of.

Morgan Seger: (26:14) Yeah. Well, I really respect that approach because if we look historically, every year we're able to collect more data, and some people are using it really well, but if they don't have a system to interpret the data and even visualize it in a way where it makes sense, we're just going to keep piling up data. And even if they get it instantaneously, it's going to be too much for them to make an action on. So, making sure that communication is streamlined and easy to understand, it's going to be what it's all about. Otherwise, it'll just be an unused feature and we'll have robots in the field doing whatever we would have done anyway.

Ajay Sharda: (26:52) Well, it will be something where somebody has gone and ransacked the warehouse, because these guys are going to be creating piles and piles of data and dumping on you, and miss a beat then you're behind and you'll never be able to catch up, honestly, Morgan. And so, the problem in some of the ag within the production cycle also is that, not every operation is taken care of by the same guy. Sometimes I have a co-op doing liquid application, somebody might be doing... Even within a family, dad does all the spraying, and the son is doing all the planting, and somebody else does all the harvesting part of it.

Ajay Sharda: (27:36) So, if we do not have a common understanding of how do we want to name those farms, and I have a live example of it, I have a farm which is named like six different ways. And when I want to look at data, I have to ask every single one, "How do you name that farm? And how do you name that farm?" I have a data, but I don't know which data set to access to get that information.

Ajay Sharda: (27:59) So, I think those would be real real challenges, but I think if we can streamline, and if I have a punch of a button, I can access one set of things, on a boat, on a spatial temporal basis on a timeline, on a timescale. I think that will be huge. I think that will be tremendously beneficial for people to work with those data sets and create some knowledge behind that.

Morgan Seger: (28:27) Yeah, for sure. And for me being a smaller farmer, and we had talked about this on our call the other day, I'm excited that it may allow access to newer technology than what we usually use that is better at collecting this data. A lot of the stuff that we're using right now is from 30 years ago, because it's sized appropriately for our farm. So, I can definitely see this being an equalizer where it works for large operations and small operations alike. What work are you guys doing to make sure that it's scalable, for large and small, and then also, how do you see this impacting the urban chain?

Ajay Sharda: (29:05) And this is something which you have basically really put a seed in terms of that. Our intent from the very beginning, was that this system should be modular and scalable. These are independent machines in themselves, they are self dependent on navigation, and path planning, decision in terms of making the application systems, which are smart enough to implement site-specific applications, and on the go while they are doing all things, collect every single piece of data on the go.

Ajay Sharda: (29:44) So, now what it means is that these can be really excellent and a new way for smaller farmers to have access to state-of-the-art technology. So, state-of-the-art sensing and control systems, which comes all embedded with the data collection and data management solution so to say, which based on our discussion the other day, it could be great talking about a couple of hundred acres, 50, 200, 300, 400 acres, you guys can get these systems up for running, you have a user interface, which is taking care of what needs to be done while you're away doing your other things, or even if you are there as well like we talk about.

Ajay Sharda: (30:39) I'm fully anticipating there'll be some human in the loop, for some years to come. I think there’ll be an excellent alternate way for farmers like these to have access to a great piece of technology, plus the information and the knowledge which will come along with that, because they'll be just amazed to see what's going on, and how good it is, and how nicely the job was done, or every piece of the acre was covered, or the staking knowing that, "Hey, I've seen an ear on every plant and they are doing good. There's no insect or disease going on."

Ajay Sharda: (31:24) So, I think these informations will be just amazing. And I think we touched base on that a little bit in terms of, urban and suburban farming spaces where people are doing specialty crops. Specialty means vegetable crops, high-value crops, or whether they are in controlled environment or smaller acres, these systems can be just amazing because cost of labor, or doing some of those operations could be challenging.

Ajay Sharda: (31:57) These guys can really fill in to really advance the work efficiencies and cost of the product usage, but more importantly providing a lot of information on timely management of these crops, getting the best produce in hand and potentially increasing their profit margins, and higher quality of product and produce. Ever since we've talked about this, I think this is just fantastic, I see no reason why it should not be implemented in suburban areas.

Morgan Seger: (32:38) Yeah. I think there's a huge opportunity there, and it's going to be interesting to watch how it plays out, because I could see even in rural areas where they have more vegetables moving to this quicker, because of the need for all of the data and that constant feedback that they would be getting. Because it's not like they're out there once or twice a year making an application, they're already visiting those fields more often. So, I could definitely see it filling in there. It just has me thinking, what about the farmer that just really likes being in the tractor? What are they going to do? How do you see their role evolving to fit into a new operation like this?

Ajay Sharda: (33:14) Well, Morgan, it will be driven by people who will need some intervention like that, and we are seeing this is his history repeating over and over again, when you were farming with horses and animals, and then tractors came in. And I know, and I've seen a lot of folktales where people said, "These tractors are never going to overtake what we do and how we do." And now, I don't think anybody can farm without a tractor in this time and age.

Ajay Sharda: (33:45) And same thing happened when auto-steer came. When people said like, "No, I will have my own hand on the steering." But now if you are a farmer in Midwest, Dakota, or somewhere where you have a mile long pass, you don't want your hands to be on steering because, it's helpful. So I see that. And I've asked this question to a lot of people do we use, and I also teach the precision ag course. And it's pretty interesting.

Ajay Sharda: (34:13) I asked this question in my class in 2020, and I would say probably three out of 10 people said that they are either excited or would like to have. And I was a little amazed that these are the Generation X or Z or whichever generation this is. That they still think that large equipment and the way they are farming is what and how they want to do. And that is definitely a decision and a choice, which well, everybody has to respect that, but like I said, it's going to come from the need, that when I do not have any choice or when I really need support and I do not have support to get things done.

Ajay Sharda: (35:01) And then there are other people who want to really venture out. They don't want to say that, "Hey, I want to..." And here's an example of an up-and-coming farmer. He has his own land, but he's taking over some of the rented properties, or the farms from the families who are thinking to retire, their kids are gone, they don't want to do that hectic operation anymore. And he said, "Well, I'm waiting for these guys to come. I want these guys because they are coming from a different, I won't say a worst-case scenario, but their business model is quite different. They want to establish this more as a business proposition, they see there's resources who will be able to go out and do those tasks efficiently and accurately and in a cost-effective manner."

Ajay Sharda: (35:53) So, I think it will be driven more by the need. That's my bottom-line answer, but we'll see, you never know. I want to say this thing is that, a lot of responsibility lies on all of us who are doing any work in this sphere is that we have to put a lot of confidence and trust in the hand of the user. So, we better make the systems robust, self-sufficient, because if that won't be the case, then we won't be able to help them with the right intent.

Morgan Seger: (36:30) Yeah. I can see it's getting to the point though where our need is increased yield. And I think that the flexibility that the robots will give to get into the field more appropriately, because we're not waiting for the entire field to be ready for a big rig and things like that. And then, obviously you guys are starting with in-season applications first, so that's often a value-added application. The first one to go, because it's not always, we don't deem it necessary, but could be a big impact on yield. So, I could see if we are looking at what it can add outside of just the manpower or robot power. I can see us looking at it that way and it definitely could have a lot of benefits.

Ajay Sharda: (37:16) Yes. You have the right point, things which are difficult to manage. And I think I missed that point as well, is that we are targeting the in-season management, of the early season in-season management of the things, and these are the things which are difficult to do, and these are the spaces which are difficult to access. And I think that can really push the benefits in terms of cost saving, site-specific management aspect of it, environmental component is on top of it, but also the valuable information which I get about my area and my crop.

Ajay Sharda: (37:56) I think that is immensely important from many different facets, whether it is crop and produce harvesting times, other decision-making aspect of it. And I think that would be a good starting point, that if we can prove these systems to be viable for those applications, I think people will be very much appreciative and ready for the next steps, from their own.

Morgan Seger: (38:21) Yeah. For sure. Well, I'm excited to watch the next steps. If someone wants to follow along with your work or maybe join in the conversation, because it sounds like you're looking for more people to talk to about this and keep growing the idea. How would you suggest they follow along or reach out with you?

Ajay Sharda: (38:37) Well, I have my Twitter handle, which is at KSU_precision ag. We have our website, which I'm fully anticipating to be up and running in the next month or so, it goes by and FARMS stand for, Farms Using Automation and Robotics and Ag Machine Systems. So, it goes a little bit with what we do on a daily basis. So, follow that part, reach out to us. I'll share some of my contact with you, and website is definitely one of the bigger thing to see where things are. Yes.

Morgan Seger: (39:23) Great. Well, thank you so much and I appreciate your time today.

Ajay Sharda: (39:27) Thank you, Morgan. Thank you so much.

Morgan Seger: (39:29) Well, the enthusiasm that Ajay has on this topic was very evident in our conversation. I love the excitement he has around this topic. And one of the reasons why I was so excited to talk about this on the podcast right now is because, in most episodes we air the question we ask, what is the one technology you are most excited about? And so often, robotics and autonomous vehicles come up. And so, getting a firsthand perspective from someone who's spending a lot of their time right now working on this, I think is going to be really valuable and was really exciting for me.

Morgan Seger: (40:03) I'm really looking forward to following along with his work. You can do the same at, and we will link out to his contact information. And he's also providing some pictures of the machines that they're working with that will be on our show notes at, and we would love for you to join in the conversation there. So, leave a comment about what you think the future of farming and automation looks like.

Morgan Seger: (40:32) We would love to start getting a dialogue, because I really think we have an opportunity right now to build this future together, because the people who are working on it are looking for input, now is a great time to start engaging in this process. As always, we appreciate you tuning into another episode. We would love a rating and review wherever you're finding our podcast today.

Morgan Seger: (40:52) And you can always go to, not only for show notes, but for other expert advice on our blog, you can leave a review of a precision ag product or service that you've used, and read the reviews of others. Let's grow together.

Voiceover: (41:06)

Thanks for tuning in to today's episode, to hear more podcasts like this, please rate, review and subscribe to Precision Points. Visit for show notes from this episode and read expert advice on the blog, share your experience with the precision ag products you use and check out our network of farmer reviews. Let's grow together.


Host: Morgan Seger

Morgan Seger grew up on a small farm in northwest Ohio before studying agriculture at The Ohio State University. She spent 10 years working with ag retail, specifically in ag tech, before coming to to host Precision Points Podcast. She lives and farms in western Ohio, with her husband Ben and their four children. Morgan has her own blog called Heart and Soil where she talks about her experience farming, gardening, and raising her family.


Guest: Dr. Ajay Sharda

Dr. Ajay Sharda is an assistant professor in the Biological and Agricultural Engineering Department at Kansas State University (K-State) and has a USDA-NIFA-funded project under the National Science Foundation’s National Robotics Initiative Program. This is one of the many projects within what Dr. Sharda calls the FARMS (Fusing Automation and Robotics for Agricultural Machine Systems) Lab at K-State. The project aims to design a robotic platform, along with an application system, that can provide substantial artificial intelligence to conduct knowledge-based, real-time on-farm application.

41 views0 comments


Search Loading.gif
bottom of page