Are We Even Building the Right Robot?

Nathan Faleide: Well, hello all. I'm doing my podcast from my pickup as I've been trying to redo my normal good old ag censored podcast with surely Nathan. And let's just say things are still busy. So I don't know how often I'll actually get these things out. I'm working on it. I haven't really even told people the first newer one I got out. put it up, but here we are. Anyway, something that's wanted to kind of bring up this week is it's, I don't know, been bothering me. I guess I see a lot different posts and stuff around it. I mean, there's always something about. robotics and, you know, how... this is the problem or this is the problem or that's the problem and this is why this is what they need. And I chimed in on them because of course I have been trying to think through it through a kind of logical farm standpoint, something like that. And I mean, in the end, I know what the real probably way for it to really work is, though the reality is that way we are very long ways to get there. Plus, if we actually got there, how that works in agriculture would be like the least of our problems. So I'll give you my kind of answer to start. Maybe that's the way. then, you know, maybe how we can get there or how, I don't know how it progresses. And something like that. At least I'll try. So in the end, the only way for, you know, real robotics to probably work in agriculture at the scales that people want and, you know, all the automation and all that. And this is, yeah, call it a weird futuristic negative way to think about it too, maybe. It'll just be like with bipedal, you know, humanoid robots. They're probably really not that far away from getting there as many think. I don't know what that timeframe really is. I have a sort of date and time in my head and how it'll happen. It basically will be. around 2040 timeframe. And it be earlier, honestly, you know, things are weird how they progress these days. It'll be someone like John Deere basically making a deal with whomever is building these things at large scale for other human needs. You know, they'll slap a... technical John Deere hat on them or something that has a bunch of sensors in it. Who knows? Yeah, that'll be the John Deere thing will be like the special eyes that are embedded into it for agriculture purposes. So it'll be some sort of partnership. And they'll be leased out. You can't buy them. They'll be leased out or something rented, I suppose. Because they can do all these different tasks. It'd be by the hour just like a normal human being would be. except that you want a robot and they'll automatically integrate into anything. be old, you work with old equipment or with new equipment and you have to plug it in. Now it just, you know, transfers whatever information needs to the way it does through glories of whatever the internet means to us by then. And now AI has taken over everything because it'll have to run on that naturally. And it'll be your farm helper. deer and other OE, I mean, it will work in any, anything. So basically you, you own the market, whoever does it first and well, and there's really only one that can do it globally anyways. So that's what will happen. Make some big deal with whomever makes that robot. Is it Boston Robotics? Is it Elon Musk? it Jeff Bezos or who knows? It's going to be one of those mega billionaires. Cause how else do you even have the back end money to do it the way that you'd really want to do it. You got to make all these AI data processing centers to even get to the point where this could actually even work unless you create, you know, shrink that AI data processing center to, you know, size of a brain. It's really what I suppose they're slowly trying to do just as much as, you know, computers used to be giant, filled an entire room and they have the power of like a modern day calculator. And now what is that? It shrunk down to an apple. So I could be totally wrong. Who knows? Of course, if we get to that point and have that bipedal humanoid robot, well then, like, what's the point of doing anything? That's when they take over the world and, you know, we're forced to basically, you know, probably through some other means, not be able to repopulate the earth anyways. It's the it's the the giant and powerful way is to save the environment is by reducing the need for crap ton of humans with something that they can control. I mean, if you really get dark, I mean, that's that's that's the answer. It's a pretty easy answer. You know, if you don't comply, well, what does that I mean, look at the movie I robot. I mean, that probably something like that. In some form or version, which that's always been generally, I guess, known, it's pretty easy to like, think through it. It's like, man, if that happened, yeah, that would probably happen. crap, that could happen, but might not happen because of that happening. What is too many scenarios? Anyways, that's just where I'm starting. That's where it all gets figured out. Of course, then there's all sorts of other stuff with that. In the meantime, though, how does how does actual robotic things, you know, actually scale and work. And yeah, I see robots in the field all the time. I'm right next to the grand farm where there's some demonstration ones there and talk to the owners of these things and good, nice people trying to build something to help someone like me or others, you know, use less chemical or make their job easier. Like the essence of it is totally fine. I have no problems with it. The issues I more have is like the realistic way this all works. And it's not to put any of those hardworking people down building these things. It's can they succeed the way that's. hopeful, actually. You know, as I've talked to them and others, you know, there's variant versions that, you know, fit a little better. Some, you know, it's not that they're not good technology or they can't do something. It's that for majority, it's just not like practical compared to what you can do today with just older normal equipment. It's a cool factor. It'd be fun if it just kind of worked the way you think it would work, but will it? Like even I'm, whether you can see or not, I'm driving next to a field of corn here. And I think about some of the different, you know, robots that go down the row individually, and some of are cool. They can do the job. They've proven they go down, you know, mow the weeds or have a... arm take out a weed or something like that. Sense something, do something. And that's, that's great. Cool. Demonstrated that. But then I think, you know, sometimes corn falls and stocks are big. You know, a bunch of them can fall down wind. Maybe it's just, I don't know, a random deer walks through, push down some stocks. And then it gets, it's just stuck in the middle of a cornfield. Now maybe corn is not the most obvious answer to these things because of that exact problem. But I've never actually ever heard anyone talk about that. Or the random rock. You know, because... It's one thing in soybeans, people, least up here, will use a big steel roller and knock the rocks down on the ground. Still something, they pop up, or... You know, I went through the field and I had to spray it. Maybe I still have a robot, but... I sprayed it, created a big rut. You can get stuck in that with a giant machine, let alone a little robot. then yeah, stocks falling, or it could be, I don't know, a tree branch on the edge from a shelter belt that fell down from a storm or something. And so for the robot to work, you'd have to mitigate that. Okay, well either a human goes out to mitigate it, which you know, kind of defeats the purpose. Or the robot is attached with arms that can move it out of the way or cut it out of the way. Which, totally possible. It's just, that's just one small scenario. And, you know, there's, it's just. It's cool that it can do some of these things in certain situations, maybe in smaller acreage, specialty stuff. Even I saw a video of someone on Facebook explaining their, I think they had vineyards, and they got one of those Monarch tractors. And I mean, there's a reason why it failed because literally the thing didn't work. Like, it had nice features, like, theoretically are built to have good things that were practical. Even the guy explained he's like, like, he's like, can't even like go slow on it. It wants to just like jump around and go fast. And then it's like, it's like the three point didn't work very well. And, and half the time it didn't really run properly. And just like all these little things is like, yeah, okay. And that happens. with new stuff, new equipment. Heck, the largest brands in the world have problems with their newest stuff. You know, when the, for instance, when the S series combine came up from John Deere, they changed the exhaust to come right out the back. What did it do? It caused fires all the time. Bunch of combines went on fire and crops, you know, fields led on fire because of it. You know, they fixed it, but yeah, there's growing pains. There always is, and that's okay. So I'm not putting down even them for that or anyone else. It's more so the cost of those growing pains. you would think was someone like Monarch or someone that's raised hundreds of millions of dollars. that's enough money to solve those problems and, you know, showcase and scale the tech to the way it really needs to be. And the answer is actually it's not, it's not enough money. Like, it's just isn't. Like what I want to see, any farmer, I'll just put the normal farmer cap on. is saying this, feel the sugar beets right next to me. it's corn and your soybeans right next to here. You know, for that to work, I'd practically want to go to a farm show or someone's farm and see, you know, 10, 20 of them working together, doing the job. Supposed to take out a weed or whatever it is. I've never seen it. I've never seen anyone do it. And the reason is, it's not saying it hasn't been done. that, you know, I haven't seen it. I'm around these things. I go to shows, I'm around these tech things and I don't see it because it's too expensive to showcase it that way. When each of those might cost, I don't know, 50 grand, either 25 grand, or 100 grand, well, to showcase it properly, okay, you need a million, two million dollars in the field and you're then moving them around to different places to showcase them. That's really expensive. Now, in essence, the large OEMs do this all the time at farm shows and whatnot, but. You know, they have these giant networks. They've built these enormous moats of dealer networks and people and parts and service. And that's how they, that's how they can do it. They don't have to travel, you know, all of their inventory everywhere. That's what robotics groups would have to do right now is basically take a good amount of their inventory. Even if they have that inventory. A lot of these things are built on spec because that's where they're at. And even if they have some inventory moving some of that stuff around, it's expensive, it's hard, things break, you need a whole team of people to probably just make sure you're a good chunk of your assets don't fail. And yeah, things do. Things break. I've seen a smaller demonstration of these things happening. Yeah, something broke. A bearing or a wire got loose, whatever. Okay, well when you had 10 or 20 of them in a field, who's doing that? And I've talked about this, who does the service? You still need a person. You might need a team of people because these things are specialized. So that's where they got into, we're gonna go into service. We're not gonna sell them. We're going to do the service. We'll have our hundred and we'll break them out in different parts and we'll have a team of 10 people going out there, whatever. It's kind of like spray drone groups. It's kind of how it started to work. Initially, yeah, not great. It's still imperfect, but now you have these smaller groups that have a few spray drones and they go out and do their work. And those things can go faster and cover much more ground, even though they can't compared to larger machines, but they still can relative to any kind of robot going up and down a row by a pretty decent factor. And those haven't even scaled to the point of where it can go. How are the other smaller robots in the field gonna be able to do it? I mean, literally on the drone side, you have fricking China, DJI building an insane amount of these things. They have billions of dollars and they can bring them out and they're easy to work with, startup fly parts and all that. And still that's really hard to do. You have to look at like the other economics of this and the reality. And are we, are we building the right robotic thing? You know, it's a robot. Yeah, that's what's in our head, but is it the right robotic process that someone like myself really wants? I mean, it's cool, but no, I, you know, as I've, I think I stated in some other posts or some video. For me this year, yeah, I would have liked maybe a robot because they had all these random weed plants. Just go around randomly and take them out. Of course, that's still expensive for that service and or piece of equipment when I can just go, you know, spray the whole field with Roundup at $5 an acre. You know, call it $7 with fuel sprayer or something like that and wear and tear. versus service, it'd probably be 50 bucks, 20 bucks, I don't even know. Or buy the robot that's $75,000. My sprayer cost me $15,000. And many people have cheap sprayers, whether it's pulled behind or could be $100,000 sprayer. Well, I use it for all these other things and your thing can work to kill an individual weed. Unless there's 50 of them out there, it's hard to do. Well, you can't showcase 50 of them, because you can't afford to do it. niche areas crops. Yeah, sure. But still, that's really hard if the drones can't really do it really well. Like, are we building the right robot? Are we thinking it properly? We can't get the bipedal humanoid robot yet. We're not gonna, we're not gonna get there for quite some time. And when we get there, that's all that's a whole nother thing. That's something that's becomes totally out of basically all of our control. If it's even done. properly. So then what's the answer? Well, I put a post out this morning, which is Monday. I don't know if I'll get traction. Who knows? Whatever. There's a group called, what is it? RC Robotic Arm or something like that. They're out of Canada. RC Farmarm. And they basically took this module, they built it with actuators and things to... basically put on top of a normal tractor console to operate, you know, hydraulics and the PTO and the throttle. Pretty simple. I'm honestly, I mean, you have to form fit it and you know, each model of equipment's different. So there's that they did one model and they went to other models and kind of just informed it better to fit on there. Stuff like that, you know, made it more ergonomic. Still think they're Yeah, a little pricey, but they're really not that bad considering what they are. I think anywhere from five to 20 grand, depending on kind of the complexity of the piece of equipment. Well, I mean, adding GPS, you know, modern stuff like that, that's, you know, that's same cost. So cost wise, like whatever, if it does that job properly, it's pretty good. Well, they just came out with their robotic version of it. If they can do the controls through not having to integrate into the machine because that gets complicated and you get the license issues, get the problems with the OEMs on the specifics of that. That's a problem, whether you like it or not. That's a problem. Well, they just bypass it. They're like, we'll just physically do it. know, the machine's meant to do it that way. So let's put the arm on there. Well, we did it on a 20-year-old tractor. The one that I... could afford when I was looking at it, 9,400 I think it was, John Deere four wheel drive. And then they integrated that in, you know, physically with the current GPS auto steer that was already in it. Like I have green fit auto steer in my 9270K Steiger. Could this group make some, a panel for me to control the four levers I have in the one throttle? and like the two buttons that I need to operate the thing. easily. It would probably be easier than some of the others because they're big levers. They're not like small little plastic knobs. So it would probably be actually easier if you'd different ways. And you could basically turn my 30 year old tractor into a robotic tractor. Not a 9RX that costs a million dollars, brand new, that, you know, they barely have any of those out and they've had their own issues with that. You can do it on an old thing. Yeah, that makes a little more sense. because you it also can work on a new thing. The same concept works. The integration's easier. It's more of a physical parameter and smaller thing to build than this big robotic thing that's integrated and has all these crazy sensors. Well, of course the problem is, you still need some other sensors, because if you're not in the cab, you know, there's still things breaking. Like, hydraulicos can blow, a bearing can go, you can hear it. You know, you can feel a weird vibration. It happens constantly, even on newer equipment. And there's not enough sensors even on newer stuff to detect all this stuff. You know, if there's anything that has all this weird sound that combined, of course, you're not going to automate a combine anytime soon for thousand different reasons. We can say, I'll add more sensors to understand it more and all that. Well, if we're going to do robotics, you know, things can break. Well, why do I not see any aftermarket? Like, OK, like this RC arm thing. OK, that's one component of solving the problem. Well, where's the acoustic sensor that senses odd things? Where is the visual sensor that looks for odd breaks? Not not performance, like literally its job is to look for hydraulic leaks. bent shanks, things that break. looks for the rock in front, looks for the wet hole, looks for the obstruction, whatever that might be. Or, you know, hears the issue or feels it through vibration. Like those are needed to solve these bigger, these other robotic problems. We start with the machine and then we envision it as, we'll do a totally different machine because this is what kind of a robot looks like. Well, as I'm trying to get people to do and some are going to do it, like, pimp my equipment. Make my thing, I guess, robotic so that I have the confidence so I can let it go in the field. Like I don't. That's maybe a more practical approach to get people to the point of where they can actually use it and making these littler processes and them fit together. That's actually way easier and cheaper to showcase and to scale than any field individual robot by probably a pretty big factor when you really go through the whole process and economics and everything. Of course, all those things gotta work together. I but this is how the GPS and auto steer industry worked. This is exactly how it happened. They had these different components, they put them together, they came from a different industry, they made it work for Ag, they did these actuators, they adjusted it for how an Ag machine works, and there's all these different players, and all of a they started getting accumulated. Some still exist privately like Ag leader, and then others took it on OEMs like deer and. case, I've gone through various versions of it, and Raven, and Trimble, and, well, you there's all sorts of others. A lot came in from construction industry, or some came from military type stuff. I guess, why aren't doing it that way for the robotics? Seems like a pretty natural progression. Look, we're autostars. those, that makes a crap ton of money. It's scalable and it's usable. It's resellable. Like if I got this arm for my 9270 and decided, I'm going to sell it and get a different tractor. Guess what? I can still sell that thing. If there was a specific acoustic. sensor just for my tractor of multiple sensors that are just listening for sounds. Well, I could put that on a different machine. I don't need a whole different new machine with specific things. I can just move it around. Like that's, that's how it works. Why aren't we building that? Now some are kind of like my friend Craig at FarmWave. He's trying to do that on the harvest side loss and all that. And he's toyed with other things. I've told them, hey, we should try this and this and this and to this point, it's like, yeah, that would be great. It's all awesome. I'd love to do it. Of course, the economics of building those things are also expensive. They take time, a lot of R &D. That RC arm thing, it's been out for 10 years, but you know, now they just got to the semi-robotic side of things. They take time and money, because that's what investment's for. Of course. takes time to make money from those investments and the whole industry has been kind of a chaotic mess of that. So, yeah, that's still needed and that's still hard and it's so complicated. And there's people trying, but it's like we're disillusioned with the noise of what we think robotics means for agriculture. It's basically my point. We think it's this when it's really this. But really what will probably happen is this, the, you we see the robots in the field, these small little robots. That's our perception of what we think it should be. The reality of things that work are these other components like I talked, the RC arm, farm wave, you add other pieces of that, things that haven't been developed yet. Then of course, reality will be, it'll just be, you know, a humanoid version of myself that'll be the actual solution. You should almost just like stop even caring and let them figure it out down the road and just let's farm. Let's, worry about other problems that we have. You know, if we collectively just focused more on policy, even if it's ugly and weird, yeah, we'd probably be better as an industry, but here we are. So that's my thoughts. around all that. And I do want my stuff to work in maybe a more automated way. I sometimes don't, because then you have the emotional side. I like to drive a tractor. It's kind of fun. like to drive a combine. Now getting it all set up properly with all the right parts, making sure it works and maintaining it and all that, that can suck. That's not the fun part. I want more help there, but I there really isn't a practical way besides literally a humanoid robot doing the work for me, which we're not there yet. So, you know, still got to do it. And I gotta my dogs here quick, But I have a closing thought. Hold on. Big one down there. He said, He's had some rough legs and got a little sick. He's a little slow. Working pretty hard. He's gonna lap all the rest of the day. Anyways, You know, with all that said, robots, robotic processes, automation, it's all cool, but still dealing with like farming in the way it is, which is a weird industry. And there's no, it's a lot of commonality in it. So how do you really build automated processes for things that aren't, you know, consistently the same? So then it makes the questions, should we even bother with it at all? You know, there's plenty of people in the world to do these other jobs. First, can we do them economically? you know, but I like to have a hired hand. Yeah. When my dad's been out there for free, it's been great. You can't do everything and all that. But my farm's kind of too small for me to hire somebody as it is for a lot of people. That's why kids and family get into it. basically free slave labor. That's literally why people had a lot of kids. back in the day, they needed help. It's just, a physical job. It's a lot of pieces to it. That's why people hire migrant workers. So there's a bunch of South Africans in the state here. There's not enough local people to actually do the help. So they gotta bring them in, pay them and too much for complicated stuff. Like we haven't... solve that problem really well and that we're bringing in robotics saying well we're gonna put Billions of dollars towards robots to solve the problems of human labor is like I don't know maybe what billions of dollars into solving human labor by Working with humans and solving with policy because that's what we really need. There's plenty of those people in the world To come and help and it could be better. Of course. We need the right money to do it. I need to change policy to make agricultural goods and process a little more profitable than they are. Different rules and things so that, you know, we get a little more back so that we can support these other things. Yeah, we don't do that. We worry about the future too much and what it could be because, you know, technology can replace it. mean, humans are pretty resourceful. Maybe you should solve human problems first instead of like robotic problems. Just a thought. I don't know, seems pretty novel and moral to do that. Don't you think? If we collectively put the money in that, we would with a lot of this other tech, probably be little happier. Things would be more efficient. More people would be striving probably because more people are making more money. Even if it's lesser, it's probably more than that they were, which is usually the point. That's kind of what capitalism works. If you want a capitalistic society. Ugh. See now I'm going really deep into the real problems of agriculture and the technical ones. But, I mean, technical things also can help solve some of those problems. Ugh. It's just... Kind of overly stupid. Some days. as it's so easy to get caught into the, I don't know, fanaticism, is that a word, of what could be versus looking at where things are now and helping fix the practical things that we all complain about. But we don't invest in those things. We don't invest in our current problems. It's like we almost invest more in our past problems, because those are not fixed, which kind of lead to our current problems. But then yet, we also put a bunch of money into our future problems without solving our current or really past problems while... Huh. Yeah. probably something to be built there, I guess, right? Something to be done. That's a whole nother episode or multi-series episode, maybe. But in closing, yes, all these robotic things are cool. One hand, they seem very practical. And on the other hand, they seem like a giant waste of time and money. And there's all this other crap that needs to be solved. And it's like, you're just running in a circle. And what most farmers do is say, go, I'll deal with what I got because that's what I can do. And I, I don't have the time to deal with all this other crap. Now I'm going to try to help you building those things, showcase some of that, maybe try out some reality things. Of course, I might not have time for it. It's more of an idea that I'd want to do my farm. You invite others in to showcase their stuff or just. play, play around like in a sandbox, which is kind of more of the idea. It's not really a showcase, it's more about come here and play and I'll talk to you. I'll be figured out to a point I can, but otherwise, here's a place to be. That's it. or they're all leaving at that. I don't know. We'll see. Will that be helpful? No idea. When the real core issues are, as I laid out, a little bigger than just solving that individual plant in the field. There's a lot of plants, just like there's a lot of other people. You know, sometimes you can only solve so much at scale. And, Always and always, that's the economics of scale. They don't lie. They're real. There's some food for thought and I'll... That's it for today. We'll check you later.