From the cab to the cloud: How AGCO Fendt is making smart farming real - We Talk IoT #86

Speaker A: If you use high precision guidance, machine control, including prescription maps and headland management, you already are, uh, in a process where the driver does not have to press a single button after activation. The only thing that is missing for full autonomy is the environment perception and more importantly, the control of the work quality.

Speaker B: Welcome to we talk IoT where we explore the ideas and impact behind AI driven tech of the future and how data creates real business opportunities to stay ahead of the innovation curve. Subscribe to our newsletters on the FNET Silica website. I am your host, Ruth Hejdag. A tractor that drives itself feels treated centimeter by centimeter, data flowing from every implement to every decision. Precision agriculture has been a promise for decades. But in 2024 Acco and Trimble closed a landmark joint venture, PTX Trimble, that puts the technology, the data and the connectivity under one roof. My guest today is Andreas Gatti from AccoFent and we will talk about what PTX Trimble actually changes for farmers, where autonomy stands today and whether smart farming is finally ready to deliver on its promise. Welcome to Wet Iot Andreas. I'm really glad to have you.

Speaker A: Thanks Ruth. Thanks for having me. I'm glad to be guest in your podcast.

Speaker B: Would you like the chance to introduce yourself? What is it you do at Aquafend?

Speaker A: Absolutely. So my name is Andreas, uh, from a background, I'm industrial engineer. In the past I spent or still are, uh, spending a lot of time on my relative's farm. After my studies I worked seven years in the automotive industry and they're already focused on digital ecosystems and are now with Ecco, uh, since three and a half years. So I am belonging to the brand Fendt and I'm responsible for the product marketing of smart farming solutions. So that means it includes actually the driver environment, everything that happens on the cabin of a tractor or a combine with its terminals and buttons and all the software features that live within the terminal. Additionally we have also offboard solutions, uh, where all the data from the machine is flowing to and where the farmer can then analyze his data and take the best decisions based on transparency. Echo is a global leader of ag equipment and manufacturer and has its headquarter in the US and we as a vend are situated uh, in the brand home in Magdoberdorf which is in the Algo region south to Germany.

Speaker B: PTX Trimble, as I understand it, combines the precision agriculture business with Acco's own technology portfolio. What does PTX Trimble change for fend? And what can you now offer farmers that wasn't Possible before?

Speaker A: Well, uh, PTX Trimble Trend Venture marks a uh, structural shift. Combining ECCO's integrated machine tech with Trembler's mixed fleet and aftermarket strength.

Speaker B: Uh-huh.

Speaker A: For fend, this evolves the brand into a full farm ecosystem provider and farmers gain both factory integrated solutions and retrofit options for mixed fleets or older equipment. So this enables the centralized management of multi brand equipment and by that advancing true interoperability. Previously closed brand specific platforms made this really difficult. Overall it creates a more open platform based approach that simplifies workflows and improves usability. And to your question, what has changed for fend? FEND can now offer not only high tech smart farming solutions out of the factory or over the air. Uh, software unlocks on machines that have the capable hardware, but we can also address customers operating other brands away older machines and retrofit them with the latest technology to make them become part of the digital ecosystem.

Speaker B: When you say smart farming. Smart farming now has evolved into a very large topic from GPS guided tractors to AI driven crop monitoring. How do you define smart farming?

Speaker A: So agriculture has evolved from mechanization to precision agriculture to digital ecosystems. Uh, we as FEND have like one big umbrella that is FEND1. FEND1 connects the onboard and the off board systems including the driver environment and serves as I mentioned as the umbrella. Smart farming is structured into four modules within our organization. So it's one that is guidance. Second one is data management, including agronomic and telemetric data. Third is machine control. So basically everything that SDSD implement and uh, last but not least, it's machine to machine and autonomy. So we have uh, as a proven value today, high accuracy guidance that is reducing the overlap by around about 10%. The adoption here is extremely strong as it serves also as the foundation for smart farming. Three out of four tractors leave the factory with a guidance system. In the high horsepower segment, we're even close to 100%.

Speaker B: Wow.

Speaker A: Then section control cuts the fertilizer and chemical use by also 10% really depending on your individual circumstances. And what kind of field boundaries do you have? Is it irregular ones or very rectangular ones? Variable rate control is another example that enables site specific input application and that is increasing yield with the same amount of resource input. Telemetry on the other hand delivers real time insights, improves performance and reduces downtime. And also what's not to be neglected, farmers spend a considerable amount of time on paperwork and compliance documentation. Uh, smart machines on the other hand already capture this data seamlessly in the background and that is Unlocking a major opportunity to simplify, automate and reduce the administrative workload. So as you can see, it's actually an incremental step by step approach. And it's not a toggle that says uh, smart farming on or off.

Speaker B: You already mentioned keywords like mixed fleet and retrofit. So the solution works across different brands, brands and it also works on old machinery. What does that mean in practice for a farmer running a mixed fleet?

Speaker A: Well, in most of the region's farms operate mixed fleets which in the past led to creating data silos across machines, fields and agronomy systems. Uh, FEND actually has long addressed this through standardization and including early ISOBUS adoption for cross brand compatibility. So isobus, in the end just make sure that tractor and implement speak the same language to make it uh, really simple. And with PTX Trimble we were able to scale this into the full mix fleet platform combining the hardware piece, software and data. Uh, so the focus is really on practical interoperability, delivering real farm value. Because as a farmer at the end of the day I want to use my guidance lines, my field boundaries across all my machines, no matter which color they have and I want to have a holistic and consistent documentation. I do not want to run multiple platforms for my farm organization in parallel.

Speaker B: For our listeners who might not be so familiar with the farming topic, what is the biggest benefit for farmers? Is it the uh, autonomous driving machinery or is it the data that uh, is collected by the sensors?

Speaker A: That's actually really difficult to say. What's the biggest benefit? It depends on the role that also happens on a farm. So if I am the farm owner, I'm absolutely interested in efficiency and also in reduced workload in the office. So having really a fast and holistic documentation. If I am the driver of the machine, of course everything that reduces fatigueness uh, on the machine and just reduces the workload and the complexity on the machine helps me doing a better job even on long working days. M the base is always guidance, then it comes to machine control. And of course every hour in the office is an hour missing on the field, really adding value to my farm. So therefore this is also a big topic our customers are dealing with.

Speaker B: And autonomy has been one of the most discussed directions in agricultural machinery. Where does FEND stand today and what is still standing in the way of a fully driverless farm?

Speaker A: As of today we were able to run machines on automation level two.

Speaker B: What does the number two mean?

Speaker A: We know those levels from the car industry and transferring that into an equivalent within the tractor Uh-huh. This is also not an official certification. This is just something that we, uh, evaluate on our own with our automation solutions and say, okay, this is the level we can fulfill. On the other hand, the task I would say is kind of bit simpler because it's just driving from A to B. But here you have a highly complex, unpredictable environment. So there can be a child or a ball getting in front of the vehicle at a high speed at any time.

Speaker B: Uh-huh.

Speaker A: This is rather an unusual use case for a tractor on a field.

Speaker B: True. But you might have birds that are nesting, deer that are hiding their young in the crops. That could be an obstacle, I suppose.

Speaker A: Absolutely. Uh, and that's also a topic we need to deal with. Autonomy is technically possible today for defined use cases. So also, for example, if you consider the PTX Outrun kit, which is a retrofit kit that you put on an existing vend machine, then you have already autonomous grain card operation or tillage. Looking at the equivalent, this would be level four of autonomy. But scaling then really depends on regulation. This is an example where FEND and PTX join forces and drive innovation.

Speaker B: Huh.

Speaker A: The current focus of our X factory solutions is automation and workflow coordination. Really? That support operators, improve efficiency, reduce fatigue for long working days, as I mentioned earlier. And if you're using today's Fendix factory automation, like the tractor, how it comes off the assembly line. If you use high precision guidance, machine control, including prescription maps and headland management, you already, uh, are in a process where the driver does not have to press a single button after activation.

Speaker B: Okay.

Speaker A: The only thing that is missing for full autonomy is the environment perception and more importantly the control of the work quality, because that's a major difference to the car industry as well. Agriculture needs to automate the whole process, not only the tractor, and that includes the implement and the work quality. So I need to be able to evaluate is the implement clogged? Are, uh, parameters like working depth still matching the soil conditions? Do I need to adapt my setting based on that? And this also requires a bunch of sensors, cameras that deliver reliable information also in dusty environments. And I need to have smart algorithms on how to adapt to those various situations. I've just scratched on the surface for tillage. Now I have not even talked about seeding, fertilizing, hoeing, mowing, etc.

Speaker B: It's not only driving. You actually do stuff while you drive a tractor, right? You have to monitor, for example, when you're sowing. You probably need to check if you're still in the right lane, if it's the right distance, if the seeds are being dispersed in the right way, or if you're harvesting, if you're cutting the right way, or if the output actually is usable. Right. That's what you mean by that.

Speaker A: Exactly. And so we do not think like one process which is driving from A to B, but we really have always a holistic process that needs to be automated. And once that's fulfilled and validated, we then go to the next one and start with that. The strategy basically is to start with tasks and processes where you have a low field capacity, where autonomy really adds early value. Because for transfer from field A to B you still need an operator. So if you have a task where an autonomous machine can run for, I don't know, six hours in the field, it's more interesting than if you have a task where the machine is finished in 40 minutes and the operator needs to go back to the field and transfer to the next one.

Speaker B: Understood. When we talked about precision agriculture, this also means you basically would be able to treat each plant of each square meter differently. Rather than applying everything uniformly across a field, you maybe have a patch of your field in one corner that's probably drier than the other one and it is affected by the weather differently. I suppose that's just my layman's interpretation. You have to correct me if I'm completely off here, but um, how close is FEN to making um, that a commercial reality that you can actually treat each plant individually?

Speaker A: This is an interesting question. So as you mentioned, traditional farming actually applied inputs always uniformly, but also causing inefficiencies and environmental impact. And with precision agriculture and uh, the features I've just mentioned, this enables a site specific management based on field variability. So with fan section control you already have a highly granular application. So we have up to 144 zones per boom. That means if you have for example a sprayer, uh, with 36 meter working width that would be 25 centimeters per section that we can adjust. So high precision actually is already technically possible and is also already used for specialty crops, for mechanical weeding, for spot spraying weeds, which is mostly then green on brown, and for clearly structured row crops, for example like maize. But in order to execute this kind of treatment, you also need to have a proper source of input telling a sprayer or you implement on where to apply the plant protection. So this can be either a prescription map that is being generated from drones, from satellite imagery, stuff like that, or it can be based on real time sensor based applications that can replace the prescription maps but require advanced sensors, AI and image processing and lots of processing capacity on the machine huh fence drives to always equip its machines with sufficient processing power also for future and backward compatible solutions. Nevertheless, it's difficult to enable a machine today with a processing power for future image processing because that requires a lot of processing power at the end. In a nutshell, I would say to your question. Plant level precision is technically feasible and it's of increasing importance in certain areas where plants are clearly identifiable and where there is a high value per plant and the working speed is secondary. But on the other hand, it's also clear that today farmers typically use variable rate control, mostly with management zones of 0.1 to 1 hectare, uh, which is at the end balancing data quality, equipment capability and economics.

Speaker B: We will take a short break, stay with us and we will be hearing from our guests very shortly. This podcast has brought brought to you by Afnad Silica, the Engineers of Evolution. Subscribe to our Afnad Silica newsletter or connect with us on LinkedIn if you want to learn more about us. We have put information and links in this episode's show.

Speaker A: Notes.

Speaker B: You already mentioned it's a little bit complicated when, um, analyzing all the data and with the algorithms, but the amount of data that these machines collect must be enormous. How can you consolidate the data capability and m Then of course the next question that comes to mind is um, do we also have to talk about security and privacy or is that not as critical in smart farming?

Speaker A: Modern machines generate data on location, yield, fuel use, application rates, machine health and so on. How Echo is addressing it we use a content based model that means pharma owns their data and they decide if it's shared or not.

Speaker B: Huh.

Speaker A: So customer platforms like Fend One Off Board or PTX FarmEngage provide insights to improve performance and efficiency to enable the farmer in analyzing his own data. Echo, on the other hand, if you're allowed to, we are using the data for diagnostics, for optimization, service and product development under strict security and GDPR compliance. Of course, at the end the farmer benefits with a lot of different aspects. So if he's using all that data and the services, he has less downtime because there's an early detection of issues and proactive service notifications. There will be faster repairs because the dealer already knows the problem and can react immediately. We have a better machine utilization because, uh, of the transparency on working times, idle times and performance. This also leads obviously to lower operating costs and less fuel consumption and input Use, we get more precise operations because better decisions can be made on real field data instead of just basic assumptions. And we have an improved planning reliability. So I know where my machine is and how work is progressing. But at the end you hit a very important point. Trust and data handling is absolutely critical, uh, for digital farming adoption.

Speaker B: Farming is simultaneously a victim of climate change, but also a contributor to it. How does smart farming technology help farmers adapt? And is the business case strong enough to drive adoption without regulatory pressure?

Speaker A: So agriculture is strongly affected by climate change with increasing weather variability, extreme events, and shorter, less predictable working windows. In terms of adaption, smart farming helps the farmers to adapt to these conditions. Through better planning, real time data, and more precise timing of operations, and with a higher efficiency in field capacity, we enable farmers to make better use of short and critical working and harvest windows. In terms of mitigation, precision application reduces the overuse of fertilizer and crop protection products. This then leads to a lower CO2 emission and reduced nitrous oxide emissions. And because of fewer overlaps and optimized field operations, we can also reduce the fuel consumption in the machine hours itself. Last but not least, in terms of mitigation, we have also the improved soil management. So we have machines with central tire inflation systems. That means that the machine is always operating with the best possible tire pressure and therefore, uh, protects the soil of too much soil compaction. And this also has an influence on environmental aspects. Lastly, I would say smart farming enables farmers to do more with less fuel, less fertilizer.

Speaker B: So smart farming can actually also help mitigate climate change and reduce the footprint, right?

Speaker A: Absolutely. I'm convinced that the benefits of smart farming solutions and the business case are strong enough to drive adoption. However, we also see that subsidy programs increase the speed of adoptions. So we have seen that in Italy recently with the Agricultura 4.0 program. And that also had an impact on precision farming, uh, and connected machine take rates.

Speaker B: Yeah, you mentioned the adoption, so I suppose there are also still farmers that might be a little bit more reluctant to the technology shift. I'm not sure if that's a correct assumption, but it usually is. In every field of digitalization, there is a hurdle to overcome and a barrier. The usual suspects, ah, are cost skills and connectivity activity problems. So this usually makes it harder for new technology to be adopted. What have you noticed? Is there something that surprised you?

Speaker A: I would say there's not one central obstacle that we need to overcome. So it's a combination of multiple aspects and it's also partly a Question of generations. So we also see if, uh, the next generation is taking over a farm, there is more tendency towards precision agriculture and more openness towards technology as such. Nevertheless, this is not always the case and there are always also customers who are in their 50s and 60s and still are at the edge of precision farming and up to the latest technology. How do we try to handle that? On the one hand, we try to make the product as simple as possible. So that means we have the same operator environment from the 200 vario, which can be, uh, sold as a vineyard tractor, up to the 1000 series, which is our biggest wheel tractor with more than 500 horsepower. And they all drive exactly the same way. So if you can drive one of them and can use precision technology in a tractor, it's the same for all the other machines. Second one also in terms of product is that we also want to have our machines update capable. That means if we add on technology at a later stage, it still can consume that data, uh, and it protects then also your investment into the machine because it's not from the day we sold it, it's aging, but it really is up to date and up to the latest technology. Another topic clearly is training and support. And so of course first thing is to convince the customer to buy precision technology. But then also we have complex environments, different data connections to various platforms. And so whenever the dealer cannot support anymore at a difficult level, we also have smart farming experts across our regions who then can support customers and dealers to overcome those challenges. Then we're also dependent on circumstances where we have not the biggest influence on connectivity. Gaps are one of them. So mobile network providers clearly focus on, uh, urban areas and where most people are living. Our machines mainly operate in rural areas where less people are living. And so therefore we always face also network gaps. But here we try to overcome that with, uh, wi, fi or satellite, uh, connectivity to enable the farmer using it in each and every part of the world. And there's also a topic that is, uh, mentioned from all our sales colleagues across the globe, which is skilled labor shortage. So with that technology being implemented into a farm, you also need to have the drivers that can handle it and can operate it.

Speaker B: Um, that's a very different skill set than it used, uh, to be, um, a couple of decades ago, I suppose.

Speaker A: Absolutely. I've once been on a congress, a speaker said the nuts have adopted, now the mainstream is coming. Very first moment I had to smile about that in the way he put it. But at the end I think he is right and that also shows the importance of plug and play solutions because like early adopters they accept if something does not work at the uh, very first time, but if more and more customers are using it, you really need to have sophisticated solutions and they just need to work easily and ad ah hoc. And therefore we have a uh, huge validation department that is always testing our uh, products from a customer's perspective. And unless they say it's meeting the FEND quality standards, we do not release any of those products.

Speaker B: Is there anything in the development in the area of smart farming that you are most excited about, something that's on the horizon that you are watching closely?

Speaker A: It's difficult to decide for one of them.

Speaker B: Oh, okay, you can tell me all of them.

Speaker A: I mean autonomy clearly is one topic that excites me really. Uh, also seeing the first big steps towards that and seeing machines operating autonomously in the field. Second one is interoperability as this is one of the biggest customer pain points as of today. Really bring from livestock to machines and everything into one platform and also enable you to draw decisions based on a holistic ecosystem and not just on an individual machine because each decision also has an influence on the, each and other direction and I would say also AI driven decisions. So we want to give our customers a agent that supports him in the decision making. As we are already used now to AI and put in each and every question and ask for support. I can imagine that something like this is also going to be really helpful tool for our farmers.

Speaker B: Yeah, I have been using uh, um, artificial intelligence to ask how to plant my high rise bed and if tomatoes will be happy in it or not.

Speaker A: I hope it did work out.

Speaker B: So far it is working. I'm skeptical. We will see. It's an experiment. If I don't eat myself grown tomatoes, that's okay. But if a big farm is losing harvest, that's obviously something completely different. If you had to put together a soundtrack for this episode, what song would you put on it?

Speaker A: I have two in mind. One is harder, better, faster, stronger for Daft Punk because that's kind of what the technology is about. It's always about efficiency, productivity, technological progress.

Speaker B: Very cool.

Speaker A: On the other hand, from a mindset perspective, I rather would see a sky full of stars from Coldplay because that is like representing a positive forward looking mindset, uh, not being afraid of technology. And that's what it is all about. I think it offers us new possibilities. Technology will not take away the decision making authority from farmers and so therefore have a positive mindset and be open for the technology that's coming because it will benefit us in the end.

Speaker B: How lovely. Um, I will add both of these songs on our playlist and I love the explanation for the songs. Thank you so much for contributing, uh, to the playlist. Great addition.

Speaker A: Thanks a lot.

Speaker B: Thank you so much, Andreas, for a clear eyed look at where smart farming stands today. It was terrific to have you on the show and I'll be excited to to see where it will all go. I hope to welcome you to another episode soon.

Speaker A: Thank you. Was a pleasure.

Speaker B: Thank you for listening to We Talk IoT. Stay curious and keep innovating. This was Avnath Silica's We Talk IoT. If you enjoyed this episode, please subscribe and leave a rating. Talk to you soon.