Read time: 4 minutes
WE ASK THE EXPERT
Dennis Buckmaster is a professor at Purdue University and an agricultural engineer by training. Over the past decade, he has focused on digital agriculture, data systems and AI.
DR. DENNIS BUCKMASTER
Professor of Agricultural and Biological Engineering and Dean's Fellow for Digital Agriculture, Purdue University, USA
Q: How is AI changing how farmers interact with their equipment?
I would categorize machinery-related AI into two classes. First, image-processing algorithms embedded on the machinery, for example, to differentiate weeds from plants. Second, generative AI, which is largely language-based and involves human interaction. Both come into play.
AI makes complicated operations smoother: controlling settings, machine paths and things that used to require a skilled operator. And generative AI can facilitate machine interaction. Maybe the farmer is looking at the dashboard wondering where he or she should adjust the air pressure in their planter. There is a menu path, but generative AI could help make that adjustment.
Q: How is AI already adding value?
A: Automatic adjustments are remarkable, whether for planting, spraying or harvesting with less damage to crops or wasted products. Another benefit is coordinating machinery. AI reduces the need for operator skill — or at least reduces operator stress. They don’t need racing-driver intensity every moment. Maybe they can take a sip of coffee or have phone conversations while they work!
Q: What are the biggest misconceptions about AI and farming?
A: I see two extremes. Some say: “It’s not good enough to trust,” but it is, in many well-vetted cases. The other perspective is: “It’s perfect. I’ll just trust it” — and no, you need to verify. I often say, in these early stages, AI is like an adviser earning your trust. If I meet someone with great credentials, I still make sure they’re doing the right thing. Treat AI the same way.
AI is like an adviser earning your trust. If I meet someone with great credentials, I still make sure they’re doing the right thing. Treat AI the same way
Q: How significant is the data quality challenge?
A: Interoperability has been my word of the year for several years now. We still don’t connect data on production, markets, machinery, facilities and personnel. AI could help with that. Farmers make strategic decisions — what to plant, what variety, when, where — every year. They need context to make these decisions better, but simple things, like what was planted on a given day, are often not readily accessible.
Most decisions are in the moment, not strategic. That’s where data streams come in. Where is my machinery? How full are the tanks? What’s the best route to keep things running? We need data to provide live status for smarter work in the field and around the barnyard.
Q: What kinds of sensors are creating opportunities for AI-driven decisions?
A: Internet of Things sensors give us data on soil, machinery, stationary equipment, crop status, inventory. All that complements what we get from machinery alone. Weather forecasts are important, too. A farmer may think: “That field needs water. Do I irrigate or wait for rain?” AI can weigh the likelihoods more objectively than people, but humans should always have the option to make the final decision.
Imaging technology will open insights into the state of soil, crops and grain bins. We can identify stresses and what’s causing them, leading to quicker treatment. We’re even using imagery in the barnyard, where equipment studies the feeding, breeding and grouping of cows.
Q: How can AI work in areas without strong internet coverage?
A: You can’t shut down just because you’re offline, so you’ve got to have some edge computing capacity — data storage and computer power on or near the farm — or reliable ways to transmit data. LoRaWAN (long-range wireless) can carry small bits of data several miles, so it’s good for tracking machinery status or relaying sensor data. There’s a new Wi-Fi standard called HaLow, which is license-exempt in the US. It’s lower bandwidth than home Wi-Fi, but could be great as a private on-farm network. TV white space and low-earth satellite could provide connectivity, too, with approval for use in mobile equipment. Agricultural machines stay in a defined area, such as a cluster of fields, so they don’t need huge areas of connectivity.
Q: Are there use cases that need higher bandwidth?
A: Absolutely. A drone can be used to spot anomalies, but unless you’ve got powerful edge compute, you’ve got to upload the imagery for processing — and those will be large amounts of data. That’s why agriculture needs symmetrical connectivity: upload speeds and capacity that match those available for downloads.
Q: How close are we to large-scale deployment of autonomous systems?
A: Technically, it’s close. But widespread adoption is still years away. You’ve got to get the equipment to the field and support it with fuel and supplies. Plus, there’s a huge inventory of existing equipment. A five-year-old tractor that’s not autonomy-ready doesn’t need replacing. Some machines can be upgraded to be autonomous if they have the right computer systems, but it’ll take time.
Q: Are there cultural or social barriers to autonomous farming?
A: Yes. Many farmers just enjoy the physical aspects of farming. They don’t necessarily like the office work. There’s a meme I like: “I want AI to do my laundry and dishes so I can do art and writing, not for AI to do my art and writing so that I can do laundry and dishes.” The agriculture version is: “I want AI to do the business and taxes so I can drive the tractor, not drive the tractor so I can do the business and taxes.”
Q: Where might AI surprise us over the next few years?
A: Generative AI can be an always-available adviser. Some farmers don’t want to deal with data, and may prefer to say: “Here’s my data. What does it mean?” AI could answer that without them having to write code, allowing them to analyze data in ways that are otherwise out of reach.
Q: Any final advice for decision-makers trying to evaluate AI?
A: Stay optimistic. Good AI needs good data and getting good data feels slow and unrewarding at first. But over time, it builds and you’ll hit a leapfrog moment. Farmers have been collecting yield data since the 1990s and now, with AI, that data becomes useful. We’re getting there.
Internet of Things sensors give us data on soil, machinery, stationary equipment, crop status, inventory. All that gives a different angle than machinery alone. Weather forecasts are important, too.
CNH is a world-class equipment, technology and services company that sustainably advances the noble work of agriculture and construction workers.
CNH Industrial N.V.
Corporate Office: Cranes Farm Road, Basildon, Essex, SS14 3AD
United Kingdom