I've got a brand new combine harvester and I'll give you the API key

Yes, you can stick IoT sensors on tractors - and John Deere's been doing it for 20 years

Two decades ago a tractor manufacturer was fitting smart sensors to its devices. Blending the agricultural and the technological were a natural fit, John Deere’s Georg Larscheid told The Register.

Agricultural IoT is one of the more obscure areas that has largely passed the rest of the industry by, chiefly because it sticks to what it’s good at and delivers what it says. Larscheid, John Deere’s integrated solution implementation manager, said: “It started very early, with the advent of precision farming in the mid- to late-1990s.”

Although John Deere was keen to stick sensors all over its tractors and other agricultural machinery, Larscheid admits that in the beginning “the technological foundation was simply not there to make it all work… the infrastructure was not there.”

“Even though people had the idea and thoughts around it, it could just not be implemented beyond the university lab installation. You couldn’t do this in the mainstream - the connectivity wasn’t there and the big data analysis tools weren’t there.”

So why bother, then? Larscheid says it was about “local resource management” and improving efficiency, “going from a whole field treatment to local resource, breaking up the field into smaller parts. You use data, yields data, soil data, nutrient data, for instance, to respond to the local conditions.”

“Now we have all the technology on the machines. In a broader scale, connecting those machines, centralising the data processing, that is definitely an area we believe is fundamental when we think about growing productivity in agriculture,” he says.

Today’s world of smart tractors is not dissimilar from the challenges every new tech innovator will recognise. “Once you’re sending out tech, it’s going to take a while before it achieves a relevant install base,” nods Larscheid. “That’s what we have achieved over the last 20 years.”

Moving from the idea of farmers sticking USB drives into tractors “where the farmer doesn’t have the capability to leverage that” and instead having the data transmitted to a central processing area where it can be actioned; in Larscheid’s words, “connecting those machines” to create “value-added services”.

The data rates John Deere’s smart tractors generate will be familiar to any IoT enthusiast: “within the 10s and 100s of bytes.” Yet they scale up as well. Larscheid says: “Sometimes we have real-time applications, where you need to have an online connectivity. For instance, if you doing load management on a machine, and the machine has a symptom of a failure then the machine needs to immediately implement [the solution].”

When high density data is involved, such as analysing soil nutrient content, sampling rates can be as high as 4KHz, leading to “megabyte-sized files in a hurry.”

Although standardisation is a mature topic in the mechanical road and agricultural vehicles sector, Larscheid admits that John Deere does have some “proprietary” standards for its machine sensor tech: “This is an area where standards do not exist. Usually tech innovation trumps standards.”

He continues: “If you want to be an innovative company you cannot wait for a standard to be ready. At John Deere we are committed to the precision space and therefore we are forced to implement technologies not based on standards. On the other hand, we are fully committed to supporting the development of them.”

In terms of standards it has adopted, John Deere works to ISO 11783, the standard for smart tractors and similar agricultural machines which incorporates a spec for a “universal virtual terminal” for tractors and their attachments. “We’re not just having a self-propelled unit, we have implement manufacturers who are not necessarily John Deere, and their data needs to be transmitted,” explains Larscheid. John Deere is also a member of AgGateway, a consortium dedicated to “eBusiness in agriculture.”

As for the future? “Smart farming, Farming 4.0, IoT apps for farming... these are all buzzwords that describe the same thing. We’re trying to bring automation to machines. We’re trying to bring better data based decision making, decision support to farmers, because we know that’s where a lot of the inefficiencies occur,” said Larscheid. “We cannot just rely on making machines bigger, faster, stronger… we’re getting to a point where productivity improvement has to come from other sources. We’ve maxed out on the ‘make it bigger’ dimension!”

As a worked example of how sticking sensors on Things makes them more effective at what they do, the smart tractor is a solid, if unromantic, one. Yet in farmers’ eternal battle to produce ever more food within ever-tighter margins, it may be - as Larscheid notes - the last available path to improvement. ®

Sponsored: Minds Mastering Machines - Call for papers now open


Biting the hand that feeds IT © 1998–2018