Machinery Link

Blue Isn't Always Blue

Julian Sanchez, second from right explains Deere's "See and Spray" at the Consumer Electronics Show. (Progressive Farmer image by Dan Miller)

Julian Sanchez first came to John Deere in 2004 as a summer intern. He returned in 2011 and is now the director of precision agriculture and business development. He is responsible for leading strategy, business development and portfolio management. Throughout his career at Deere, Sanchez has made significant contributions in the area of digital innovation, such as building capabilities in mobile software development.

Sanchez earned his Ph.D. in human factors psychology from Georgia Tech. He holds more than 20 patents and is a John Deere Fellow in Technology Innovation.

He sat for this interview during the Develop With Deere conference convened earlier this year.

PF: In 2020, what's a connected tractor?

Sanchez: Simply, [that tractor] has the ability to
record and store data about whatever is being done
in that tractor. And, that data can be moved off the tractor wirelessly.

PF: Is that even the right question in 2020?

P[] D[x] M[x] OOP[F] ADUNIT[] T[]

Sanchez: It was the right question. But, at this point, we feel pretty good that we have nailed down what a connected tractor is and all the technologies to make that happen. The more relevant question is, what is a more intelligent tractor? That is truly what we and others are grappling with.

Trying to define human intelligence is a field all of its own. What makes humans special and what defines intelligence in machinery is how good are the machines' ability to process context.

You can say to the computer, "This is the color blue, blue is made up of these signals. Look for these signals, and if you detect these signals that are blue and therefore, do this." It's really not intelligence. It's a very predetermined pathway. Intelligence is the ability for the computer to figure out something is blue while there is a weird light cast onto it, or it's nighttime, but it is still blue. That is context, and that is the ability to discern context. Blue is blue, but the context around that blue is changing. It might kind of look like blue, but the machine has to be intelligent enough to see the things [that might not make it look exactly blue].

That's why [artificial intelligence] is a breakthrough. Ten years ago, as a programmer, you would have to come up with every scenario the machine was going to be in, so it could detect blue or detect green. Now with AI, you show the machine examples of what it might encounter, and the machine forms an opinion or "interpolates" all the possibilities.

PF: So, is there a way to measure intelligence or intelligent capabilities on machinery?

Sanchez: You know how companies competed on horsepower, for example. In the next 10 years, there will be some way for us to say ours is more intelligent than theirs. How good are we at detecting things, for example.

PF: Tell us about Deere's decision to attend the Consumer Electronics Show in January for the second straight year.

Sanchez: Our decision to go back [after 2019] was more complex than our decision to go the first time. We actually doubled down on our original intention. We engaged the tech community about agriculture and by doing so, established our brand as a tech company.

PF: What does it mean to be a tech company?

Sanchez: Deere has always been a technology innovation company. It focused on innovating with iron and electronics and transmission, on drive trains. We're doubling down on that. [But] you have to be willing to reinvent who you are and how you get results. That's the essence of any tech company. They (a tech company) didn't just pivot to get where they are today. They are still pivoting. If you stand still, the ground changes underneath you.

PF: Why is it important for Deere to establish that reputation as a tech company?

Sanchez: Recruitment is one. We are in the battle for talent, to attract some of the best digital developers, some of the best [artificial intelligence] engineers. We don't need just three or four, we need dozens ... someday we'll need hundreds of these people. You can't do that unless you are viewed as a tech company. I worked at the University of Illinois, at our John Deere facility on campus. John Deere is a brand name in Illinois. But, it was hard to enter into conversations with college students [on tech career paths] when you were facing tech companies. They just have a leg up on you.

PF: What are the challenges for digital developers and AI engineers as they enter John Deere?

Sanchez: When you were [once] hired as an engineer or technical person at Deere, you knew what your job, your training would be for many years. A mechanical engineer will design mechanical things, even though the tools might change. Our employees [today] are not just responsible for being good at what they do ... the expectation is that within 10 years, that person is going to have to reinvent themselves.

[PF_0420]

Past Issues

and