Trimble Working on Predictive Technologies

Trimble to Bring Implement-Based Tools to Its Steering, Guidance Technology

Dan Miller
By  Dan Miller , Progressive Farmer Senior Editor
Trimble plans to mount on implements its newest technology systems that will also operate across mixed machinery lines. (Photo courtesy of Trimble)

As many original equipment manufacturers (OEMs) vertically integrate, Trimble's approach is different. It's a bit like Apple vs. Android, Trimble says.

Trimble, based in Westminster, Colorado, bills itself as the largest independent precision ag company in the world. Its research and product development work focuses on building near-real-time management capabilities for farmers who often operate mixed fleets of machinery.

OEMs like John Deere and CNH Industrial are motivated to sell iron and develop tech for their equipment. That hinders the work of smaller and mid-sized players with mixed lines. Trimble is designing its technology to give analytical power to all farmers, across mixed fleets.

DTN/Progressive Farmer spoke with Jim Chambers, senior vice president and general manager worldwide of Trimble's agricultural division, about Trimble's work, products informed by data collections and put to work with artificial intelligence. The question-and-answer interview begins here.


DTN/Progressive Farmer: Pick the timeframe, two years, five years: What do you see in the future for Trimble Ag?

Jim Chambers: What's got us here today has been our leadership in steering and guidance. Going forward, we still expect to continue to be a leader in steering and guidance. We're bringing new technologies into the market in the next year that continue to enhance our performance in that area. But we see a lot of technology moving to the implement. We can drive straight in the field, but can we make that implement more effective and more efficient.


DTN/PF: How is precision farming evolving?

Chambers: The first phase of precision ag has been about managing variability in the field caused by things that don't vary frequently, such as slope and soil type and water-holding capacity and organic matter. You create zones and you could do that in the off-season. Now we need to deal with the variability that happens in the field very frequently from weather to pest infestations. Some of that means being able to (manage) in near real-time. But some of that means that I have to be more predictive in what is going to happen. It is (data) that will drive those (artificial intelligence) engines. So, our view is (to) put more technology on the implement. And while we're doing those tasks with the implement -- spraying or planting more precisely -- let's gather as much data as we can, so that we can feed these AI engines to (create) more predictive models.


DTN/PF: Talk about predictive modeling.

Chambers: In some regards, we have predictive technologies. We gather data and the insights from yield and slope, those sorts of things. You can set up field zones, based on (that information). But I would say in terms of bringing offers to the market that are predictive in nature, like "Hey, you're going to have a pest infestation in the next 48 hours, you need to get out there in front of that," that that type of technology, we've got the ability to do that, but it's more in the (research and development) phase. To pull that off, you need more data than just one farm (and data from commercial input suppliers). I can't just take your data from your farm and only use that. It takes massive data sets to really get it meaningful because you're only going to get this chance. If I tell you it's going to happen and it doesn't happen, you might say "I'll give you that one." Then, I tell you again it's going to happen (and) it doesn't happen, you're going to say "Forget it."


DTN/PF: So, the camera, the AI capabilities with them, are tools to map disease and infestations, or map weed species in a field?

Chambers: The cameras themselves are not the secret sauce -- they are commercially available. The secret sauce are the AI algorithms (that) translate images into something insightful. With resistant weeds, it's important to get an actual map of exactly what we (have) and where they are at. (We might) have to change tillage practices and crop rotations.


DTN/PF: In gathering data, it's not so much gathering data for data's sake, correct?

Chambers: We're not an agronomy company, we're not trying to replace the agronomist. But we're trying to get data and information into their hands so they can make decisions. We look at workflows. We're trying to understand what are all the steps in a job. What are all the things that a farmer needs to do in that application? What is he satisfied with, and not? So here, it's not data for data's sake. But if I have this information, what would I do differently and how do you need to receive it and act on it? This is where Trimble turns data into insights. And then we want to make it as easy to execute it in the field and document what was done. Did we get the results we desired? How do we need to modify the next plan?


DTN/PF: Data collection and transfer requires good connectivity -- high-speed broadband services. If you would, explain the broadband environment.

Chambers: We have experts inside (Trimble) that can go much, much deeper than I can. When we talk about near real-time (analysis), the ability to transfer significant amounts of data, you're not going to get that done (over) old 3G systems -- you don't even have decent cell service. Trimble, through industry associations, is a big supporter of building that infrastructure. That being said, we recognize that that's a long-term challenge. So, secondarily, we are constantly looking at how can we minimize the amount of data that actually needs to be transferred. Let's slim that down to what are the critical things we need.

Dan Miller can be reached at

Follow him on Twitter @DMillerPF

Dan Miller