The Center for Data Innovation spoke with Mike Preiner, co-founder of San Francisco-based precision agriculture company Granular. Preiner discussed the kinds of data that help farmers make better decisions, and how applied physics relates to problem solving in agriculture.
Joshua New: A smart house or a smart city is pretty easy for most people to wrap their head around, but a smart farm is a different story. Could you walk through what a smart farm looks like, in terms of the technology involved and how it operates differently than a traditional farm?
Mike Preiner: I would never call a less technology-savvy farm “dumb,” but we’re definitely seeing a lot of ways that farms are becoming increasingly data-driven and smarter as a result. If you could imagine a smart city or house, a smart farm really isn’t that different. At its core, the concept is really the same—using information technology to increase efficiency by gaining better knowledge and making better decisions.
So for example, when people think of a smart city, they might think they can log into a computer and see where all the traffic is and take a different route to work. This is the same kind of thing that happens on a smart farm—you can track the locations of all your tractors. If you have to schedule a grain delivery, you can see where the closest tractor is in real time and see what my team is doing to guide your decision making. And take smart homes. If you’re on vacation and you want to adjust your thermostat at home, you can do it right on your smartphone. You can do the same on a farm—a farmer can remotely monitor what’s going in their grain storage. A lot of the fundamental benefits of information technology, like using mobile tools to make smarter decisions, are the same whether you’re on a farm, in a home, or in a city.
New: Of course Granular uses data from farms themselves, but are there any less obvious data sources that farmers find valuable that they couldn’t easily access before?
Mike Preiner: Absolutely. There are a wide variety of sources because farming is a very interdisciplinary business. Of course, things like weather data are hugely valuable. Granular pulls in historical, real-time, and forecasted weather information from government data sources and models.
Other types of government data sources are valuable, too. Elevation data and topographic information, as well as soil data from the United States Geological Survey, are really important to farmers. We can give this data directly to farmers, but we also package this data and turn it into models so we can say things like, “based on the weather and soil information, it’s too wet to plant that type of seed right now.”
There are a couple other buckets of data that are useful to farmers, such as private data—not from the farm itself but from third parties. This could be a database of seed properties, for example, that describe seed traits and agronomic characteristics. Even live market data can be really useful. Grain is a commodity, and when grain farmers know what the market prices and futures prices are they can make much better decisions.
New: You hold a PhD in applied physics. What’s the connection to data-driven agriculture there?
Mike Preiner: I grew up in a small rural town doing primarily aquaculture. But in a bigger scale, with hard sciences, applied physics in particular, you develop a background of multi-level problem solving. There’s the really high, general systems-level of problems solving, but you also have to solve really detailed, intricate problems that require a lot of technicality. The tools hard sciences give you to do that apply very well in farming.
As I mentioned, farming is very interdisciplinary. To be a good farmer, you and your team have to be able to connect the dots between things like where the market is headed to the latest tech advances to what the weather is doing. There’s so many different parts that you need a systems-level overview, but to really be efficient you need to get really into the details.
And of course, the basic data analysis and thinking are also really important for agriculture problems. You’re often dealing with very sparse amounts of data. There’s an expression that goes, “you only get 40 shots to be a good farmer [assuming a farmer works for 40 years],” meaning that farming really only happens once per year. When something happens one year, you need to be able to identify if this is a trend or if it’s likely to happen again. A math or science background helps immensely with that.
New: Can you describe the AcreValue tool? Its value seems pretty obvious—is there anything else like it in the marketplace?
Mike Preiner: Anyone that’s ever bought a house will know Zillow, and AcreValue is basically Zillow for farmland. When you log on, you look at a map and get a value estimate for any given field or parcel of land. It’s a pretty simple concept and it sounds obvious, but the tools and infrastructure to do this didn’t really exist until recently. You couldn’t do this kind of large-scale data analysis, but you also couldn’t reach enough consumers. People have always wondered what the market value of farmland is, but we’ve only recently advanced enough to be able to pull in things like geospatial data and soil information to accurately determine this price. We’re one of the first companies to do this, but we’re only just beginning to see what’s possible in this space.
New: Precision agriculture seems to be increasingly popular, almost approaching buzzword status. Is this just hype, or is it really taking off on a large scale?
Mike Preiner: I think the term precision agriculture encompasses a huge amount of technologies, all of which have different impacts on farming and adoption rates. It’s an umbrella term, so to understand if it’s taking off you have to peel back the layers and look underlying technologies.
So, for example, in the past 10 or 15 years we’ve gone from navigation systems that tell you which direction to go on a tractor to tractors that literally drive themselves with accuracy down to a fraction of an inch. Then there’s applications that can automatically turn on and off fertilizer or chemical systems as you go across a field.
These technologies are definitely transforming the industry and making it more efficient, but in some sense this is becoming the new normal. I was at a conference a few years ago where someone said, “In a few years, we’re not even going to talk about precision agriculture anymore. Not because we won’t be doing it, but just because we’ll be calling it ‘agriculture.’” In the businesses, this kind of technology is the standard package.
It’s part of a process that’s been going on for a hundred years. It used to to be that 90 percent of the population farmed. Now, it’s just two percent, thanks to a steady increase of technological innovation. Nobody talks about the increase in mechanization 60 or 70 years ago because everyone assumes people are using tractors and machines. Precision agriculture is the next stage of an evolutionary process, and pretty soon, we’ll be looking at what’s next.