Big data is big business.
Two years ago Monsanto paid $1 billion for Climate Corp., a firm that specializes in digitizing and aggregating a long list of data collected from farmers’ fields. The numbers are crunched and sold to farmers so they can make better agronomic decisions and more money.
It all started with yield monitors, GPS technology and yield maps and evolved to multiple layers of information from soil type and fertility to topography, rainfall, heat units and crop variety. That information is then connected to seeders and fertilizer and pesticide applicators to vary seeding rates and applications on the fly. Instead of treating a field all the same, inputs can be adjusted to suit conditions, in effect creating many fields within a field.
The technology behind precision agriculture and micro-agronomy, produces the information, which when aggregated across many fields, is big data.
Other large companies, including DuPont Pioneer and John Deere, are offering big data services. The focus has been on the United States, but Canada and Brazil top Climate Corp.’s expansion list.
Climate Corp.’s chief scientist David Fischhoff told Forbes in 2014 big data will change the face of agriculture.
“We can see things we’ve never seen and know things we’ve never known so we can tap more of the potential in every seed, every acre and every dollar we invest and use a new level of decision-making in the pursuit of higher yields,” a company video says.
Do farmers see it the same way? A farmer-panel on big data at the annual Manitoba Agronomists’ Conference Dec. 16 was cautious.
“It is worth something,” Roland farmer John Bergen said. “It’s not worth something yet. Our data is worth something too. In exchange for our data I guess we should have access to everyone else’s.”
Although farmers are skeptical about yield data, Bergen added, “I think big data can… take away some of those reasons not to believe,” if the variables are reduced.
What would interest Bergen? Being able to access 2015 soybean yields grown on similar soil to his, planted before the spring snow to see what impact, if any, seeding into cooler soils had.
“Once we can do that then I think we’re into something really good,” he said.
Adam Gurr, who farms at Rapid City and also does independent research through his company Agritruth Research, said he will likely buy big data services in the future, but mainly to access data management and analytical tools.
Dan Hacault of Swan River stressed that data from his own farm is what’s most important to him.
“For my farm it’s the small data I want,” he said. “It’s the stuff I collect for my own farm that’s more value to me than the amalgamation of a pile of data.”
Earlier in the conference Dan Frieberg of Premier Crop Systems in West Des Moines, Iowa also focused on the value of capturing data from individual fields and using it to assist the owner to make better agronomic decisions.
Big data is crude, but it’s better than it was, he said.
“We have to do better than averaging the responses from tiny plots and pretending that those results fit across huge spatial variability that exists within fields,” Frieberg said.
“Every agronomist and every farmer knows there are differences across their fields.
“The real key is let’s not pretend all this land is the same agronomically because it’s not, and we can use the equipment that growers have already invested in and help them with their data to do a better job of managing what they do in agronomy.”
Local data is important, said Dan Wright, Monsanto Canada’s trait launch lead, who is working on Climate Corp.’s Canadian expansion, but aggregated data can reveal important information too, he said. For example, based on data from 3,800 fields Climate Corp. found 10 per cent didn’t get enough nitrogen, costing the farmer an average of $54 an acre in reduced yield. It also found in 40 per cent of the fields too much nitrogen was applied costing producers an average of $13 an acre.
So besides boosting yields, collecting and analyzing data can help the environment. Nutrient run-off is attracting more government attention and regulation.
Only an estimated 10 per cent of Manitoba farmers are collecting data now from their entire farm, said Mitch Rezansoff, integrated solutions manager for John Deere dealer Enns Brothers. “Why aren’t more farmers doing it? It goes back to ‘show me the benefit,’” he said.
And that’s just what Rezansoff has been doing. On one farm he discovered operation overlaps totalling 200 acres, resulting in $24,000 worth of extra inputs being applied. Modifying the anhydrous ammonia application with sectional control saved $6,400, paying for itself in a year, he said.
Paying for sectional control on an air seeder, either factory installed or added later, was recoverable after three and four years.
In another case using yield maps Rezansoff found a 4,300-acre farmer had zero yield on 731 or 17 per cent of his acres over several years due to excess water. Topographical maps were used to see how to remove the water. If buying a used scraper to improve drainage could cut the lost acres to 200, it would pay for the scraper in one year, not including the time to do the work.
Speakers agreed more data will help farmers make better decisions and make them more efficient, more productive, sustainable and profitable.
“Like it or not big data is the new reality in farming,” Rezansoff said. “It’s here to stay. It will become more embedded into the equipment, the technology, the capabilities.”
What’s perhaps less clear is the extent farmers will focus on their data compared to what has been aggregated. It’s easy to imagine both. Either way farmers will have to see the value and be confident of the data itself.
Gurr said he’ll test recommendations based on pooled data on his farm before fully adopting them.
Bergen noted big data companies are focused on the U.S. Midwest, where crops and conditions are different from here. Firms will have to invest a lot to get the data necessary to build predictive models that apply to Western Canada, he said.
And while local and big data will become more important, farmers or their agronomists won’t be replaced.
“Farmers’ intuition is paramount,” Rezansoff said. “Can data replace the farmer’s experience and intuition? The answer is no. It complements it.”