By Jaclyn Krymowski
Like much of the world, data now plays an integral role in our daily life. This is extended to modern agriculture from research to daily farm management. While this affords farmers many advantages and opportunities, it opens serious questions on data ownership and security.
In a 2017 survey by Agriculture and Agri-Food Canada, 55 per cent of respondents said farm data privacy and security was either a barrier or significant barrier for them to adopt precision ag technology.
Despite such concerns, companies collecting agricultural data continues to grow. According to Research and Markets, the entire precision agriculture sector alone will be worth $18.4 billion by 2024.
To do their jobs, such technologies rely on complicated software that continually collect, interpret and store data. The same applies to advanced livestock recordkeeping software, robotic equipment, field maps and more.
However, just because something is intended for on-farm use does not mean the farmer and employees are the only ones with access to it. Ultimately, the agreements between buyer and software company are intended to determine and uphold privacy—or are they?
“Many companies don’t have a company wide data policy that everyone involved knows and understands,” explains Todd Janzen, attorney and co-founder of Janzen Agricultural Law, LLC. “They make all sorts of promises to farmers, but they’ve never taken the time to actually sit down and figure out what are company-wide policies with respect to farmer data.”
Value of data
As digital connection becomes a bigger part of daily life, more of us have become aware to the value of data. Massive amounts of consumer information are collected and stored every day from the devices we use, providing incredible value to advertisers.
Cüneyt Gürcan Akçora is a professor of computer science and statistics at the University of Manitoba and explains how in most industries they can do back-testing on this past information and compare it to the current year to understand its value. Agriculture data is not priced this way.
“In agriculture, the main issue is the company doesn’t want data to leave their servers and they don’t want competitors to know what they are doing,” says Akçora.
Janzen also points out it’s a common pitfall for ag data companies to assume they can create and implement their own policies similar to what other industries have done.
“That mindset doesn’t appreciate that farming is a very unique business,” says Janzen, noting how farmers have different needs and concerns compared to most industries. “Agricultural data is proprietary business data, and it should be treated like consumer data is on Facebook or other social media platforms.”
One person who believes that fellow farmers are truly unaware of what their data represents is Brian Tischler. He farms 3,000 acres in Mannville, Alta., and firmly believes many farmers do not comprehend the true value their data generates, stores and provides third parties.
“Take something as simple as a moisture probe connected to the internet and giving georeferenced hour-by-hour moisture data,” he says. “What is that worth?”
If valuable, there needs to be a way to quantify it. Tischler suggests it may need to be farmers like himself who advocate for research avenues or explore opportunities for compensation.
Something else worth considering is how often companies are not only collecting, but also learning and storing data.
Tischler compares this idea to how daily, data is constantly being collected when we are unaware including online activity, purchases, GPS tracking and which businesses we frequent.
However, having all this information at the fingertips is extremely useful to farmers as it accumulates over time.
“The more data that you have, the better decisions you can make going forward,” says Janzen.
For example, the information collected on a 1,000-acre wheat field that includes weather, soil conditions and yield could be used to predict what might happen if a farmer made the move to plant 20,000 acres of wheat in the same area.
For things like this, farmers need not only their data but a system and service to process it. It takes a very complicated series of computer processing, analysis and machine learning to take seemingly arbitrary bits of information and translate them into something that is valuable and understandable.
“We have to model the data and learn a machine learning model from it even when we do not understand the nature of the data,” says Akçora. “Once the data is put into a digital format, algorithms can find relationship between parts of data.”
Data and privacy
When farmers agree to use software and other data-collecting technology, they enter into a private agreement with the provider. Terms and policies will vary according to individual companies and the services provided.
One of the issues, Tischler believes, is companies having their own systems without a national standard. With a unified platform and public access, private companies could still profit by interpreting available data in a meaningful way to the farmer.
“My philosophy is farm data should be like Google Earth and be public, then the services could be provided from that data,” says Tischler. “I think that’s where the future of business lies.”
Such a future could be possible, according to Akçora. “New concepts, such as federated learning, are currently being developed to train models on pieces of data that reside on different serves and locations.”
David Weins, chair of Dairy Farmers of Manitoba, milks 240 cows near Grunthal, Man. On his dairy farm, they use a variety of advance technology for daily farm management. This includes things like robotic milkers, automatic feeders and herd management software. All of these continually collect, store and interpret different kinds of data.
While the data is intended only for on-farm purposes, Weins notes the agreements they sign onto with the companies for these systems aren’t as clear cut as he would like.
“How much access they have to our data on the farm I am not altogether clear on,” he says. “Also, the level of security threats and security breaches isn’t clear either.”
He notes a lot of these legitimate concerns and questions are often not discussed in farming circles. He agrees it may be up to farmers to make their concerns and questions known.
Who has the final say?
A common concern is what should happen if data security becomes compromised or questionable, either by the company providing the service or foreign tampering.
While there exist certain recommended data handling guidelines, such as those put forth by the Consultative Group for International Agricultural Research, all legal issues and concerns come down to the original agreement made between the two parties.
“At the end of the day, what matters is that contract between the farmer and the data company. You can’t expect a government agency is going to swoop in and solve the problem for you,” says Janzen.
Weins believes that stronger regulation in these sensitive areas is important. Like Tischler, he believes having more standardization would be ideal.
“I think this is one of those things that kind of creeps up on you a little bit here and there until we’re not as aware to the dangers,” he says.
Getting farmers to appreciate the value of their data is important to getting solutions made around ag data privacy and security, says Tischler.
“Imagine asking the companies I buy inputs from to install my app that tracks their stock information, their purchase cost, and how much they retail it for,” says Tischler. “The app would constantly send me that information because that’s exactly what’s happening with farmers data. Why does it sound crazy when the data flow is reversed?”