Big Data is a chicken-or-egg scenario on the farm
How 5G will impact the future of farming
According to McKinsey, the agricultural sector is the least digitized industry in the U.S. Yet, agriculture already produces large volumes of data -- especially from IoT. There's hardly a lack of scenarios for digitizing farming, with the common thread for all the use cases is that they each involve Big Data.
Among the obvious use cases, precision farming uses sensory data to tell farmers exactly where to plant and how much to water and how to fertilize. With a baseline of rules, this provides an excellent case for using machine learning that can subsequently adapt those rules to the specifics of the actual field and correlation with yield data. Food safety and spoilage prevention can be enhanced through use of smart devices that detect ambient humidity, temperature, chemical contamination, and the presence of gasses signaling the presence of harmful microbes.
SEE: The Future of Food (ZDNet/TechRepublic special feature)
There are other Big Data use cases not necessarily specific to agriculture with the potential for eliminating cost and waste such as asset management and predictive/prescriptive maintenance of farm equipment, and supply chain optimization for getting crops efficiently to market.
But Big Data on the farm is a chicken-or-egg scenario. The use cases are there. But then there's the matter of infrastructure, data ownership, and economics. Which one has to come first? We had a chance to get gain some insights on the last mile bottlenecks to Big Data on the farm as we sat in on a recent Big Data in Agriculture roundtable convened outside Washington.
Let's start with connectivity. Ironically, the farm is one of the most- and least-connected industries out there.
According to Mark N. Lewellen, manager of spectrum advocacy for Deere & Company, 60 - 70% of farm machinery is already connected to GPS systems. On the plus side, farmers use those GPS systems to automatically guide their tractors and combines on where to plant and maintain their crops. In emerging regions, such as Brazil, where huge farms totaling thousands of acres are being cleared from the brush, these machines have become robotically guided and are operating 24 x 7. While few if any US farms are that automated, the resulting wealth of data provides bounties to all of the suppliers: the manufacturer of the equipment, and the suppliers of seed, fertilizers, and herbicides that are applied to the land.
Also: Agriculture 4.0: How digital farming is revolutionizing the future of food (cover story PDF)(TechRepublic) | Smart farming: How IoT, robotics, and AI are tackling one of the biggest problems of the century (cover story PDF) (TechRepublic) | How self-driving tractors, AI, and precision agriculture will save us from the impending food crisis (cover story PDF) (TechRepublic)
But beyond GPS, there's a huge gap with connectivity. Drive cross-country on the interstate, and you'll usually be within range of a cell tower. Get off on one of those exits onto a back road, and you're likely to lose all those signal bars.
It's a sign of the times that a major agricultural equipment supplier felt the need to recruit a communications specialist. Lewellen's mission is lobbying for more broadband coverage. The irony is that, while mobile carriers typically use population densities for network planning, the USDA has compiled surveys that have economically mapped farmland. At first glance, it looks tempting to use the USDA maps as a cudgel for getting more bandwidth out there. But, the challenge remains the scale of the necessary build-out: cell towers may "only" cost $250,000 a pop, but to blanket the countryside, they'd have to build towers every five miles.
Another key structural challenge is working out who owns the data. When your GPS-equipped tractor or combine gets machine data from the equipment and agronomic data on field conditions, does the data belong to the farmer or the manufacturer?
According to Todd J. Janzen, an attorney who specializes in agricultural law, there is nothing on the books that protects agricultural data. Janzen has proposed a certification system for data transparency that would act a form of Good Housekeeping seal documenting how agricultural suppliers handle farm data. The program, which is still in its early stages, currently counts a handful of suppliers that have signed on.
Let's suppose that farmers can gain trust that their data will be handled responsibly, even in the event that their equipment or feed supplier changes hands. And let's suppose that magically, broadband materializes. What economic incentive would they have to sign up for services to digitize their farms given that their business is low margin to begin with?
Food Trust is one of the use cases that IBM is promoting for its blockchain initiative. The trigger is food safety. According to IBM, every year, one in 10 people fall ill due to contaminated food. IBM is promoting blockchain as the backbone of a global food safety solution because of its ability to provide what it terms a "trusted" environment for transactions. In this case, those transactions would be comprised overwhelmingly of IoT data from sensors that are deployed with food.
Also: A day in the data science life: Salesforce's Dr. Shrestha Basu Mallick | Port of Rotterdam plots IoT rollout, efficiency push with IBM, Cisco, Axians | Data scientist: The cult of pan-galactic data doesn't work for business | TechRepublic: Why the reality of big data is finally catching up to its hype
The challenge with deploying a blockchain for tracking food safety is that for all parties involved -- farmers, logistics companies, and food processors -- this becomes a cost center.
Here's the kicker. You're still tracking food safety, but what if the business proposition for the blockchain is flipped to tracking food provenance? The value proposition is that farmers could charge more for their harvest if they can tag their crops from field to checkout aisle. The farm-to-market movement has gained adherents from upscale restaurants and from Whole Foods shoppers who value knowing where their food is coming from and having the assurances that those free range eggs really came from un-caged hens. But it's easy to dismiss such programs as appealing only to the 1%.
Results from China, which has had its share of food contamination problems, reveal that the market for such tagged food could be wider. According to a survey conducted for Wal-Mart in China, food shoppers indicated that they would be willing to pay premiums of up to 30% more if they had a smartphone app that could scan the QR code of the food item to identify the source and wholesomeness.
For farmers, such programs could disintermediate the middleman where it comes to pricing power. The hurdle is that it would require the farmer to invest in the tagging and sensor technology up front. IBM has yet to come up with the pricing model to get the farmer connected.