For Farmers, A Picture is Worth More Than a Thousand Data Sets

There are several large orchards and tree farms scattered among the hills at the eastern edge of Pennsylvania along the Delaware River — managed and maintained by the same family and the same set of rules for more than 150 years. The rules to farming in the region may or may not be written down or archived. They are, essentially, based on visual information, processed through years of experience and instinct. For example, a pair of human eyes observes the yellowing leaves of a walnut tree, and a diagnosis of Thousand Cankers Disease is reached even before fingers get a chance to touch and process the soft, dark spots inside its withering limbs. This is a disease created by a fungus carried by tiny, burrowing Walnut Twig Beetles, which the human eye can only see up close. The human eye can’t zoom out to see the larger swarm of beetles, or observe the direction they are moving, or produce a time-lapse slideshow of images that display the larger environmental impact of the blight.

One Eastern Pennsylvania farmer in the region says the condition of tree crowns are often the first visible signs of poor health or disease. “It’s more likely you’ll see discolored leaves and thinning foliage before you ever notice the insect causing the problem,” he says, pulling out his smartphone and swiping through pictures of tree tops taken at a curious overhead angle. “I took these photos from a drone. My son bought me one a couple of years ago and I now use about three or four to keep an eye on the orchards.” The farmer asked not to be named for the story because he hasn’t yet registered his drones with the Federal Aviation Administration (FAA). When asked if he had ever used or considered using satellite imagery or sensing technology to monitor his fields, he says he hasn’t heard much about the technology, other than Google Maps, but adds that he wouldn’t be opposed to using data analytics to manage his farm. “Farmers may have a reputation for sticking with the old, but we constantly have to modernize and evolve to stay afloat,” he says. “It’s even more important now with the weather changing. It’s early June, we’re in the mountains and its 100 degrees for the third day in a row. I can’t say that I know how that’s going to play out.” Approximately 40 percent of the United States is made up of farmland. Environmental uncertainty of this size and scale can have a crippling effect on even the most robust economies. How can satellites help manage those uncertainties? Some would say through good, old-fashioned research and disruptive engineering. Multispectral cameras on satellites are now powerful enough to monitor crop production, evaluate developments and shifts in the physical environment, and produce highly focused weather forecasts with incredible detail. By saving farms and helping to manage and preserve our natural environment, satellites may also be responsible for saving millions (and potentially billions) of human lives.

Martien van Nieuwkoop, director of the World Bank’s Agriculture Global Practice, says that current global crop yields and food production is not keeping up with global population growth and changing diets. By 2050, global food production will need to increase by 70 percent in order to feed a projected population of an estimated 9 billion people. In order to solve this problem, World Bank is investing $4 billion in a global Climate-Smart Agriculture (CSA) campaign. “The actions of this campaign will be driven by satellite imagery and sensing data,” says Van Nieuwkoop. “Food shortage is more than just a production challenge. Its severity is amplified by climate change and an increase in extreme weather events that are devastating for crops and livestock. Unfortunately for us, food production is one of the world’s major contributors to climate change and an increase in global Green House Gases (GHG) emissions. We know this because of satellite data and the data we use to manage our cropland, livestock, forests and fisheries — essentially in order to eat — will need to get more comprehensive and accessible.”

One reason that commercial satellite imagery and sensing data hasn’t been accessible to our tree farming friend in Eastern Pennsylvania and other farmers in the United States might be that the U.S. Department of Agriculture (USDA) hasn’t done much to promote the technology. In fact, the agency has produced a lot of mixed messaging about the effectiveness of satellites to monitor and forecast crop yield. This past April, the USDA’s World Outlook Board, which produces one of the nation’s most popular and widely-used world crop forecasts, stated in a report that the agency still does not believe that satellites can accurately predict annual corn, wheat or soybean harvests and is still mostly relying on farmer surveys and random field samples for crop yield information. World Outlook Board Chair Seth Meyer doesn’t dispute that satellite data is a useful tool, but rather believes that satellites aren’t powerful enough to be the USDA’s main data procurement tool.

“It’s only been a few years now that satellites have been able to differentiate between different crops, and accurately determine crop conditions,” says Meyer. “But, the level of precision isn’t quite where we need it to be in order to rely on it for our forecasts.”

The USDA’s thinking on satellite data has since changed, but only very recently. In June, the USDA announced it would begin incorporating data from NASA’s Soil Moisture Active Passive (SMAP) satellite to monitor global croplands and make commodity and crop forecasts, as well as reports on regional droughts and floods. The resources will be made available on the USDA’s Crop Explorer website, maintained by the agency’s Foreign Agricultural Service. SMAP, launched in 2015, is NASA’s first satellite mission devoted to soil moisture, temperature, precipitation, and vegetation health research. “Variations in global agricultural productivity have tremendous economic, social and humanitarian consequences,” says NASA Goddard Research Scientist John Boten. “There’s a lot of need for understanding, monitoring and forecasting crops globally. [SMAP] is a very straightforward approach to applying that data.”

The Canadian government’s version of the UDSA World Outlook Board is Statistics Canada. Unlike its American counterpart, Statistics Canada relies heavily on satellite imagery and sensing data for the analytics behind its national monthly crop forecasts. Statistics Canada’s head of remote sensing analysis Gordon Reichert says that the increasing capability of commercial imagery satellites combined with the integration of machine learning algorithms to quickly analyze current and historical trends, has improved the accuracy of the reports and saved the agency a significant amount of money and time.

Even if national farming organizations were to get on the same page and educated farmers about how satellite data could enhance their individual businesses, the questions remain — Is space-procured imagery analytics for agricultural and environmental monitoring a sustainable business for commercial satellite providers? If so, how do you create a successful business model for selling customized Big Data-powered analysis to individual farmers without having to invest in an army of sales representatives? Investors believe that there are potentially unlimited revenue opportunities for those who are first in line with the right answers to these questions.

Research and Markets believes that the geospatial analytics segment of the overall big data market will grow from $3.41 Billion in 2017 to $13.21 Billion by 2022. There’s a near consensus among analyst reports that the remote sensing, climate change modeling, and geospatial analytics market will all enjoy a Compound Annual Growth Rate (CAGR) of 30 percent or higher during the next five years.

The accelerated growth rate for these markets can be attributed to a number of enabling factors aligning at once: the adoption of artificial intelligence and machine learning in an increasingly competitive global enterprise market; democratization of Cloud computing and hosting; a global increase in demand for visual surveillance; faster and cheaper access to space; government agencies around the world shifting from hard data-driven operations to analytics; and a burgeoning startup community backed by significant private capital investment to name a few.

SSL Chief Strategy Officer Adam Marks believes that satellite imagery and sensing analysis is and will continue to be an essential commodity for an expanding set of end-users, especially in rapidly digitalizing industries such as energy and utilities, maritime, healthcare, transportation, and agriculture. SSL is now part of Maxar Technologies, a vertically integrated company that includes a satellite imagery company DigitalGlobe, multisource data analytics provider Radiant Solutions and space subsystems and robotics manufacturer MDA. “We believe that the market value of geospatial intelligence will only continue to grow,” says Marks.

Another commercial satellite industry leader in the smart agriculture space is Planet, which in March of 2018, signed a deal with DowDuPont’s agricultural software business, Granular, to provide three years of access to the operators six-year archive containing millions of high-resolution satellite images. Granular will feed Planet’s imagery into its agricultural analytics tools and then provide the combined data sets to farmers.

Granular Co-Founder and CEO Sid Gorham says that farmers desperately need more visibility into problem solving and proven, optimal responses to unexpected changes in weather and crop health. “Farming is different today from how it used to be and seasons are shifting,” says Gorham. “With Granular, farmers can craft crop and field plans, delegate duties to employees, track inventory, and predict revenue and yield. And by transforming Planet imagery into Granular software, we’re gathering perhaps more data than has ever been available about crops.”

In Europe, e-Geos, owned by Italian spaceflight services company Telespazio, uses both satellites and airplanes to collect multi-temporal, optical and Synthetic Aperture Radar (SAR) imagery, and to extract data sets from parcels of farm land. The imagery and data are combined with GPS tracking to provide farmers with the information they need to improve the management of their land. The analyzed data could provide insight on more efficient uses of fertilizer and irrigation, improving crop quality and yield, and cutting energy and operational expenses.

“Sustainability in agriculture is the result of a balance of productivity and quality that can be very difficult for farmers and land managers to achieve,” says e-Geos CEO Massimo Claudio Comparini. “Farmers are dependent on many different types of operations that are connected to their operations, from seed producers to tractor manufacturers and government regulatory agencies. Along with the actual crops and livestock, these supportive elements also play a critical role in creating the infrastructure of the overall farm. Managing this infrastructure requires the ability to manage and process massive amount of information, quickly, because the future of farming is what we call ‘precision farming.’ Satellites, and in general Earth observation solutions, collect and organize information in a way that allows managers to quickly solve problems, save time and money, and prepare for the future.”

Marks sees smart agriculture as an application with tremendous investment and socioeconomic value and one that will lead commercial satellites into a new space renaissance. “These are the kinds of exciting new ideas that are generally good for the whole industry and enrich the experiences of life on Earth.” VS