Satellites Are Changing Our Daily Lives
We all have used this technology in everyday movement around our Urban and Rural environment. As each yearly quarter approaches the technology gets better and better. It has intertwined so neatly in how we go from an unknown place to another unknown area. Some of us can remember the phonebook being our guiding star. Others from way passed generations can remember the green books that gave us locations to sleep and eat in early travel. Those days are long gone and we look forward to even better mapping as satellites become more sophisticated. The imagery we receive currently can map to a few yards of anywhere on this globe.
A number of companies use satellite imagery to predict annual farm yields—mostly focusing on major crops like wheat, corn and soy—and produce estimations that are useful for farmers and commodities traders alike. But Vinsight, a small startup in California, has decided to instead focus on grapes and almonds, two of the state’s most valuable agricultural products. “Even the U.S. Department of Agriculture doesn’t forecast those crops,” says founder Megan Nunes. “We thought it made sense to apply the technology to a sector that really needs it.” Grape and almond farmers typically see a 30-40% error rate when they predict their seasonal outputs, Nunes says. With Vinsight’s technology, which uses machine learning to analyze satellite images, taking into account external factors like location, weather, and historical performance, farmers can get a yield prediction that is three or four times more accurate. That information, says Nunes, can help them save on labor costs, estimate their revenue for the year, and strike better deals with buyers.
Another analytics company, FarmShots, digs even deeper. The North Carolina–based startup analyzes specific fields and patches of farmland, studying factors like light absorption and land elevation to detect the presence of pests or diseases on individual fields. FarmShots alerts its clients when there’s a problem, and is currently building out its recognition algorithms to easily identify the cause, such as fungus or trapped rainwater. Through a recent partnership with John Deere, the company’s technology has been built into tractors and other equipment, so those findings can automatically direct machines on the ground. “You don’t want to put on an even coat of fertilizer,” says FarmShots CEO Joshua Miller. “We generate a map to instruct the machine to put less fertilizer on the areas that are healthy and more on the areas that are unhealthy.”
Before the era of CubeSats, government-owned satellites tended to cover only the more populated areas of earth—leaving remote corners of the ocean in the dark. Shipping routes in the Arctic, for example, weren’t covered by satellites or signal towers, which led to a dearth of knowledge about who was passing through and what they were doing. In March, Spire partnered with the National Geospatial-Intelligence Agency (NGA) and Ball Aerospace to monitor these blind spots. “The suspicion is that there’s more traffic there than we realize,” says Nick Allain, head of Creative and Brand at Spire. “Whose water are they actually traversing? Where is an oil spill most likely to happen? Are ships meeting in the middle and sharing things they’re not supposed to be sharing?” The partnership will look to gather that information and come up with solutions that prevent black-market trading and make Arctic sea routes safer for shippers.
Farther south, Spire has used its capabilities to prevent similar dangers in the Indian Ocean. It works with the Indonesian government to cut down on illegal fishing activity by flagging ships that are in restricted waters, and is currently testing a new capability that can detect what kind of fish a boat is trawling for based on its patterns at sea. “By looking at the pattern, you can say, ‘Wait, they’re not supposed to be fishing for crab, they’re supposed to be fishing for tuna,’” says Allain. “It was hard to track that before.” Spire also provides its data to piracy forecasting companies, helping them keep an eye on risky spots in the ocean.
When disaster hits—whether it be a tsunami or a mortar attack—satellite images can be a crucial tool for governments and aid organizations trying to assess the damage and direct relief efforts. Planet, which has launched nearly 150 satellites into orbit, uses its images to create maps that show an affected area before and after a disaster, helping field workers quickly identify roads that have been blocked or important buildings, such as schools or hospitals, that have been damaged. “It can really increase the efficiency and the effectiveness of response efforts,” says Tara O’Shea, a program manager for impact initiatives at Planet. “Rather than having to send folks out into the field to survey, which can be costly and time intensive, we can fulfill that need with our imagery.” After a hurricane hit Haiti last fall, the company jumped into action and produced before-and-after maps of the country within days.
Around that time, Planet decided to form a disaster response team focused on these efforts. Many aid organizations or community groups don’t have staff members with image processing and geospatial analysis capabilities, which are critical for turning the satellite images into useful information. “People on the ground just need a simple PDF that they can laminate and carry outside,” says O’Shea. To get those assets to first responders faster, Planet worked with the Digital Humanitarian Network to enlist about 20 volunteers from around the world; after receiving training from Planet on how to use its platform and work with its imagery, those volunteers were put on call, ready to respond to a disaster at any time of day.
For researchers, the implications of satellite imagery are nearly endless—they can be used to monitor everything from deforestation in the Amazon to the annual bloom of tropical plants. In April, Planet launched a new program to facilitate this work, opening its platform to anyone with a university affiliation. Scientists from Stanford to the University of Oslo use Planet’s images and data, as well as factors like snowfall and sea level, to track the movement of Greenland’s Jakobshavn glacier, which is famous for shedding massive amounts of ice into the ocean each year. “It’s an area of intense scrutiny—it’s one of the hot spots in the cryosphere community,” says Joe Mascaro, a program manager at Planet. In Florida, Mascaro says, another researcher is using Planet’s platform to study the effect of an invasive ant species in India by surveying the health and population of Acacia trees.
Orbital Insight has also launched initiatives in this space: for the past two years, it has been working with the World Resources Institute to keep an eye on deforestation. By looking for warning signs, such as new roads being built in undeveloped areas, the company is hoping to prevent deforestation before it happens.
The U.S. government uses satellites for more than just keeping an eye on North Korea. The Defense Department recently granted image analysis company Descartes Labs $1.5 million to study food security in the Middle East and North Africa. “The US spends a lot of money surveying farmers, but if you go into developing economies, those numbers don’t exist,” says cofounder and CEO Mark Johnson. “Across the Middle East and North Africa, where it’s not cash crops but crops that sustain the population, there’s no good way of alerting people to a food shortage.”
To remedy that, Descartes is scanning farmland—both large-scale operations and smaller fields in rural areas—for early signs of famine, which can precede sociopolitical discord. That makes it quicker, easier, and cheaper to identify such regions and try to prevent conflict. “If we see a shortage, we can send in humanitarian resources rather than waiting for famine and unrest,” says Johnson. “We have all these pictures, and if we analyze them, we can send people in at the right time to the right places.”
In 2015, Orbital Insight partnered with the World Bank to study how well its technology could measure poverty rates and economic growth, focusing on a small part of Sri Lanka. The two organizations are now using those insights to test the technology in Mexico, using satellite imagery, machine learning, and survey data to gauge how many people live below the poverty line in different municipalities. “Traditional household survey data is the gold standard for accuracy in poverty measurement, but it is expensive to collect,” says David Newhouse, Senior Economist, Poverty, at the World Bank. Surveys are also conducted infrequently and often fail to accurately capture rural areas.
Using satellite imagery to analyze an area as large as Mexico will likely produce less accurate results than a household survey, Newhouse admits, but the frequency with which it can be done could still be beneficial. The Mexican government uses poverty maps to direct social funding, but those maps are only produced every five years or so, says Carlos Rodriguez-Castelan, senior economist and poverty global lead at the World Bank. “With more frequent data, we could have a better picture of people moving in and out of poverty, and who really needs assistance in a particular area,” says Rodriguez-Castelan. “This could be a critical tool to better target governmental programs and investments towards those who need them the most, at a much more granular level.”