During the last several years, numerous companies announced their intention to launch low-cost Earth-observation satellite systems, most of which will be constellations of satellites. This implies a significant increase in the amount of Earth-observation data being collected and a growing number of supply choices, but what are the industry implications?
First things first: should we be surprised? The commercial Earth-observation sector is still evolving. Although commercial data have been around for some time, only in the last 15 years have we witnessed the emergence of commercial high-resolution data, which generates the most revenue for data sales.
About 15 years ago, 80-centimeter ground-resolution panchromatic data from IKONOS had a revisit of 3-5 days with off-nadir pointing (much higher with true nadir). The data also were expensive. Since then, high (metric) ground resolution supply has increased, while data prices decreased. Comparable DigitalGlobe/Airbus systems today offer data in the $20-25/square kilometer range; IKONOS is priced lower. It’s possible to procure one-meter-resolution data for much less, but with compromises in the native geolocation accuracy, radiometric resolution, mechanisms for service delivery, etc.
Very-high-resolution data with combined high accuracy is a key requirement for markets such as defense—an important consideration when defense made up 59 percent of the $1.5 billion data market in 2013. U.S. defense end users, in particular, demand high-accuracy data, which require expensive solutions (e.g., agile systems based on control-moment gyro technology, star-trackers, etc.) on a larger platform to increase stability. This increases the cost to develop the satellite, so data are charged at a higher price.
This image captured by Planet Labs satellite “Doves” shows the West Greenland Glaciers, which drain the western Greenland Ice Sheet through Karrat Fjord and into Baffin Bay.
The maximum ground resolution of satellite imagery allowed by the U.S. government is 25 centimeters. DigitalGlobe offers 31-centimeter resolution for panchromatic data; no further companies to date are expected to challenge this limit. Geolocation accuracy due to the aforementioned solutions also is now down below five meters when measured on the CE90 deviation scale. The technical and regulatory limits of these factors are being reached.
Lowering a system’s revisit time increases the frequency in which data can be collected and services delivered—an important factor for monitoring applications relying on change detection. The increasing supply from individual companies has reduced the revisit of their overall systems, as has the launch of initial constellations such as RapidEye and COSMO-Skymed.
However, few systems to date have daily coverage; even then, cloud cover may inhibit data collection. Services requiring weekly updates, such as precision agriculture or emerging-market/business-intelligence and location-based services, can be difficult to guarantee delivery. Most end users, especially outside defense, are sensor/data agnostic—they want the solution, less the data. They then want a guarantee on the services delivery at the best price possible; if cost-benefit is demonstrated, then there are opportunities.
Building a low-cost constellation is cheaper, and a camera capable of one-meter-resolution data capture now is much reduced in cost. Although a simpler solution with a smaller platform implies reduced stability, advances in data processing on the ground now increase data accuracy using ground control points (GCPs) and/or existing digital elevation models (DEMs). Recently, Blackbridge announced that by orthorectifying data using Landsat GCPs, it was able to bring data accuracy of the RapidEye constellation down to 10 meters CE90—suitable for most applications.
This is an important concept: increasing the activity in data processing on the ground implies increased operational cost, but lower-cost satellites vastly decrease the capital expenditure (CAPEX) required to manufacture them in the first place. Satellite constellations based on smallsat technology can bring revisit to below daily at a lower cost than a satellite with very-high resolution and high geolocation accuracy. For these new constellations, the premise is largely similar: increased data collection through numerous satellites delivered at a low cost, with services focused on monitoring change detection at high frequency.
This image from Skybox Imaging’s SkySat-1 is a video frame showing the Burj Khalifa in Dubai, the tallest building in the world.
With lower CAPEX, smallsats can charge much lower prices for data and services, with the potential to be disruptive. There’s also less risk; Planet Labs lost a batch of satellites on the Antares launch headed to the International Space Stations, but the company could absorb much of this impact by producing more satellites than required and launching those without increasing cost too much. It also plans to launch its cubesats frequently on varied launch vehicles.
This approach also is more attractive to investors: a lower CAPEX to support and scalable projects with the potential to open up new markets. In the last couple years, there has been an influx of venture-capital investment in the Earth-observation sector, supporting initiatives such as Skybox Imaging, Planet Labs and OmniEarth, among others.
From announced missions to date, it’s estimated that more than $500 million in venture capital has been invested in emerging Earth-observation operating companies. This represents an unprecedented level of private investment coming into the sector, as new entrepreneurs enter and new markets are explored. A decade ago, the idea of Google acquiring an Earth-observation company for $500 million would have seemed far-fetched.
It’s of interest to note that the current Earth-observation environment is largely a U.S. storyline. Of the companies that intend to launch a low-cost constellation, only Satellogic (Argentina) and NorStar Space Data (Canada) are non-U.S. based. Most of the investment also is from the United States, much from Silicon Valley.
If data can be produced with decent ground resolution and accuracy, and at a much lower price than what’s currently available, does this spell trouble for existing operators with higher-cost systems? The answer is likely yes and no.
For the existing market, particularly in the defense sector, there are requirements currently beyond the scope of the announced new constellations, including higher ground resolutions, secure/encrypted data delivery, higher native geolocation accuracy, speed of delivery, etc.
There may be some impact in more cost-sensitive markets, such as natural-resources monitoring, but the new companies can open markets further through lower-cost, more-frequent data collection. Areas being explored include location-based services supporting market intelligence, providing regular inputs on crop yield/health, supporting competitive assessments (energy, retail sectors, etc.) and supporting financial communities.
It’s also important to note that none of the newly announced constellation initiatives have reached full capacity; few have launched a satellite (Skybox has two satellites in orbit, Planet Labs aims to launch 131 cubesats by mid-July, BlackSky Global aims to launch the first of its satellites before the end of 2015). For these constellations to come to fruition, additional investments likely are required. It’s also doubtful that all the announced missions will succeed; there’s a risk that failures of these announced missions will scare off future investment. But whatever the outcome, the Earth-observation ecosystem continues to evolve.
Smallsats Disrupting GEOINT Community
A likely driver of the smallsat revolution will be buy-in from the geospatial intelligence (GEOINT) community, which is the major funder of today’s commercial-satellite market.
One of the most challenging things affecting our people in the near future is the smallsat revolution,” says Robert Cardillo, director of the U.S. National Geospatial-Intelligence Agency, at the U.S. Geospatial Intelligence Foundation’s GEOINT 2015 Symposium held June 22-25 in Washington, D.C. “Some are uncomfortable with this seemingly uncontrolled movement of more and more sensors into space. And while I recognize that there are two sides to the world’s growing transparency, I am energized and enthused about this development.”