Today, there are around 300 or so Earth observation (EO) satellites in the skies above, operated by a mixture of governments and commercial companies and constantly taking pictures of the planet. The number of commercial imaging satellites could balloon to nearly 1300 or more craft in the next five years — if all 20 or more start-ups deploy all of the satellites they want as described in public statements.
Everything moving on the surface of the planet is already being recorded in an unprecedented way, with terabytes of imagery being logged each day in visible and near-infrared (near-IR) light. New constellations will add radar to see through darkness and clouds in the next two years while full-motion, 24×7 video of anywhere on Earth may be available within 5 to 7 years.
Leading the commercial EO constellation revolution is Planet. Over 100 small Dove spacecraft the size of a loaf of bread record the entire planet’s surface on a daily basis at a resolution of 3 to 5 meters in visible light, with higher-resolution imagery of under 1 meter plus HD video taken by the company’s 13 larger SkySats. Planet logs over 6 terabytes (TB) of imagery a day, every day, and has accumulated over 6 petabytes of images to date since the company started operations in 2013.
A number of start-ups are building satellites capable of 1 meter to 2.5 meter resolution. The typical satellite size, like Planet’s SkySat series, is roughly the size of a dorm fridge and delivering 1 meter resolution. However, some are pushing engineering tradeoffs to produce smaller and lower-cost imaging satellite beyond the already-compact Planet Dove 3U-size CubeSats.
Smaller satellites are cheaper and faster to build and put into space, with the ability to put a dozen or more into orbit with a single launch. Seattle-based BlackSky plans to build a constellation of 60 satellites while Canon — yes, the same Canon that builds copiers and digital cameras — says it wants to launch up to 100 satellites to photograph the entire Earth on an hourly or less basis.
EarthNow intends to go behind still images, launching a constellation of up to 500 satellites to provide real-time, 24×7 video coverage of the entire Earth. incorporating on-board processing to interpret what it sees in real time. But the company is still in early stages, with a rough estimate of $1 billion to build, launch, and operate the constellation and no certain dates set on when the venture would start putting hardware into orbit.
But you can’t see details through clouds or at night with visible imaging. Radar satellites provide a different view not dependent upon the sun or affected by weather. At least four start-ups are building commercial radar satellite constellations, with sector leader Finland-based ICEYE funded to the tune of over $50 million and planning to launch 9 radar sats by the end of 2019.
Still others are working on multispectral and hyperspectral — multiple wavelengths of light — imaging satellites to provide still more detail and in-depth information on everything from crop health to hidden underground structures. Hyperspectral imaging doesn’t provide the fine 1 meter or less resolution detail that (visible) imaging does, but can easily identify materials, find structures, and detect emissions.
Readers should note the estimate of nearly 1300 satellites is heavily weighted by the EarthNow video constellation of 500 satellites. Initial start-up costs for an imaging company are much more variable and larger than in the IoT world, due to the need for more expensive equipment (cameras) and higher data rates to download data. A satellite IoT company can build and launch a pair of satellites for under $5 million. Imaging start-ups require anywhere from $10 to $15 million to build and launch one or two satellites, depending on the size and capabilities of the satellite. ICEYE has launched an initial pathfinder-style radar satellite with around $10 million in funding, for example.
New entrants to the imaging field will have to compete against well-established Planet. The company has an established database of 6 terabytes (and constantly growing!) of imagery, a number of government customers, and is treating imaging as a Big Data business with all the trappings of APIs and machine learning to provide easier access.
Start-ups will have to compete on pricing, better features, and easy access. They will also have to work to expand the market for imaging data beyond traditional areas such as national security, natural resource usage, transportation, energy, and urban resource planning. The insurance industry may be an area for expansion in the near future.