IEEE Geoscience and Remote Sensing Magazine - December 2023 - 172

data using a form factor that is quite inexpensive compared
to historical missions of a similar specification. The average
system that can capture 25-cm X-band data in the ground
plane costs about as much to operate each year as Umbra
has raised in its entire history as a company. And that's just
one satellite. We have six satellites in orbit already; seven
and eight launch later this year (Figure 1).
GRSS: Why is running an open data program important
to your company?
Capella: Capella's mission is to make SAR more accessible,
demonstrate its unique value, and help solve the hardest
challenges we face on Earth. To do this, we are working
with a broad community of the brightest data scientists and
software engineers.
Our collective efforts and imagination are unlocking
the most value from these data to transform how we
live. Capella's open data program is part of the tiered
Capella SAR community (open data, data cooperative,
and data grant). The community program enables users
to have increasing access to SAR data, and to help them
develop tools and algorithms to solve complex and urgent
problems.
Capella's open data is a continually growing, timely
dataset that consists of hundreds of images, covering every
continent with examples from agriculture and aquaculture,
energy and natural resources, infrastructure, maritime, environmental,
as well as humanitarian and natural disaster
use cases (Figure 2). With the creative commons license,
this makes Capella's data ideal for computer vision and machine
learning research and development.
Umbra: If you can make something just as good or better
than what others offer for one-tenth the cost, you have
two basic strategies to consider pursuing:
◗ You can artificially constrain supply and try to sell data
at similar prices to historical norms, just with a much
better margin. In this strategy, you are using an arbitrage
to siphon cash from an existing market.
◗ You can use your supply advantage to flood the market
and try to grow it overall (not just for yourself) by finding
new markets and customers (and sucking the margin
out of your competitors' businesses at the same time).
We've opted for the second option, because we believe
the commercial market for high-resolution SAR data is
essentially nascent, but will be very large in the future
based on both advances in technology-like cloud computing,
artificial intelligence-and demand-driven by
climate change.
We've released more open data than all of our competiFIGURE
1. Umbra Lab-provided SAR image of New York, NY USA.
tors combined. In fact, we probably release more data each
month than all of our competitors have released in their
history. We are seeding opportunity for entrepreneurs who
want to find novel applications for these data without having
to spend hundreds of thousands of dollars speculatively
to test their ideas.
GRSS: Does running an open data program make
good economic sense?
Capella: Yes. When Capella was founded more than
seven years ago, very high-resolution SAR was a rare commodity
that was available almost exclusively to specialized
governments and research intuitions with large budgets.
The community program is designed to give very high-resolution,
ultralow latency data directly to researchers, nonprofits,
developers, and disaster response organizations to
foster innovation and discover the next game-changing applications
of SAR.
Today, the community program, which includes our
FIGURE 2. From Capella Space Corporation SAR: image of the
Muruntau gold deposit, which is situated in the Qizilqum Desert of
Uzbekistan. It is being mined in the world's largest open-pit gold
mine with production believed to be of the order of two million
ounces per annum.
172
open data offering, has proved mutually beneficial for
both Capella and our research community, as these users
uncover new use cases and applications for high-quality,
high-resolution SAR imagery. Already we have had early
adopters reach out expressing interest in commercializing
automated detection algorithms and machine learning
IEEE GEOSCIENCE AND REMOTE SENSING MAGAZINE DECEMBER 2023
IEEE Geoscience and Remote Sensing Magazine - December 2023

Table of Contents for the Digital Edition of IEEE Geoscience and Remote Sensing Magazine - December 2023
