Is Space just cool, cold, and costly?

by JMill. March 15, 2024 | Space for Earthlings 🌏🛰️

Space is ‘cool’. The proliferation of NASA-branded swag, space-themed science-fiction media, and adulation of Space-faring heroes such as the Apollo Mission teams shows the general public’s overall appetite for thinking about Space. (This is a typical U.S.-centric perspective.) Space has indeed maintained a ‘coolness’ factor for decades, if not millennia: as Scarlett Koller of Mithril Technologies stated when I interviewed her on Tough Tech Today, “I actually think it’s very human, really there isn’t a culture that we know of that hasn’t named the stars that they could see.” 

In regard to marketable selling points, enhancing brand reputation through Space involvement might not resonate with all stakeholders, as some might view it as an unnecessary diversion from core business operations. It could be argued that BP‘s involvement in space solar power projects does not mitigate the environmental impact of its oil and gas business, nor is Budweiser’s “space beer” more than a marketing piece. 

Space is cold. Most of us know that Space is literally mostly a frozen, desolate vacuum, interspersed with blazing hot balls of rock and gas every few parsecs, plus or minus a lightyear. Our cozy part of the universe is anomalous (but it is not the only), and most everywhere else is a brutal environment that is challenging for even our species’ best autonomous robots.

Space is also costly – but unit economics are changing favorably. The public had been trained for a half-century that Space is for nation-states and unlimited budgets. In the late 2000s and early 2010s, that perception began to evolve as several billionaires, including Elon Musk, Jeff Bezos, and Richard Branson, poured their personal wealth into launch services. However, in the 2020s, we see a clear transition to venture capital investment. As Ashlee Vance, author of When the Heavens Went On Sale notes, “Trying out an idea in space no longer requires congressional approval or some wild-eyed dreamer willing to risk his personal fortunes; it just requires a couple of people in a room agreeing that they’re willing to spend someone else’s money on a huge risk.”(When the Heavens Went On Sale by Vance. HarperCollins. 2023. Page 17.) “In 2021, private capital [invested] more money in the space industry than NASA will spend on everything,” stated Shanti Rao, NASA-JPL physicist and startup investor, and based on data sourced from Space Capital’s industry reporting.

Lets focus more on costs, since there are several and each is a key consideration for senior business leaders.

A drop in launching cost spurs an increase in launch

What happens when it costs less to launch goods around the planet on a SpaceX Starship rocket compared to the status quo of airlifting via Lockheed Martin KC-130 cargo plane? As the cost to launch a heavy object into low-earth orbit has decreased by several orders of magnitude, it has implications not only on how much ‘stuff’ we can launch to Space, but what also can return back to Earth. It cost $100,000 per kilogram to ride the U.S. Space Shuttle in 1980, or $1,000 per kilogram to ride SpaceX’s Falcon Heavy in 2020. Starship, with anticipated operation beginning in 2025, may achieve a price point of just $100 per kilogram. Launch cost is an important instigator of space development in much the way that low-cost access to large bodies of water such as navigable rivers helped port cities to grow on the top of lowered transportation costs. (For a deeper look at this, check out Tomas Pueyo’s thread.

Hidden costs of orbiting data streams

GeoInt, specifically Earth Observation, is among the fastest growing segments serving terrestrial businesses like financial services, mining, logistics, et cetera. Companies providing capabilities for earth sensing tend to focus on the cost of data. While important, it is not the only cost for which terrestrial leaders must account. 

When pricing a Space-origin data stream, Earthlings should probe:

• Cost of data. This would be the dollar-per-square-kilometer cost for a multispectral image. Planet Labs, an early provider of imagery data, provides subscription services including daily global coverage at a resolution of multiple meters per image pixel and rapid taskable imagery that is a higher resolution of deca-centimeters per image pixel. See below for a visual comparison provided by Maxar of differing image resolutions.

• Cost of compute. Once a dataset has been retrieved, there are often large computational needs to analyze the data for whatever the organization cares about. For example, it is common to want to look for new or anomalous objects, or to count or measure certain objects while ignoring others. Machines are great for this task, but powering these processing systems often requires dedicated and tuned computational infrastructure. The Utilis company may provide a plot of water leak locations to utility managers once a quarter or twice a year.
• Cost of human resources. If ‘off-the-shelf’ products are unavailable but an Earthling organization really wants a particular capability, the development must occur in-house or through partnership. Both approaches require redirecting internal human resources and, likely, increasing hiring for Space specialists. This is typically a very expensive and strategic endeavor before any return on investment could be recouped.
• Cost of change. An organization’s cultural norms can make it difficult to strategically evolve the organization into a new market position. While value-based evaluation of if and how to incorporate Space capabilities into the organization is probably the best metric from a shareholder’s perspective, from the perspective of the rank-and-file employees there can be expected skepticism and reluctance toward such a shift.

Beware revenue model (mis)alignment. The Space industry mostly runs on a project-based model with customers paying for Earth Observation capabilities on-demand, which is usually when no other option remains for the customer. Earth observation is often priced per square kilometer or “per scene”. This does not scale well when an organization needs to construct a time-series analysis over a wide-spanning area. A better alignment of incentives is to provide a subscription model with customers relying on space-based resources on a recurring basis, which makes the capability strategic to the (non-Space) organizations. For most companies, the adoption of geospatial data is not an inevitability. Having the technology available does not make it ready to be adopted. As stated by Aravind Ravichandran of TerraWatch Space, a consulting firm focused on earth observation applications, “The adoption chasm is real – as weird as it sounds, launching satellites and beaming the data down does not seem like the hardest bit. …It is super important to compare the relative benefits of what we bring to the table instead of just offering something that is ‘cool’ because data comes from space!” Noted earlier, there can be a cost to change, which can manifest as employee reluctance to adopt what an organization has strategically identified as a good value investment in Space capability. When a new Space initiative is financially modeled and shows opportunities to “free up a lot of [the employee’s] time and that person can now go and do something else or so the productivity of the person improves,” Ravichandran of TerraWatch described to me, “that’s more of an organization perspective” rather than accounting for concerns from an employee’s point-of-view.

Just because the data is Space-origin doesn’t make it the best or defensible for a business. For some forms of earth observation, sometimes it is easier (and lower cost) to simply hire a human to fly an airplane overhead or to contract someone to stand on a street corner and count cars. As a case example on this approach, see the Radiolab episode Eye in the Sky for a story on tracking crimes with airplanes. Here is an excerpt: “In 2004, when casualties in Iraq were rising due to roadside bombs, Ross McNutt and his team came up with an idea. With a small plane and a 44 mega-pixel camera, they figured out how to watch an entire city all at once, all day long. Whenever a bomb detonated, they could zoom onto that spot and then, because this eye in the sky had been there all along, they could scroll back in time and see – literally see – who planted it.”

Thus, persistent surveillance is feasible with satellite systems but also with other modalities. See this Washington Post article for an example. Companies like Fraym offer services to synthesize multi-modal approaches so that their customers can avoid mucking through this form of data analysis.

Earth observation data sets costs a lot if one wants to monitor a large area often. Pricing is usually dollars-per-square-kilometer. The surveying cadence matters, as costs grow quickly if an organization needs hourly or daily snapshots rather than monthly or quarterly. As the headline of a 2022 piece by BCG states, “Satellites Are the Next Frontier for Industrial Companies”, though the piece was not intended to delve substantively into the nuanced cost characteristics of such data sources. Data sources that are newer, smaller, more numerous and more expendable in low earth orbit (LEO) and medium earth orbit (MEO) are reducing the cost curve compared to geostationary earth orbit (GEO) systems.

𝝭