By Abbey White, Staff Writer, SatNews
Dispatch from SmallSat Symposium. Coverage and analysis from across the conference, tracking the forces shaping the next phase of the SmallSat market.
MOUNTAIN VIEW. The polite fiction that the commercial space industry exists primarily for crop yields and climate monitoring has ended. During the symposium’s Future of Geospatial Intelligence session, industry titans stopped pretending their product is intended for sale to NGOs, admitting they are instead building a new global backbone for national defense.
For years, the future of Earth observation was said to lie in commoditized data for hedge funds and farmers. That thesis is disproven. The real money ($10.4 billion flooding into the sector over the last year) is chasing hardware-heavy infrastructure designed for Great Power Competition. The session’s panelists made it clear we are no longer discussing a commercial marketplace. We are looking instead at a militarized mesh network where latency or time-lag is the enemy and the customer is almost exclusively the state.
The 99% Reality
Dr. Marino Fragnito, Senior Vice President at Thales Alenia Space, minced no words concerning revenue’s true source. Though his company builds massive Copernicus sentinels for European science, he was blunt about the infrastructure they deliver to export clients.
“When we deliver our space infrastructure to a customer, I have to say that 99% of the need comes from a defense application,” Fragnito told the audience. “All the rest is collateral.”
Fragnito is saying the quiet part said out loud. The dual-use narrative has collapsed into a single-use reality with commercial benefits serving merely as a side dish. Anxiety, not curiosity drives the market. Nations no longer want to buy subscriptions to American commercial constellations. They want what the industry calls sovereign capabilities: the ability to own the satellite, control the shutter, and keep their data off the cloud.
The Rise of the Middle Powers
The most striking takeaway from the session was the fragmentation of the global commons. The world where everyone shares the same commercial picture of Earth is giving place to another world where every middle-tier power demands its own private glass in orbit.
Fragnito noted that this trend is sweeping across Europe and Asia.
“We recently signed a contract with Indonesia,” he said. “In Asia Pacific all the countries around Indonesia are looking for sovereign capabilities in Earth observation now.”
These customers, he explained, explicitly reject the shared-infrastructure models that defined the last decade.
“They don’t want a cloud-based system. They don’t want to rely on somebody else. They want their own ground infrastructure.”
This shift to sovereignty complicates the business model for pure-play analytics firms. If every nation builds its own vertical stack to avoid reliance on the US, the global market fractures. For hardware primes like Thales, however, the shift is a bonanza of opportunities.
The Trust Gap in the Kill Chain
While hardware manufacturers sell sovereignty, analytics providers are attempting to sell speed without sacrificing trust. The conference buzzword is tipping and cueing, the automated process by which a wide-area sensor detects a target and instantly triggers a high-resolution camera to identify it.
Automation, however, brings a new terror: AI hallucination. A false positive in a kinetic kill chain isn’t just a glitch; it turns into wasted munition or, far worse, a diplomatic fiasco. Peter Kant, CEO of Enabled Intelligence, argued that human analysts are not going extinct but their role is shifting—from counting cars to policing algorithms.
“One of the added features of a human analyst is, typically, when you got a picture, an EO picture or a SAR picture, you weren’t worried it was lying to you,” Kant said. “Now the human analyst, because they’re using AI tools, is trying to understand, okay, what did the AI tool tell me and where is it wrong?”
This creates a paradox. The speed of war demands automation, but the risk of war demands verification. Kant emphasized that while the industry is awash in bounding boxes, the need for trusted, human-verified labels has only grown. His firm’s recent $708 million contract with the NGA [National Geospatial-Intelligence Agency] confirms this sentiment.
Latency is the Enemy
The final nail in the coffin for the commercial narrative is the industry’s obsession with speed. A farmer does not need to know the health of their corn within five minutes. A theater commander, however, needs to know a mobile missile launcher’s location in seconds.
Dan Adams, Head of KSAT USA, when asked about the biggest unmet need in imaging, didn’t request better resolution or more spectral bands. Rather, he summarized the entire technical roadmap in one word.
“Latency. Just latency,” Adams said. “A lot of data needs to come down faster. Move faster.”
This drive for speed is pushing compute to the edge. The days of downloading terabytes of raw data to a ground station in Svalbard to be processed days later are ending. The future requires processing on the satellite itself, filtering out the clouds, and sending only the answer down.
The Verdict
The industry gathering in Mountain View this week is healthier than it has been in years, but it is also harder. The speculative capital that chased balloon internet and asteroid mining has dried up, replaced by defense dollars demanding hardware that works, answers that are verified, and data that stays inside sovereign borders.
As James Crawford of Privateer noted during the panel, the geopolitical reality has forced the industry’s hand.
“The amount of their revenue that’s coming from sales of hardware—you know, entire satellites—indicates the trend very clearly.”
Now grown up, the space industry has enlisted. The end of the commercial revolution in orbit marks the start of its strategic industrialization.
