The Satellite That Fits in a Shoebox — and Images the Entire Earth
The Satellite That Fits
in a Shoebox — and Images the Entire Earth
A CubeSat started as a grad school project at Cal Poly in 1999. Today it’s a $22 billion industry photographing every square kilometer of Earth, every single day. Not bad for something that weighs 1.3 kg.
what if we made a satellite the size of a soda can — and made it a standard?
Twenty-five years later, that standard runs the fastest-growing segment of the entire space industry.
Let’s be honest: when most people picture a satellite, they imagine something the size of a school bus, wrapped in gold foil, built by aerospace engineers over a decade, costing half a billion dollars. That image isn’t wrong — it’s just increasingly outdated.
The CubeSat — a 10×10×10 cm unit weighing 1.33 kg — democratized access to orbit in a way that nobody fully predicted. What started as a teaching tool for engineering students evolved into the backbone of a new space economy. And in 2026, that economy is starting to look very real indeed.
2026–2035 (Novaspace)
at 12.7% CAGR
fastest growing in space
From Dorm Room to Daily Earth Imaging
The CubeSat standard was never supposed to be a commercial breakthrough. Professors Jordi Puig-Suari and Bob Twiggs created it so that engineering students could actually build — and fly — a real satellite as part of their coursework. The genius was in the constraints: fixed dimensions, standardized interfaces, deployer compatibility. Constraints, it turns out, are often the mother of entire industries.
The first commercial pivot came when someone realized: if a university can put a CubeSat in orbit for under $1 million, what could a well-funded startup do with a hundred of them? That someone, more or less, was Robbie Schingler and Will Marshall at Planet Labs. In 2010, they left NASA — which, yes, takes a certain audacity — and started building what would become the world’s largest Earth observation constellation.
CubeSats didn’t become commercially viable in a vacuum. Three things happened simultaneously that made the whole thing work:
| Revolution | What Changed | Why It Mattered for CubeSats |
|---|---|---|
| Smartphone components | Miniaturized cameras, sensors, processors went mainstream | Satellite-grade hardware at consumer prices |
| Rideshare launches | SpaceX Transporter, ISRO PSLV made orbit accessible | $5,000/kg down to under $1,000/kg |
| Cloud computing | AWS, Google Cloud made processing petabytes trivial | Constellation data actually became usable |
Take any one of those away and the CubeSat industry looks very different. Put them all together and suddenly you have Planet Labs imaging every square kilometer of Earth’s surface, every day, and selling that data to governments, farmers, hedge funds, and militaries — all from a constellation of satellites that would have been science fiction in 2005.
What CubeSats Are Actually Good At
Here’s something that surprises people: CubeSats aren’t trying to replace big satellites. They’re doing things that big satellites genuinely cannot do. The key is in the word constellation.
A single traditional Earth observation satellite passes over any given point on Earth roughly once every few days. That’s fine for mapping. It’s useless for monitoring. If you want to know whether a North Korean missile facility is being expanded, whether a Brazilian farm is being illegally cleared, or whether a shipping route is congested right now — you need daily, sometimes hourly, coverage. That requires many satellites spread across many orbits. Which is exactly what a constellation of 200 CubeSats provides, at a fraction of the cost of a single traditional bird.
By 2034, over 35% of newly manufactured CubeSats will have onboard AI processing — up from under 8% in 2025. The satellite isn’t just a camera anymore. It’s a flying edge computing node. Data gets analyzed before it ever hits the ground.
Planet Labs Just Proved the Business Model Works
For years, the polite skeptic’s question about the small satellite industry was: "Sure, but does it make money?" Fair question. Planet Labs was burning cash for years, Spire Global kept missing estimates, and the whole sector felt perpetually on the edge of "interesting technology, unclear economics."
Then March 2026 happened. Planet Labs reported its fiscal year 2026 results — and for the first time, it posted adjusted EBITDA profitability. Full year revenue of $307.7 million, up 26% year-over-year. Backlog of $900 million, up 79%. The company had finally crossed the line that every space startup talks about and very few reach.
The kicker? The growth was driven by government and defense contracts — Planet signed a deal to deliver sovereign satellite capability to Sweden in just four months from contract signing. That’s not a science project. That’s a product.
GAAP net loss for Planet Labs FY2026 was still $(246.9) million — largely due to a $161 million non-cash warrant revaluation. The path to full GAAP profitability is still ahead. And Spire Global is at $75–85M revenue guidance for 2026, still running operating losses. The business model works; the financials are still maturing.
The Ecosystem Building Around CubeSats
The Honest Limits — What CubeSats Still Can’t Do
Let’s not get too carried away. CubeSats are remarkable, but they have real physical constraints that aren’t going away soon.
A CubeSat’s small aperture limits optical resolution. Planet’s standard Dove satellites deliver about 3–5 meter resolution — good enough to detect a ship, not good enough to read a license plate. Their newer Pelican satellites push toward 30 cm, but at that point you’re adding mass and cost that starts to look more like a traditional small satellite than a true CubeSat.
A standard 1U CubeSat generates about 2 watts of power from its solar panels. That sounds fine until you realize a decent radar transmitter needs 50–100 watts. This is why SAR CubeSats — like ICEYE ’s — are technically 85 kg microsatellites, not true CubeSats. Power physics doesn’t care about your standardized form factor.
Traditional GEO satellites are designed for 15–20 year lifespans. CubeSats in LEO typically last 3–5 years before orbital decay. That’s not necessarily a problem — it’s actually a feature in some ways, because it forces regular constellation refresh with newer technology. But it means the business model requires constant manufacturing and launch cadence. You’re not buying infrastructure; you’re subscribing to it.
The CubeSat revolution is no longer a revolution — it’s an industry. Planet Labs’ first profitable year, 16,900 small satellites projected for the next decade, and defense ministries around the world building sovereign constellations: the numbers speak for themselves. What started as a grad school teaching tool is now imaging crop failures in real time, tracking illegal fishing fleets, and providing governments with intelligence that used to require billion-dollar programs. The next phase isn’t about proving it works — it’s about who can execute at scale. And the answer to that question, as the Novaspace report put it, “will determine who succeeds.”
Comments
Post a Comment