Andrew Peterson’s founding story begins with two books. A colleague handed him one about cows and another about startups: Peter Thiel’s Zero to One. Peterson moved past the bovine volume and lingered over the contrarian question in the second book. What important truth did he know that few people agreed with? After years inside aerospace and radar programs, he had an answer precise enough to require diagrams.
The largest radar satellites were expensive because the industry kept asking a familiar instrument to become larger, sharper and more intricate. Meanwhile, a different set of curves was moving quickly. Launch costs were falling. Small satellites were becoming practical. Storage that once demanded a stack of spinning hard drives could fit near a thumbnail. Graphics processors could chew through workloads that earlier space programs could barely accommodate.
Peterson’s answer was to recast the instrument as a distributed system. Put several radar satellites into a coordinated formation. Have them look at the same place from different angles at the same moment. Fuse those observations into three-dimensional information. Replace one exquisite machine with a fleet whose members become more useful together.
“I was happy being a technical guy forever. I never thought I’d start a company. I still don’t introduce myself as an entrepreneur. But then I had an idea that was way too good.”Andrew Peterson
A career spent inside the signal
Peterson studied aerospace engineering at Cal Poly Pomona, then earned a master’s degree in the same field at San Diego State. His professional path crossed sensing, control and computation. At General Atomics Aeronautical Systems, he worked on synthetic-aperture radar image formation and guidance, navigation and control. At Moog’s space and defense group, he worked on beam-control systems connected to the Vera C. Rubin Observatory, the large telescope built to survey the southern sky.
The projects look different from the outside: radar images, defense control systems, a telescope. The shared craft is getting a complicated physical system to point, coordinate and turn faint signals into useful information. Array Labs, which Peterson started with Isaac Robledo in 2021, gathered those threads into one company.
The co-founder pairing gave the company a useful split screen. Peterson brought the technical conviction. Robledo came from finance, real estate analysis and business development. Array’s first angel investor was Brian McClendon, one of the people behind Google Earth. It was a connection with narrative symmetry: a builder of the familiar digital globe backing an attempt to make the globe dimensional and frequently refreshed.
The geometry of looking together
Conventional Earth imagery often gives a clean view from above. Peterson likes to point out the limitation with everyday language: there are many pictures of rooftops and the tops of people’s heads, but people do not live on a flat plane. Terrain, buildings, mines, bridges and vehicles occupy volume. A useful model of the physical world needs depth.
Radar brings its own advantage. It does not need daylight, and cloud cover does not close the curtain. The hard part is scale. A larger antenna becomes punishingly difficult and expensive. Peterson’s proposed escape is the same one computing used when clock speeds stopped giving easy gains: parallelize the work. In his analogy, the traditional satellite behaves like a CPU; the cluster behaves more like a GPU.
Array’s first production cluster is designed around four distributed satellites. Their radar observations are meant to arrive from several positions, giving the processing system the geometry it needs to reconstruct depth. The company is also working on algorithms that recover useful depth from radar phase history that other processes discard. In favorable conditions, that technique can reduce the number of passes needed even before the coordinated fleet is operating.
A parking lot before orbit
Space companies have a feedback problem. A software team can ship in the morning and learn by lunch. A satellite team may wait through procurement, integration and a launch calendar before physics delivers its verdict. Array tried to pull some of that learning back to Earth. Its early team built a large radio-frequency test range, the sort of apparatus that looks like industrial scaffolding crossed with an antenna array. It arrived in a Silicon Valley parking lot and became a way to test image-formation software without waiting for orbit.
Peterson said that building a little in public mattered because the company was attempting something unfamiliar and needed conversations with potential users. The phrasing carries two founder instincts at once. Public work can recruit engineers and partners, but it can also expose a technical team to the untidy vocabulary of demand.
“Our customers have been clear with us: speed and accuracy are the top priorities.”Andrew Peterson
That listening changed the shape of Array Labs. The company began with a sweeping data vision: a high-resolution 3D map of Earth, refreshed often enough to become infrastructure. Government customers also wanted the radar itself. Others wanted complete, sovereign satellite systems. Commercial buyers wanted answers and analytics rather than folders of imagery. Array now describes a business spanning standalone radar payloads, integrated satellite systems and 3D radar data.
One technical core, three paths to a customer
The move is less a pivot than a widening of the delivery mechanism. The difficult technology stays in the center. Customers choose how close they want to stand to the hardware. It also gives Array intermediate markets while the full constellation advances through qualification and launch.
The engineer learns distribution
Peterson once repeated a line that first-time founders care about product while second-time founders care about distribution. He is a first-time founder trying to borrow the second-time lesson. The company has paired with Umbra and Raytheon on 3D radar products, worked with U.S. Air Force programs on formation flight, communications, antennas and image algorithms, and won work tied to DARPA and a U.S. Navy study of space-based airborne moving target indication.
In January 2026, Array announced a $20 million Series A to expand production, hire across engineering and go-to-market roles, finish flight qualification and prepare its formation-flying cluster. Six months later, it announced an additional $21 million strategic round anchored by Mitsubishi Electric. The investment came with a partnership to develop satellite-based maritime and aircraft tracking services for defense and security customers in the Asia-Pacific region.
Those relationships reveal the company Peterson is becoming responsible for. It is no longer a two-person thought experiment about better pictures. It is a manufacturer, a government contractor, a geospatial data supplier and a future satellite operator. Each identity carries a different sales cycle and definition of done.
The personal transition is just as telling. Peterson’s public writing offers an open door to technical founders who are earlier in their journey. In a recent conversation about teams, he emphasized self-awareness, empathy and the ability to work with others - qualities far removed from antenna gain or orbital mechanics. The recovering engineer has not abandoned the engineering. He has expanded the system boundary to include the people.
What the formation is for
The applications have shifted as the business has matured, but the ambition has stayed recognizable. Persistent 3D sensing could monitor industrial sites, measure stockpiles, assess infrastructure, supply ground truth to autonomous systems and help governments track activity across large areas. The shared need is a physical world that changes faster than traditional mapping programs can revisit it.
Peterson’s wager is architectural. When technology changes the cost of coordination, the old optimum stops being optimal. The winning instrument may not be the largest object in orbit. It may be the formation whose pieces are inexpensive enough to manufacture, precise enough to fly together and smart enough to combine their signals.
There is still a great distance between a working ground range and a persistent orbital service. Flight qualification has to become flight. A cluster has to act like a single instrument. Production has to repeat what a lab can do once. Customers have to receive something useful on their own clock. Array Labs has organized its recent capital, partnerships and hiring around those ordinary-sounding verbs: build, qualify, launch, deliver.
That may be the cleanest way to understand Peterson. His company carries a planetary-scale premise, but his method is decomposition. Split the antenna across satellites. Split the market across forms of delivery. Split the technical risk into tests that can happen now and tests that must wait for orbit. Then coordinate the pieces. The engineer who never planned to become an entrepreneur is building a business the same way he wants to build a radar: as a system whose parts become more capable together.