Formation Flight

December 6, 2023
Palo Alto, CA

Array Labs is developing radar satellite clusters for the first real-time, high-quality 3D Earth model, enhancing affordable 3D data use in AR/XR, defense, climate, and insurance.

How do you build an antenna 500X larger than the world record holder, at 1% of the cost?

Can you you maintain control of a flock of dozens of satellites?

While keeping them all pointing on target?

Finally...can you surf in space?

In today’s update, we’re going to take a deep dive into one of the most innovative, interesting aspects of the Array Labs story: formation flight.


A defining characteristic of Array Labs’ constellation is the distributed nature of our  constellation. 

Instead of trying to build a single, giant antenna, our tiny radar birds fly as a flock. This allows us to create an aperture that is tens of miles in diameter, compared with the current record holder, the Orion SIGINT satellite, which has a diameter of around 300 ft (100m). This formation flight enables our satellites to collect high-resolution information as if we deployed a single dish that’s 500 times greater than the largest spacecraft antenna ever flown - at a tiny fraction of the cost. 

Current space antenna record holder: USA-202 Orion SIGINT satellite, complete with an estimated 100m diameter antenna

Fortunately for us, targeting and maintaining orbits is a well-established field, with companies like Planet and SpaceX using various techniques to hold their satellites in position as they fly overhead.

A simulation of the Starlink Constellation

The above simulation shows you how the birds of the world’s most famous satellite constellation – SpaceX’s Starlink – fly. Next, we’ll show an example a bit closer to home – here’s Planet Labs' simulation of its constellation: 

Satellites from SpaceX and Planet are distributed in large constellations around the globe for maximum coverage, but their orbits are not perfect and must be maintained and adjusted over time. Slight deviations in the upper atmosphere, the magnetic field of the earth and even the sun itself will gradually push satellites off course. 


To adjust a satellite’s orbit, an operator must change the speed of the spacecraft at the right time and in the right place. This is done via some type of propulsion system, which sends out a high-velocity stream of particles in one direction, which, in turn, pushes the satellites in the opposite.

The major differences between different space propulsion systems is in exactly how these high-velocity particles are created. Sometimes pressurized gas is simply released, like air from a balloon (cold-gas thrusters). Sometimes different chemicals are burned, which allow for much greater efficiency and higher thrust (chemical propulsion). For super high efficiency, electric fields can be used to accelerate tiny ions to incredibly high speeds (electric propulsion systems). Each of these technologies can be used (and have been used) to establish and maintain satellites' orbits. 

But there’s a completely different way to find particles in space…  


As you move higher in altitude, air gets thinner. Move from sunny Palo Alto to the heights of the Sierra Nevada mountains, and you’ll notice fewer molecules of air; it's a bit harder to breathe, and you won’t be able to exercise as hard without getting winded. If you somehow were able to stick your head out of the window of a jet airplane flying along at 30,000 feet, you won’t have enough air at that altitude to even remain conscious for long. 

It’s easy to imagine that a bit above this altitude, air simply ‘stops’ - that there simply aren’t any more molecules left - but this isn’t exactly true. Instead, the air just continues to get less and less dense. While your body probably wouldn't notice a difference between being at 100 km (extremely low orbit) and 500 km (‘standard’ low orbit), when you’re a satellite flying along at 17,000 mph (orbital speeds) and bumping into lots and lots of air molecules, there’s a huge difference.

Bumping into all of these particles creates drag, causing satellites to slow down and re-enter the atmosphere, and the difference between 100 km and 500 km altitude is the difference between staying in orbit for hours vs years. 

Many satellite operators would love to live in a world with zero air in LEO altitudes - where things placed into orbit simply keep flying forever. However, these faintest wisps of our planet’s atmosphere present an opportunity.

In fact, some satellite operators have found that they can make use of these particles to change the orbit of their spacecraft. Instead of bringing a supply of high-velocity particles with them into space, they simply make use of the particles which are already there. 

For example, Planet Labs controls their Dove satellites using a purely aerodynamic approach. By changing the way their satellites are oriented, they control the area exposed to the orbital ‘wind’, along with the associated drag. By using this drag control approach, they can keep their satellites exactly where they want them. But this control doesn’t come for free…


Drag control for a satellite is a bit like controlling your car only using the brake pedal. Using drag alone, satellites can slow down by pumping the brakes. While it’s effective, this approach quickly reduces the amount of time the satellite can stay in orbit. 

Of course, drag control is better than no control, but it would be much nicer if we could steer side to side as well. Continuing the car metaphor, staying in formation and not colliding with other drivers is much easier if you can use the steering wheel too. We wondered: 

What would it be like we took a different approach, and steered the spacecraft by creating lift?

We began with detailed simulations of lift and drag generation using advanced modeling and simulation tools like Mathworks’ MATLAB/SIMULINK and A.I. Solutions’ FreeFlyer. 

Using these tools, we rapidly tested new concepts for flight architectures, deployment schemes, and formation maintenance in our quest to find the most efficient, lowest cost way to deploy and control our satellites.

Simulation results showcasing one potential formation’s expansion and contraction over a few orbits

We found that the ability to create lift can increase satellite lifetimes by more than 50% compared to drag-only approaches, while still maintaining the same level of control. In other words, steering is substantially more efficient than just braking.


It’s clear that using the atmosphere to create lift offers huge improvements, so why hasn’t it been done before? 

Well, it hasn’t been possible, and the reason has to do with chemistry. 

You see, the upper atmosphere doesn’t just have fewer air particles than the air we have at sea level, it also has a completely different composition. Unlike sea level air, the upper atmosphere is mostly made-up of atomic oxygen (O), the incredibly reactive molecule which is created when UV light from the sun breaks typical Oxygen molecules (O2) apart.

Atomic Oxygen (also known as a free radical) wants to bond with any compound it can, and when a spacecraft comes by at 17,000 mph the atomic oxygen slams into the surface, chemically bonds to whatever it can, and sticks like glue. Not only do these reactions physically pit and damage the surfaces of the spacecraft like a cosmic sandblaster, the sticky, oxidizing particles slow the craft down, creating lots of drag. Atomic oxygen is sticky and destructive. This has always been the problem with generating lift in space - if the particles bounce off, then you get lift, but if they don’t  - all you get is drag. 

To achieve lift generation we needed to find a material that can survive the onslaught of orbital atomic oxygen and provide this ricochet for the entire duration of the mission. For a long time, the space community has looked into ways to counter this persistent menace – and shield the space assets we’ve put sweat, tears, and literally years into building, deploying, and flying. 

The destructive power of the fast free radical is its two for one punch that wreaks havoc through high energy impact and oxidation, chipping away at Kapton tape thermal blankets, fragile optical coatings, and solar panels. Over time, glass coverings become so clouded that light can no longer pass through, destroying solar output. Materials literally disintegrate, with particulates contaminating sensitive instrumentation like antennas.

So, we decided to look into whether we could do things differently…by mitigating degradation from the cosmic sandblaster, and reflecting AO strikes, rather than accepting them as a way of life in LEO. To do this we needed a material that could protect and reflect. 


By rethinking spacecraft design from scratch, soup to nuts, you could spend years toiling in the lab, spinning your wheels, and breaking the bank. 

But, by rethinking specific elements of spacecraft design from scratch, ones where a new subsystem design or material could lead to an order-of-magnitude improvement in cost savings, you could potentially overcome once-intractable problems with physics-first solutions.

Here, we’ll present our first update into the work that we’ve been doing on the cosmic sandblaster front. 

Earlier in 2023, we started working from first principles and asked whether atomic oxygen could be a feature, rather than a bug, for formation flying. 

And this summer, we partnered with a university and their state-of-the-art aerospace environmental testing facility, to test a novel satellite coating that could mitigate the effects of atomic oxygen bombardment in LEO. We started by focusing on the protective qualities first, if we were going to use this coating to aid in our formation flight for our mission it would have to survive. 

After a preliminary technical study, we identified a coating that—on paper—fit the bill for a protective spacecraft armor perfectly. 

We won’t name the material, but for our purposes here, let’s call it Material X – it exists outside the lab, as an ultra-thin, ultra-tough castoff originally deemed too pricey for earthly uses. Material X has since been taken up for some terrestrial applications, such as manufacturing semiconductors. Material X has one potentially huge cosmic advantage: Array’s engineers and our university partner suspect that its properties allow it to reflect atomic oxygen, rather than absorbing the bombardment. 

We’re pleased to say that the results from our first space weathering/sandblasting experiment with Material X look extremely promising. We tested the coating on two common spacecraft materials, Kapton Tape and Solar Panels, and we have demonstrated dramatically improved atomic oxygen resistance in both cases.

Figure 1: Test of coated Kapton Tape in 10 year exposure test

What’d we test? We tested our coating on Kapton tape, a ubiquitous miracle material for spacecraft. This polyamide film can endure extreme thermal swings from -269°F to 400°F under a harsh solar glare. So, Kapton effectively helps hold satellites together despite space’s profound hostility.

To simulate a decade in low Earth orbit, we blasted two Kapton strips with atomic oxygen. One we left vulnerable. The other we covered with a thin layer of Array Labs’ custom Material X coating. 

Ten years of bombardment in 14 hours – let the sandblasting commence!

Kapton tape samples undergoing atomic oxygen testing

What’d we find? When we removed the two samples from the test chamber, it was a tale of two tapes. Catastrophe struck one sample, with atomic oxygen completely obliterating the uncoated Kapton, and calamity seemingly averted for its X-coated twin. The armored strip also has had precisely zero measurable weight loss (which is an effective proxy for subtle surface erosion). 


Kapton was merely Phase One for our experiment. The real test would be integrating a coating into spacecrafts' workhorse: the solar panel. 

Power generation fails without functioning photovoltaics, and solar panels make up a good portion of a satellite. 

  • The panels themselves consist of a wafer layer sandwiched between two layers of thin glass. 
  • Known as cover glass, the outward-facing layer is what actually interacts with atomic oxygen (and suffers the sandblasting). 
  • Degradation to the glass cripples otherwise healthy solar cells. Clouding glass blocks photons from penetrating to the wafer layer sandwiched underneath, progressively throttling electrical output over time.

In Phase 2 of this experiment, we again prepared a thin coating of Material X. We needed to understand how the coating impacts light transmission, which directly correlates to the wattage we can pull from our solar panels. 

Too much efficiency loss, and we’d need to make drastic satellite reconfigurations to utilize our formation flight techniques.

Fortunately, initial tests suggest the coating actually increases solar efficiency in low Earth orbit rather than hampering it. This net positive outcome is encouraging as we evaluate applying the technology to spacecraft.

Without disclosing details of Material X (we’re holding that close to the chest), let’s survey the results from our light pass-through test:

We found that the coating has an initial 5% negative impact on the beginning of life transmission of energy. However, after exposure to the atomic oxygen, the uncoated glass degrades significantly, with its transmission percentage dropping well below the coated glass, especially for the higher energy shorter wavelengths! As expected from our kapton tape, the coated sample showed no change at all before and after the exposure. 

Uncoated Coverglass (Top) vs Coated Coverglass (Bottom) after AO Exposure,
Note the brown discoloration of the uncoated sample.

What's the conclusion from our experiments with Material X? Our experimental atomic oxygen-proof spacecraft armor is working on both kapton tape and solar panel coverglass. We have successfully proved the material has the right protective qualities and this means that we may be able to create usable lift if the right reflective qualities can be demonstrated as well. We’ll save the discussion of how we can prove that for a later update, but as a hint sometimes it's better to just go out and test your theories in the field.


It’s all fine and dandy for us to pontificate on the enormous potential that *Array* sees in Material X, but what about everyone else? Or anyone else? 

Well, it turns out we’re not the only ones who think this is a good idea. One institution, the United States government, has taken an interest in our research and its implications. 

To that end, we’re pleased to announce that last week, Array Labs was selected and funded for a SBIR contract from the US Air Force. Specifically the awarding entity was AFRL (or, the US Air Force Research Laboratory), supporting our research into “Novel Protective Coatings that can improve satellite aerodynamic performance."

This is very, very fresh off the press – and we’ll know more on next steps in the award process in the coming weeks. But we’re pleased with these results, as well as the pathways into bigger, meatier Phase II SBIR contracts that this opens up. 

Oh, and one more thing. We may have buried the lead here, but we’re very excited to announce that we’ll be launching these satellite coatings to space next year! X, our first orbital demonstrator mission, is set to launch with SpaceX in the first six months of 2024. 

For now, minds and models race ahead of reality. This is no surefire bet and more work remains to be done. But we’re highly confident that these protective coatings will be a force multiplier for formation flight, and consequently, a key enabler of our mission and vision. We’ll report back in due time once we have results – and space heritage – under our belts. 

In the meantime, stay tuned for more on our ground-based technical derisking efforts in Q1.  That will do it for Field Note #4. Thanks for sticking with us, and we’ll see you back here soon…