Rendezvous Robotics raises $20M to build self-assembling spacecraft

In 2022, a handful of small tiles floated inside the International Space Station. They nudged, corrected, and clicked into place. The experiment fit inside a locker, yet it answered a question Ariel Ekblaw had carried since her student days at MIT: could a spacecraft assemble itself in orbit?

Ekblaw had sketched the idea as a Buckyball on her desk. She named it TESSERAE, after the mosaic tiles of antiquity. The first versions were small enough to hold in two hands, built to prove that magnets and software could replace a robotic arm.

On October 28, 2026, Rendezvous Robotics, the Golden, Colorado company she co-founded to commercialize the idea, closed a $20 million seed round. The money is not for a finished satellite. It is for a seven-tile demonstration that will try to assemble a spacecraft in open orbit.

A spacecraft that arrives as a stack of tiles

The company's answer to launch limits is a modular tile about one metre across. Each hexagon carries sensors, a processor, electromagnetic actuators, a mechanical latch, and power and data interconnects. A stack of tiles fits inside a rocket fairing. Once released, the tiles find each other, align, and lock into a larger structure.

Every tile is a spacecraft in miniature. That is the design bet. The tile that carries the antenna is the same tile that carries the computer and the latch, so a mission can grow by adding area instead of by designing a bigger bus.

Rendezvous did not invent the physics. It is commercializing a decade of work that began at MIT and continued through the Aurelia Institute. Earlier tile generations flew on Blue Origin's New Shepard in 2019, on a SpaceX cargo mission to the ISS in 2020, and on the Ax-1 private mission in 2022.

Pole A: the fairing stops being the boundary

The case for self-assembly starts with a mismatch. Satellite operators want football-field solar arrays, kilometre-scale antennas, megawatt power, and more heat rejection. Rockets still impose a fixed envelope.

Joe Landon, Rendezvous co-founder and CEO, puts the demand side plainly.

Nobody's talking about making satellites smaller and less capable. They're talking about more capable satellites, more power, more heat dissipation, bigger antennas, more sensitive communication.— Joe Landon, co-founder and CEO, Rendezvous Robotics

Deployable structures, the origami approach, squeeze more surface into the same fairing. Self-assembly changes the unit of design. A satellite becomes a swarm of standard tiles that connect after launch. Operators can add tiles later, replace a damaged section, or reconfigure the array for a new mission.

Landon argues the model could go further, with tiles staged in orbit as a depot that dispenses replacement parts on demand. That vision is not part of the 2028 demonstration. It is the reason investors are willing to fund a hardware company with a long road to revenue.

Pole B: the last centimetre is the hard part

The ISS demonstrations happened inside a controlled experiment enclosure. Free space is less forgiving. Tiles must find each other without a station wall, a crew, or an external robotic arm holding the pieces steady.

Rendezvous plans to send seven tiles to low Earth orbit no earlier than April 2028. The tiles will separate from a rideshare, manoeuvre, align, latch, and form one functional spacecraft. Then they will separate and assemble into a different configuration.

Seven tiles is a deliberately modest number. A failure in alignment, latching, or power transfer shows up early, while the hardware is still cheap to change.

The next ISS test is larger in tile count but narrower in ambition. In December 2026, 32 fifth-generation tiles are scheduled to assemble into a three-dimensional structure inside the station. The Gen-5 hardware carries 192 cameras and 384 electropermanent magnets. Cameras check the bond state. Magnets do the grabbing.

That enclosure test matters because it isolates one question: can the control software close the loop on hundreds of bonds in microgravity? The 2028 free-flyer adds the questions the enclosure cannot answer, including thermal cycling, vacuum, and orbital rendezvous.

None of this is routine. Autonomous docking has a short list of commercial successes and a long list of slips. Rendezvous is proposing it without an external robotic arm and without a crew. The magnets and the software have to do the work.

The trade-off hiding in the tile count

Modularity is not free. Every additional tile adds a latch, a power junction, a data link, and a control problem. A fifty-tile satellite has more interfaces than a monolithic bus with the same mass. Interfaces are where thermal gradients, vibration, and software faults concentrate.

Astro-mechanical engineers will recognise the argument. The industry has spent decades optimising hinges, deployment mechanisms, and one-shot latches. Self-assembly replaces those mechanisms with software and magnets, and moves the complexity into the control loop.

Rendezvous is pitching the opposite trade. A larger number of small, identical units gives redundancy and replaceability. A failed tile can be isolated or swapped. A failed monolithic subsystem can end the mission.

The tile count in each test maps to that trade. Seven tiles in free flight is the smallest configuration that can prove assembly, separation, and reassembly without becoming a swarm-management experiment. Thirty-two tiles inside the ISS is large enough to stress the bonding logic while keeping the experiment inside a controlled enclosure.

That sequencing is sensible for a company with 22 employees and a first priced round. It also leaves the hardest question for last: whether the economics improve when the tile count rises. A depot with hundreds of tiles needs a production line, a test line, and a launch cadence that do not exist yet.

For now, the engineering claim is narrow. A spacecraft can be built from identical pieces that find each other. The commercial claim is broader. If the pieces work, the fairing is no longer the boundary.

The $20M buys a date, not a product

The round was co-led by MaC Venture Capital and Toyota Ventures. The University of Michigan Endowment, Astera, Industrious Ventures, Aurelia Foundry, ATX Venture Partners, and Mana Ventures also participated. Rendezvous describes it as its first priced round, after a $3 million pre-seed in 2025.

The company has 22 employees, according to its LinkedIn page. The money is earmarked for larger tiles, ground demonstrations, the November 2026 parabolic flight campaign, and team expansion.

Rendezvous is not alone in orbital infrastructure. Starfish Space launched its first Otter servicing vehicle on October 1, 2026, and is building a pipeline of satellite inspection and disposal missions. Its business is the repair and retirement of existing spacecraft. Rendezvous is betting that the next generation will be assembled from modules in the first place.

The difference matters for the sales pitch. A servicing company sells an intervention after a satellite is already designed. A modular assembly company asks the satellite designer to change the architecture before launch. The second sale is harder, and the payoff is larger if it works.

What April 2028 has to prove

Three results would validate the pitch. The tiles must find each other without ground intervention. The latch must survive the load path and carry power and data. The assembled spacecraft must separate and reassemble into a different shape.

Each result is testable. Each can fail in public. That is the point of a seven-tile demo instead of a paper design.

The alternative is not a single competitor. It is the familiar engineering answer: fold more capability into the fairing, accept the size limit, and launch a bigger rocket. That path has decades of flight history behind it.

Rendezvous is asking investors to accept a longer timeline in exchange for a different ceiling. The $20 million does not buy a constellation. It buys the chance to prove that a spacecraft can arrive in pieces and become one machine in orbit.

If the tiles can do that, the fairing stops defining what a satellite can be.