Featured
Technical Deep-Dive: Inversion’s First Mission on Orbit
What is Ray?
Ray is our first fully in-house developed spacecraft and reentry capsule. It was conceived at the very beginning of Inversion to give us as many real-world learnings as possible, as quickly as possible, and as cheaply as possible.
Ray serves as a technical maturation platform — building the heritage, processes, and tools that lay the foundation for Arc, our next-generation vehicle for precision cargo delivery to anywhere on Earth in under an hour.
The Mission Begins
January 14, 2025 was a defining moment for the team at Inversion. It was the day we watched as Ray launched on a Falcon 9 out of Vandenberg Space Force Base. As the rocket lifted off from the California coast, most of the company cheered from our new headquarters. Meanwhile, a small group of operators in Mission Control monitored the livestream, holding their breath for the first critical milestone: clean separation.
The callout came: “Ray separation confirmed”. Ray’s in-house designed separation system deployed cleanly. In fact, the separation was so steady that our attitude-control logged only a single token pulse before declaring the vehicle had fully de-tumbled. With this first proof point in hand, Ray was stable, on-orbit, and operational. Our mission had truly begun.
First Contact & Early Checkout
Four minutes and 21 seconds after deployment the mission control dashboard lit up with state-of-health telemetry on Ray. In that instant the entire avionics, GNC, and ground-software stack, over 90 percent written by Inversion engineers, proved itself on orbit.
With this being a test mission, many of our objectives were validating the functionality and performance of each individual component. What this looked like was extensive time-on-console running through checkouts, procedures, and validations. Over the next couple of weeks we:
- Ran full functionality checks on system telemetry, attitude estimation and control systems, and sensor system hardware.
- Configured Ray into its nominal on-orbit configuration, orienting its solar panels towards the sun and its antennas towards our ground stations.
At this point, Ray had been fully commissioned for its early-mission.
Settling In & Stretching Our Legs
With the fundamentals stable, we began handing more authority to autonomy scripts. This meant that operators didn’t have to be on console at all times, letting our custom ground software infrastructure handle the incoming data and outgoing commands to the vehicle. By February, Ray was flying lights-out most of the time. With more operator availability, we delved deeper into the data being received and continued checking off test objectives.
One test that warranted some cheers in Mission Control was capturing our first image on orbit.
We also began evaluating the performance of various components. Multiple sensors and components saw degraded performance compared to expectations, but the forward-thinking and robustness of Ray meant we were able to handle it in stride.
One of the more bold milestones came next: a planned on-orbit software update — something that is rarely attempted on a maiden flight. Before launch, we had intentionally left room to improve the spacecraft’s core flight software in orbit — an inherently risky move that many satellite operators avoid. But we trusted our architecture. The update went smoothly, and we followed it with multiple subsequent updates, all executed successfully and without issue.
We managed a radiation-induced hardware fault when a pair of memory components failed while passing through the South Atlantic Anomaly. Thanks to thoughtful engineering, we seamlessly switched over to our radiation-tolerant memory — hardware developed in-house from scratch — so smoothly it barely warranted a line in the logbook.
We conducted extensive tests of Ray’s Attitude Control System (ACS), spinning the vehicle up and down to validate the performance of the cold gas thrusters and the guidance, navigation, and control (GNC) algorithms — all developed in-house from scratch.
(top) Shows the duty cycle of the two ACS thrusters being tested: first, spinning Ray up and second, spinning Ray down.
(bottom) Shows the spin rate of the vehicle throughout the test.
At this point, Ray was operating as a fully functional satellite, with all orbital systems performing above expectations. This milestone alone surpasses what most companies aim for on a first mission — but it was only one phase of ours.
The Phasing Burn Attempt
Of course, we launched Ray with the intention to bring it back. With the orbital portion of the vehicle going well, we began preparation for the first of a series of operations for ultimate reentry. First up was a phasing burn. This was a full-scale test of our propulsion system, designed to lower Ray’s orbit and adjust its period so that, days later, it would align precisely with our planned landing zone off the California coast.
In the early hours of March 15th, the vehicle was sent its final burn authorization command and began its autonomous procedure to fire its engine. The vehicle calibrated its precise attitude, slewed to its burn attitude, conducted a propulsion settling pulse of the cold gas thrusters, and began its ignition logic.
It was then that our main anomaly occurred. An over-current event on a Bipolar Junction Transistor used to trigger the spark plug used for igniting our engine’s ignition chamber had shorted. In the moment, this had caused Ray’s propulsion controller to lock up, and Ray began to tumble. Automated aborts and operator commands were able to mitigate the scenario and swiftly recover Ray to its nominal orbital configuration.
Immediately, we stepped into an extensive anomaly investigation. Through numerous ground and on-orbit tests, we were able to narrow the failure mode to a particular unknown behavior caused by our flight power system along with the spark plug being actuated in vacuum. The extent we had to go to replicate this failure mode demonstrates the rarity of a situation like this and was invaluable for informing future propulsion system designs.
We then explored alternative methods to ignite the rocket engine, including some creative solutions to actuate the spark plug at significantly reduced power levels. Despite successfully triggering the ignition system, the energy output wasn’t sufficient to execute the de-orbit burn.
While this meant we couldn’t bring Ray back, the robustness of the spacecraft allowed us to continue collecting valuable data and fully diagnose the issue — turning a missed maneuver into a meaningful learning opportunity.
Orbit Change Maneuvers
Ignition or not, we still had propellant and plenty to still demonstrate and prove. We then conducted two cold-flow delta-V maneuvers to demonstrate Ray’s orbital maneuverability.
First, Ray pointed itself towards a specific portion of the sky. This was determined to be the optimal place to receive data for its startracker to receive a precise attitude estimation.
Armed with this precise understanding of the direction Ray is facing, it then slewed to its burn attitude. Once in place, operators on the ground gave the final authorization command to commence the maneuver. After conducting a brief propellant settling fire of its ACS thrusters, the vehicle began to flow propellant through its main engine. While this was happening, the ACS thrusters were keeping the vehicle stable. Onboard, Ray uses an advanced targeting system to ensure that the engine is cut off at the precise moment to ensure an accurate entry. In this case, it would be to enable an accurate orbital insertion. Once that trigger was met, the engine was shut off. Ray was able to successfully insert itself into a semimajor axis within 4.5 m of its target, validating Inversion’s unique capability of accurate delivery from space.
(middle) Overview of the quaternions showing successful GNC control during the phase of the mission.
(bottom) Specific force observed on the vehicle during the maneuver. The peaks during the burn represent ACS thruster firings to keep the vehicle pointed in the right direction.
To show this was repeatable, we also conducted an orbit raising maneuver with the same overall sequence.
- Orbit-lowering maneuver: Dropped perigee 35 km.
- Orbit-raising maneuver: Added 40 km to perigee and 10 km to apogee.
These maneuvers represented a fraction of the overall propellant onboard our system and validated all of the remaining orbital systems onboard Ray: a fully functional propulsion system, guidance algorithms, cold gas thruster control, and more.
Wrapping the Mission
With every feasible test complete, we have formally closed Ray’s primary mission. The spacecraft remains healthy, we maintain full command and control over the vehicle, and have substantial propellant reserves for future maneuvers should we choose. We will be collecting data on how our systems age, which will help to inform future long durations storage on orbit missions for delivery from space.
Why Ray Matters
Ray was built from day one as a test vehicle to inform the design of our future spacecraft that will make precision delivery from space a reality.
Nearly every system on board was developed in-house by a team of just 25 people. The outcome was a fully functional, maneuverable spacecraft with a reentry capsule — an achievement that places Inversion in rare company. Just as important, Ray’s bill of materials came in well under $1 million, demonstrating our ability to build advanced space systems at low cost.
While we’re disappointed Ray won’t be returning home, the lessons learned are already being applied to the next vehicle on our roadmap: Arc, our next-generation reentry system.
This is only the beginning. The next chapter is already taking shape, and we’re excited to share more soon.
Thank you for following along!
-Austin Briggs
