As space-based research and manufacturing ramp up, a market is emerging for returning payloads to Earth affordably and reliably. German startup ATMOS Space Cargo is meeting that need with a new generation of reentry vehicles built around an inflatable heat shield.
Already flight-tested, the novel system positions Europe as a serious contender in the field of orbital logistics. NSF spoke to Sebastian Klaus, the founder & CEO of ATMOS Space Cargo, about surviving reentry and why Europe’s next big space breakthrough might be on the way back down — not up.
The technology builds upon NASA’s Low-Earth Orbit Flight Test of an Inflatable Decelerator (LOFTID), which successfully demonstrated its technology in November 2022. The mission aimed to test a new type of aeroshell, designed to slow down spacecraft during atmospheric entry by utilizing its large surface area to generate drag. LOFID was itself part of the agency’s Hypersonic Inflatable Aerodynamic Decelerator (HIAD) project, which flew a series of suborbital tests starting in 2009.
The project sought to develop lighter and more compact inflatable heat shields that might be used for lunar cargo return or for landing heavy payloads on Mars, in addition to the more immediate benefits of low-cost Earth return missions. Having proven that these inflatable shields could deploy and maintain their shape, the tests went on to validate their structural integrity during hypersonic reentry and prove that the concept could survive real-world atmospheric heating and deceleration.
“We have designed our vehicle pretty much based on where NASA took off, and we added some special details in the inflation technology,” Klaus explains. “It builds upon decades of atmospheric reentry research, but is the most advanced technology that is out there right now.”
The concept involves extracting high-energy air from the boundary layer between the vehicle’s heat shield and the surrounding shock layer that is formed in front of it. This captured flow is then used to pressurize and inflate the vehicle. It’s a novel and somewhat counterintuitive approach, exposing the structure to hot plasma, controlling and repurposing the hot boundary layer air as a functional asset.
The LOFTID demonstrator is recovered on the deck of the recovery ship. (Credit: NASA)
Phoenix 1
The company’s debut Phoenix 1 recently flew on SpaceX’s Bandwagon-3 rideshare mission on April 21 — a mid-inclination mission to a 500 km low-Earth orbit (LEO). Phoenix 1 was built quickly as a minimum viable product (MVP) on a limited budget to gather initial flight data.
The capsule remained attached to Falcon 9’s upper stage while the other payloads were deployed. Phoenix 1 finally separated around 90 minutes into the mission over Los Angeles, following the upper stage’s deorbit maneuver.
“SpaceX did us a great favor by giving us a little kick to go into the atmosphere,” Klaus emphasizes, “which meant we didn’t have to bring our own propulsion.” This simplified the capsule’s initial design and also removed the risk of any issue with onboard propulsion, potentially leaving the capsule in orbit.
Onboard view of Phoenix 1 before it was deployed during the Bandwagon-3 mission. (Credit: SpaceX)
Around 30 minutes later, Phoenix 1 began its reentry and ultimately reached a splashdown in the South Atlantic Ocean. The original plan was for Phoenix to instead reenter east of Madagascar in the Indian Ocean, and ATMOS had spent six months planning the recovery of the capsule using ships. However, just five weeks prior to the launch, the trajectory of the Bandwagon mission was altered to satisfy constraints of one of the main payloads. This is an accepted risk for the smaller payloads on these rideshare missions, which are paying a significantly cheaper price to fly.
The alteration changed the splashdown location enough to scrap those months of recovery preparations and left little time to make new plans. The news dropped on a Friday, and ATMOS immediately scrambled its guidance, navigation, and control (GNC) team, which worked through the weekend. By Sunday evening, they had a solution, including the identification of different ground stations they could use during Phoenix 1’s descent.
“In space terms, five weeks is nothing, so we literally put an antenna in the jungle in Brazil, worked with a ground station in Ecuador, and had a lot of amateurs in the community supporting us.” These amateur ground stations picked up a lot of UHF data, which helped ATMOS collect over 135,000 telemetry points along the track.
Render of Phoenix 1 in orbit. (Credit: ATMOS Space Cargo)
The new landing location, around 2,000 km off the coast of Brazil, was beyond the range of the chartered aircraft, which had hoped to capture imagery and data as Phoenix completed its descent. “The last phase of the reentry was blacked out,” Klaus notes, “we didn’t have data all the way to splashdown because it was just out of the range of ground stations. We rented a little single-engine propeller plane with a Starlink terminal under the windshield of the cockpit. We had a live connection to our mission control center here in Europe. It was wild. We had to improvise a bit!”
Four payloads onboard also returned data during this demonstration — a radiation monitor from the German Aerospace Center (DLR), as well as payloads for the United Kingdom’s Frontier Space and Japan’s IDDK.
Designing for reusability
Whereas traditional reentry capsules commonly separate from a service module, which contains propulsion, solar panels, and tanks, Phoenix is intended to be fully reusable. “Our goal is to bring everything back,” says Klaus, noting that Phoenix 2 will return with its own propulsion system, solar panels, avionics, and communications when it flies next summer. This is an economic decision as much as a sustainable one, and ATMOS is approaching this with realistic expectations. “We’ve all seen with SpaceX over and over again how difficult it is to bring something back from orbital velocities and get it to be reusable,” he adds, “but that’s what we’re aiming for.”
Render of Phoenix 1 with its inflatable heatshield. (Credit: ATMOS Space Cargo)
ATMOS draws its name from its primary focus on returning payloads through the atmosphere and back to Earth. The company grew quickly following the summer of 2023 after receiving a round of seed funding. Studying NASA’s International Space Station (ISS) Utilization Statistics document, it became clear that biomedical research is the predominant type of research conducted aboard the ISS and would be a promising downmass market. Examples include protein crystallisation, 3D printing of human tissue, and research into cancers and the aging process.
The team set out to build a vehicle that is designed around the standard-sized payloads used on the ISS, known as “mid-deck lockers.” “We said, if we build a vehicle that has the same conditions as the ISS, in terms of pressurized atmosphere, providing power and data, and if we can make that in a way where it’s completely automated, or remotely controlled, then, probably, we will be able to find customers,” Klaus explained. The design was set to accommodate three of these lockers, providing a 100 kg payload capacity and 100 watts of power.
With the principles established, the team began to construct its first prototype. “We always said it’s going to be an inflatable heat shield,” says Klaus. “It was quite the project to do in less than two years! One of the things that is unique about this area of Europe is that there are a lot of partners and suppliers that you do not necessarily have in all places in the US, for example.” With a strong automotive industry in Germany that was already familiar with hardware-in-the-loop testing, ATMOS was able to leverage local resources to develop its project on a budget.
Render of Phoenix during reentry. (Credit: ATMOS Space Cargo)
The finished design for Phoenix 1 had a total mass of 250 kg and measured 1.5 m in diameter and 1.2 m in length, specifically designed to fit into an XL slot on SpaceX’s Transporter and Bandwagon rideshare missions.
Phoenix 2
The next Phoenix to launch will carry its own propulsion system, using thrusters from European startup ISP Tech. “The good thing about them is they are green propulsion; ethane and nitrous oxide,” Klaus noted. The more standard choice of Hydrazine used in thrusters is toxic, he points out, and requires the support crew to wear protective suits both when fuelling or returning to a landed vehicle, such as Boeing’s X-37B, due to the risk of residue in the thrusters. “You don’t want to have that on your vehicle, especially if it’s a reusable vehicle that you want to recover and then relaunch,” he said.
These small thrusters provide an impressive 290 seconds of specific impulse. Six of these will be used for pitch and yaw control, while smaller thrusters will provide roll control, collectively enabling Phoenix to control its reentry trajectory when deorbiting and to perform any necessary station-keeping.
The next flight will remain in orbit for two to four weeks before slowing down to around 150 meters per second, where it will enter the atmosphere. “The plan for the next flight is to fly to Santa Maria. There’s an island in the Atlantic Ocean that belongs to Portugal, so it will be the first reentry towards European territory!” Klaus said. The capsule will be filled to capacity with 100 kg of payload. There remains a small amount of risk, as they have yet to receive data all the way to splashdown. However, ATMOS is fully confident in the orbital phase of the mission, where most of the data will actually be generated.
The company is equally confident in the accuracy of the capsule’s landing, which, in simulations, achieved a precision of less than 100 m from the target. “That’s something you cannot even achieve right now with, say, a Dragon because these classic vehicles pop parachutes at an altitude of a few kilometers and from there just drift down with the wind. We can steer all the way until we touch the water, and that enables us to land very precisely.”
Render of the Phoenix control surfaces (Credit: ATMOS Space Cargo)
The donut design of the vehicle is purposely asymmetric, with the vehicle’s center of mass offset to one side. “That makes the vehicle fly at an angle of attack that creates a lift vector,” Klaus explains. By adjusting the direction of this lift, known as a bank angle maneuver, the vehicle can curve left, right, up, or down as it reenters the atmosphere. Two aerodynamic surfaces control this, along with some small roll control thrusters located at the back of the capsule. “The cool thing about our technology is we can use it all the way from orbit, through the entry interface point where we feel the atmosphere for the first time, all the way to a splashdown in the ocean.”
ATMOS has few competitors, of which Varda Space and Inversion Space both apply parachutes for the final descent of their return capsules. Phoenix doesn’t require parachutes due to its large surface area and low mass, allowing its shape to essentially act as a parachute. The ballistic coefficient, as it’s called, is lower than NASA’s LOFTID demonstrator. “You’re flying very slow,” explains Klaus. “At splashdown, you have less than 15 meters per second of terminal velocity, and that’s the equivalent of jumping from a five or 10-meter stair into a pool. That’s what keeps the system very simple, and in spaceflight, simplicity is king. The less systems you have, the less that can fail.”
With this in mind, ATMOS is comfortable with ocean landings, where the shocks of impact are much lower than on land. Seawater corrosion remains a challenge that can affect systems and reusability, but SpaceX has proven this to be solvable with its Dragon capsules.
Render of Phoenix as it is deployed into orbit. (Credit: ATMOS Space Cargo)
ATMOS intends to launch the enhanced Phoenix 2 vehicle next summer, with another capsule following either late that same year or in early 2027. Launches will then begin to ramp up to three or four annually, building up to monthly flights.
“The goal is that, eventually, we will be able to enable a rideshare service from space back down [to Earth],” notes Klaus. “What you’re getting with Transporter or Bandwagon for the way up, we want to enable that for the way down. If you have a mid-deck locker or a cubesat-sized payload, maybe a University payload, we want to be able to send that on a regular basis to space and back again. I think that’s going to make a huge difference, especially for the research community.”
The vehicle is designed to be launcher-agnostic and is compatible with a standard 24-inch ESPA separation ring — the structural adaptor used to carry and deploy secondary payloads alongside a primary satellite. This would allow Phoenix to launch on New Glenn, Ariane 6, and Vega, or aboard smaller rockets like Electron and Spectrum. “Right now, SpaceX is the most reliable launcher and the most cost-efficient launcher, so I think there’s a high chance we might be flying with SpaceX on the next mission,” he notes.
Future use-cases
For that mission, Atmos has been working with the Portuguese space agency to acquire the necessary reentry license in time for its second mission. “We’re very much looking forward to bringing that vehicle back to European soil,” Klaus adds. Looking further into the future, ATMOS is closely monitoring progress on the various commercial space stations that will replace the ISS and, inevitably, host more research customers with payloads to return to Earth.
Phoenix prototype prepares for drop testing. (Credit: ATMOS Space Cargo)
A Phoenix 3 is already in the design stage, which would increase payload capacity to around 1,000 kg. This can be scaled even further, Klaus points out, until they reach a calculated limit of around 25,000 kg. At this point, the square-cubic rule dictates that the combined mass of the gas and inflation system would become prohibitively heavy in comparison to the heat shield. Nonetheless, he notes, “25 tons means a fully filled shipping container is something we could bring back from space with this technology, so it opens a lot of possibilities.”
The company is currently exploring rapid response and point-to-point applications for the military through a partnership with ARX Robotics. The technology may even be applied to returning smaller rocket upper stages in the future. This idea was originally included in an unsuccessful pitch to the European Space Agency in collaboration with Isar Aerospace in 2022, but has since been contracted by the European Commission for its Inflatable Concept Aeroshell for the Reuse of Upper Stages (ICARUS) project. Klaus stresses that returning upper stages is especially difficult, as SpaceX’s Starship has proven.
“One of the use-cases we’re looking at is the Vega upper stage. It’s a scalable technology, so you might as well use it to bring back something like an Ariane 6 upper stage, and that’s what inflatable heat shields are really designed for — to slow you down from orbital velocities,” Klaus explained.
The ICARUS project is aiming for a suborbital demonstration in around two years, after which it would hopefully move onto an orbital test. Klaus can see similar potential for the technology in returning satellites for servicing and reuse.
“There are great companies out there [such as] Astroscale and Clearspace already working on the capability to grab a satellite in space and then drag it into the atmosphere. Now, if you combine that technology with ours, that’s a really cool capability we haven’t had since the Space Shuttle days!”
(Lead image: A render of the Phoenix craft reentering Earth’s atmosphere. Credit: Atmos Space Cargo)
