Two bodies flew around the Moon on Artemis I and came home. Neither was alive. Both were female. One was wearing a vest, and the difference between them is now the best evidence we have about keeping people alive in deep space.
Here is the number that governs human spaceflight. NASA allows an astronaut 600 millisieverts of radiation across an entire career, a limit set to hold the lifetime risk of dying from a radiation-induced cancer at about three per cent.
A single large solar storm can spend more than a third of that in hours. A three-year round trip to Mars would spend all of it.
So what does a crew do when a storm arrives? Today they hide. They stack supplies into a barricade, climb behind it and stop working until it passes. The state of the art is a fort made of luggage.
In 2015 the Israeli company StemRad began building an alternative with Lockheed Martin. The idea came from an odd place. StemRad's chief executive, Oren Milstein, trained under a supervisor who treated the Chernobyl first responders, and what killed those men within months was damage to their bone marrow. Shield the marrow and you save the person. Not the whole body, just the parts that matter most. Seven years later that idea flew around the Moon.
Here is what happened
Two mannequins flew a lunar mission as a controlled experiment. MARE put two phantom torsos aboard Orion for Artemis I, the uncrewed 25-day flight around the Moon. They have names, Helga and Zohar, they are shaped to mimic the female body, and between them they carried more than 12,000 radiation sensors, sliced through the torso so readings line up with real organs.
One wore the shielding, one did not. Zohar wore the AstroRad vest. Helga flew bare. Same spacecraft, same trajectory, same radiation environment, one variable.
The vest is thousands of small plastic rods. Hexagonal high-density polyethylene, tessellated between two layers of stretch fabric so the rods slide as the wearer moves. Polyethylene is hydrogen-rich, which is unusually good at stopping charged particles. It is thickest over breasts, lungs, stomach, colon, ovaries and red bone marrow.
The results arrived in stages. MARE was run by the German Aerospace Center with NASA, the Israel Space Agency, StemRad, Lockheed Martin and ESA, which supplied five active dosimeters around the cabin. First findings appeared in Nature in 2024. The vest analysis was published in Science Advances on 12 August 2026.
Both test bodies were female. This was deliberate. Women carry a higher predicted risk of radiation-induced cancer, which means the most vulnerable crew member sets the mission duration limit for everybody.
What they found
There is a catch in the method that makes the result more interesting, not less.
No solar storm arrived. Artemis I flew through a quiet Sun. The one thing the vest was built for never happened during the mission, which could have wasted the entire experiment.
The Van Allen belts stood in. Orion crossed the inner belt, where trapped protons sit in roughly the same energy range as storm protons. That gave the team real particles, at the right energies, hitting real detectors inside a real spacecraft.
So they built digital twins. Software models of Helga, Zohar, the vest, every detector and the shielding of Orion itself. They fired hundreds of billions of simulated protons and electrons at it, then checked the predicted detector readings against what the sensors actually recorded in flight. The two agreed closely.
Then they ran the storms that did happen, in history. With a model anchored to real flight data, they simulated the great solar particle events of August 1972 and October 1989.
The vest cut effective dose by about 60 per cent in a 1972-type event, and by almost 40 per cent in a 1989-type one. In deep-space terms that is sparing an astronaut the equivalent of up to 193 and 131 days of ordinary cruise exposure.
Why this is a frontier story
It replaces hiding with working. The vest gives protection comparable to the best storm-shelter arrangement a crew can build out of stowage, while the person wearing it can move around the cabin and keep doing the mission. A shelter protects a body. A garment protects a body that is still flying the spacecraft.
The real deliverable is the model, not the garment. There is now a simulation of organ dose on a lunar trajectory, checked against measurements from an actual lunar trajectory. You cannot get that from a laboratory beam line or from low Earth orbit, and it is what lets anyone ask what 1972 would have done to this crew, in this vehicle, on this route, and trust the answer.
It turns a survival question into a design question. Deep-space dose has been an input you plan around and accept. Once it is predictable per organ, per seat and per orientation, mission length becomes something you engineer rather than endure. That step has to happen before anyone honestly plans three years to Mars.
Protection stops being infrastructure and becomes equipment. Vehicle shielding is fixed when the spacecraft is built and improves on a decade-long cycle. A garment is sized to a person, swapped between missions and carried by the crew rather than the ship. The team is already looking at flying the vest empty and filling it with recycled polyethylene in orbit, which would make crew shielding something you top up rather than launch.
The winning principle is uneven, not more. Concentrating mass over the most vulnerable organs beat spreading the same mass evenly across the body by roughly 30 per cent. That runs against the intuition that fuller coverage is safer, and it is a rule that transfers to habitats, rovers and built-in storm shelters, not just to vests.
The ship turned out to be a lever too
The vest is the headline, but it is not all the flight measured.
Dose is not uniform inside the capsule. Detectors spread through Orion found a patchwork, with the best-shielded areas offering up to four times more protection than the worst. Where a crew member sits is a decision with a dose attached.
Turning the spacecraft halved the dose. During a 90-degree rotation near the Van Allen belts, exposure fell by about half. Face the vehicle differently and the crew sits behind more of its own structure.
Two of the three levers are nearly free. A seating plan costs nothing and an attitude command costs a little propellant. The vest is the one that must be designed, built, launched and then actually worn.
The honest catch
A strong result with clear boundaries, and the boundaries deserve stating plainly.
The headline numbers are modelled, not measured in a storm. No solar particle event occurred during Artemis I. The 60 and 40 per cent figures come from a simulation whose credibility rests entirely on how well it matched the belt-crossing measurements. That is a reasonable way to do this. It is not the same as having flown through 1972.
Mannequins are not people. Dose is read at detector positions and mapped onto organ locations. Nobody wore this on Artemis I.
The vest helps most where the threat is easiest. Solar particle events are the shieldable half of the problem. Galactic cosmic rays, the constant background that dominates a multi-year Mars mission, are far less affected.
The mouse study is a mouse study. A separate paper published on 13 August found lasting blood vessel damage and accelerated arterial plaque in mice under simulated cosmic radiation. A signal to investigate, not a conclusion about astronauts.
Comfort is unresolved. A garment only protects you if the crew will actually wear it, in a hot spacecraft, for hours, while working. The team is still cutting mass.
EDITOR'S TAKE
The quiet fact in this paper is not the vest. It is that both test bodies were female, because women hit their radiation exposure limit sooner, which means the length of a mission is set by the most vulnerable person you send. That is a design constraint dressed up as a biology fact, and it has sat in the background of human spaceflight for decades. Cut organ dose by 60 per cent and you do not just protect a crew, you potentially extend how long a mixed crew is allowed to stay, which quietly changes who gets to go. The other thing worth holding onto is where the idea came from. A vest designed around Chernobyl bone marrow injuries is now the best answer anyone has for a solar storm on the way to Mars. The frontier usually borrows before it invents. Watch whether the garment becomes standard Artemis equipment rather than an experiment, and watch the cardiovascular literature. If the headline risk migrates from cancer to heart disease appearing a decade after landing, the acceptable mission length changes again.
Quick questions
Did the radiation vest actually work?
Yes, with an important qualification about how that was established. Artemis I did not encounter a solar storm, so the vest was never tested against the threat it was designed for. What the mission did provide was a crossing of the inner Van Allen belt, where trapped protons sit in a similar energy range to solar storm protons. The team used that to build a detailed computer model of the whole experiment, digital twins of both phantoms, the vest, every detector and Orion's own shielding, and fired hundreds of billions of simulated particles at it. When the model's predicted readings matched the sensors' real in-flight readings closely, they used it to simulate two of the largest solar particle events on record. In an August 1972-type event the vest reduced effective dose by around 60 per cent, and in an October 1989-type event by almost 40 per cent, sparing the equivalent of up to 193 and 131 days of ordinary deep-space exposure. The paper also found that concentrating shielding over the most vulnerable organs beat spreading the same mass evenly across the body by roughly 30 per cent. The results were published in Science Advances on 12 August 2026.
How dangerous is space radiation, really?
Dangerous enough to be a primary limit on mission design. NASA caps an astronaut's career exposure at 600 millisieverts, a figure chosen to keep the lifetime risk of dying from a radiation-induced cancer at roughly three per cent. A single large solar particle event can deliver more than a third of that in hours, and a three-year round trip to Mars would equal or exceed the whole career allowance. There is also a slower concern that has been less discussed. A study published on 13 August 2026 found that simulated galactic cosmic radiation caused lasting damage to blood vessel linings in mice and accelerated arterial plaque, which raises the possibility that cardiovascular disease appearing years after a flight belongs alongside cancer in the risk picture. On the reassuring side, Artemis I showed the best-shielded parts of Orion stayed below 150 millisieverts during a significant solar event, comparable to a year aboard the International Space Station.
Does this help with a Mars mission?
Partly, and it is worth being precise about which part. A vest is most effective against solar particle events, the acute threat that arrives in hours and can be acted on, and those occur on any journey. The dominant problem on a multi-year Mars mission is the constant background of galactic cosmic rays, which are far too energetic for a wearable garment to stop, and which is what pushes a three-year round trip past the career dose limit. So this is a real contribution to crew safety rather than a solution to deep-space radiation. Mars will still need a combination of approaches: faster transits so there is less exposure time, storm shelters built into the vehicle, seating and orientation planned around dose rather than convenience, and possibly medical countermeasures that do not yet exist. What has changed is that these can now be traded against each other using a model that has been checked against real flight data.
Sources
Science Advances: the peer-reviewed analysis of the MARE experiment and the AstroRad vest, published 12 August 2026.
Physics World: the modelling method, the 60 and 40 per cent figures, and interviews with StemRad's Oren Milstein and Jordan Houri, 26 August 2026.
ESA Orion blog: the cabin shielding gradient, the sub-150 millisievert solar event result and the 90-degree rotation, 4 December 2024.
German Aerospace Center (DLR): lead on MARE, with NASA, the Israel Space Agency, StemRad, Lockheed Martin and ESA.
Clinical and Translational Medicine: simulated galactic cosmic radiation, endothelial damage and accelerated atherosclerosis in mice, published 13 August 2026.
Frontier Signal explains frontier technology in plain English. Agency and laboratory figures should be independently verified. This is general information, not medical advice.
