As humanity extends its infrastructure toward the Moon through the Artemis program, the reliability of computing systems faces an invisible enemy: cosmic rays. In the vacuum of space, a single-bit error can mean the difference between a historic landing and a catastrophic loss.
1. The Bit Flip Phenomenon (SEU)
On Earth, the atmosphere and magnetic field shield us from most charged particles. In deep space, however, high-energy protons can pass through spacecraft shielding and collide with the semiconductors inside a chip.
That impact can trigger a Single Event Upset (SEU), also known as a bit flip: the particle alters the electrical state of a transistor from 0 to 1. If that happens inside a critical memory address or a flight instruction, the system can crash or execute unintended maneuvers.
2. The Artemis II Dilemma: Why Fly with "Old Technology"?
Recent reports have sparked debate after revealing that astronauts on the Artemis II mission will travel with systems that, by commercial standards, seem outdated. According to reports from La Vanguardia, NASA uses previous-generation hardware and software, including Windows 8 variants, for critical load-management and experiment functions such as the EPM.
Why not use the latest i9 processor or an Apple M3?
Proven Reliability: In space, "new" often means risk. A system that has operated for a decade without logical failures is far more valuable than a faster but unproven one.
Physical Resistance: Modern chips have transistors so tiny that a single radiation particle can affect several bits at once. Older chips, built with larger process nodes, require more energy to flip a bit and are therefore naturally more robust.
3. Hardware Redundancy: The Rule of Three
In crewed missions, trust does not rest on a single processor but on Triple Modular Redundancy (TMR).
Majority Voting: Three identical systems perform the same calculations. A majority-vote component compares the outputs. If computers A and B say "turn left" and C says "right" because of a bit flip, the system ignores C and executes the correct command.
Rad-Hard Processors: Devices designed with silicon-on-insulator substrates to make them physically less vulnerable to ionization.
4. Software Redundancy: The Last Line of Defense
When hardware fails, software must be able to self-heal:
EDAC (Error Detection and Correction): Algorithms such as Hamming or Reed-Solomon monitor RAM continuously. If a bit changes, the software detects and corrects it before the processor uses it.
Design Diversity: In spacecraft such as Orion, multiple software layers are used. If a logical error affects the main application, backup systems with simplified and ultra-stable codebases can take over.
Watchdog Timers: Independent processes that reboot the system if they detect the primary software has frozen because of a radiation event.
5. The Future: COTS vs. Rad-Hard
The current trend, led by SpaceX and increasingly adopted by NASA for non-critical assets, is to use modern commercial processors (COTS) in parallel. It is more efficient to use six powerful processors with extreme redundancy logic than a single ultra-protected processor that is 100 times slower. Even so, for Artemis II life-support and mission-critical systems, NASA still prioritizes architectural robustness over raw computing speed.
Conclusion
The use of "old" software in Artemis II is not a lack of innovation. It is mastery of safety engineering. The bit flip reminds us that code is a physical entity, exposed to the laws of the universe. Redundancy and stability are the real engines that will allow humanity to return to the Moon and, eventually, reach Mars.