Artemis II Co-Pilot: A.I.

From Pocket Calculators to Super Advanced Technology

In the 1960s and 1970s, NASA’s Apollo program achieved what many called impossible: sending astronauts to the Moon using technology that today fits in a child’s toy.

The Apollo Guidance Computer, the spacecraft’s onboard brain, had just 4 kilobytes of Random Access Memory (RAM) and 72 kilobytes of Read Only Memory (ROM)—less processing power than a modern pocket calculator.

Astronauts relied on banks of switches, manual controls, and constant voice contact with Earth-based mission controllers who performed most complex calculations on the ground.

Today, with Artemis II—the first manned lunar mission since Apollo 17 in 1972—NASA is sending four astronauts on a 10-day flyby of the Moon using the Orion spacecraft. And this time, Artificial Intelligence serves as a tireless digital partner, transforming how humans travel into deep space.

Note: Artemis II is the name of the mission. The rocket itself is called Orion.

Read on for details.

When it comes to this modern day manned mission to the Moon, the leap in raw computing power alone is staggering.

Orion carries four redundant flight computers, each built around ruggedized IBM PowerPC processors derived from commercial airliner technology. These systems operate 20,000 times faster than the Apollo Guidance Computer and boast 128,000 times more memory.

While Apollo’s single computer handled narrow tasks like lunar descent guidance, Orion’s computers manage virtually every system. This means life support, navigation, power, and propulsion. There are almost no manual switches except for rare emergency overrides.

Everything from trajectory adjustments to system checks runs through these high-speed, radiation-hardened machines, allowing the spacecraft to fly autonomously for long stretches, a capability proven during the unmanned Artemis I test flight.

But you might say the real revolution lies in how AI turns the computing power into intelligent oversight.

During Orion’s development and testing, Lockheed Martin integrated NEC’s System (Japanese technology) Invariant Analysis Technology (SIAT). This AI tool examined data streaming from more than 150,000 sensors across the spacecraft and mapped out over 22 billion logical relationships between systems.

The result is an anomaly-detection engine that spots subtle behavioral shifts in everything from power distribution to life-support equipment long before they become problems.

SIAT continues to operate during flight, providing real-time monitoring that was impossible in the Apollo era. Crews and ground teams had to catch issues through limited telemetry and human judgment alone. In many ways, it’s incredible we made it to the Moon!

Working with SIAT are digital twin simulations. These are virtual replicas of Orion and its systems running in parallel with the real spacecraft. Obviously, this is a huge safety and performance addition over the Apollo days.

The AI-driven models constantly compare predicted performance against actual sensor readings, flagging drifts in oxygen levels, temperature, radiation exposure, or propulsion health.

On Artemis II, the vast majority of trajectory planning, life-support monitoring, and navigation decisions happen autonomously through advanced algorithms. Astronauts still take manual control at key moments to test handling and build experience, but AI handles the relentless data crunching that would otherwise overwhelm a crew thousands of miles from Earth.

Communication delays with mission control, which can stretch for seconds or minutes in deep space, are less risky when onboard AI can analyze options and suggest solutions instantly.

The missions themselves highlight this shift.

Apollo flights aimed for bold objectives such as lunar orbit insertion, landings on the surface, and precise returns, often with crews performing critical maneuvers by hand while Earth teams calculated every burn.

But Artemis II follows a safer free-return trajectory. This is where the Moon’s gravity acts as a natural slingshot to bring the crew home without relying on a major engine firing. This prioritizes testing deep-space systems over immediate landing, with AI ensuring the spacecraft stays on course and healthy throughout the roughly 4-day outbound leg and high-speed lunar flyby.

In the Apollo era, astronauts were pioneers operating at the edge of human and machine limits. I have had the pleasure of meeting and interviewing a number of these pioneers. Some of the stories they told will be featured in an upcoming podcast I’m planning.

Today’s Artemis II crew—Reid Wiseman, Victor Glover, Christina Koch, and Jeremy Hansen—operate as a skilled team supported by AI that processes sensor data at scales no human could match. It predicts issues proactively, and frees the humans to focus on science, observation, and the experience of deep space.

Orion itself offers 30 percent more habitable volume than the Apollo command module, powered by solar arrays rather than fuel cells, and features a modern glass cockpit with digital displays instead of toggle switches.

This evolution does not undercut the human courage and ingenuity it takes to plan and serve on a Moon mission.

But you have to really think with wonder about how man reached the Moon in the Apollo era with little more than slide-rule precision and raw determination.

The next chapter is being written through a true partnership between astronauts and intelligent machines. AI is truly acting as co-pilot.

(Let’s hope it doesn’t all turn out like it does too often in the movies.)