Six Astronauts Is the Magic Number: New Study Reveals Optimal Crew Size for NASA’s Moon Base
Featured image: NASA concept art of a future moon base near the lunar south pole. Credit: NASA
As NASA formalizes plans for a permanent base near the Moon’s south pole under the newly renamed Moon Base Program, a fundamental question has gone largely unasked: how many astronauts should live there? A new study published May 27 in PLOS ONE provides the first rigorous attempt at an answer, and the finding challenges some of the conventional wisdom about crew selection for deep-space missions.
The answer, based on agent-based modeling of thousands of simulated lunar missions: six astronauts is the optimal crew size. Four is the minimum, and the worst case.
“People can be very, very well trained, but for long-duration or deep-space missions, there will always be a human factor involved,” said Anamaria Berea, associate professor at George Mason University and lead investigator of the study.
The study, “Lunar Base Agent-Based Modeling: A Benchmark for Simulating Crewed Space Missions,” was funded by George Mason University and used an open-source Python framework called Agent_Astronaut to simulate how crews of different sizes and compositions would perform under various mission scenarios. The model ran 10,000 Monte Carlo iterations across nine distinct cases, varying crew size, resupply frequency, mission duration, and environmental hazards.
Why training is not enough
The study’s central finding is that the conventional approach of emphasizing psychosocial training and crew compatibility screening is insufficient for long-duration lunar missions. Mission design parameters, the researchers found, are more powerful determinants of success than individual crew characteristics.
The Agent_Astronaut model incorporates each astronaut’s skill level across four domains (EVA, science, engineering, flight operations), personality type using the DISC framework (Dominant, Influencer, Steady, Conscientious), emotional coping capacity, and interpersonal tension. The key equation combines skill, emotional state, and a technology-learning factor to determine whether each task is completed.
In the baseline case of four astronauts on a nominal mission, the model found that crews completed only about 20 percent of scheduled tasks: a figure the study calls “acceptable for a typical manufacturing process” but clearly inadequate for a high-stakes lunar outpost. The low completion rate suggests that even well-trained teams are “having challenges overcoming psychological stressors and environmental disruptions.”
What the scenarios revealed
The most dramatic improvement came from increasing crew size. Expanding from four to ten astronauts produced a 31 percent improvement in the synthetic Task Load Index score, the single largest effect of any variable tested. Larger crews provide better skill specialization, higher probability of compatible personality pairings, and a shared maintenance burden that mirrors real-world findings from the International Space Station.
The optimal balance between crew size and logistical cost settled at six. Below that, the psychological and operational penalties of a small, tightly coupled team become acute. Above that, the marginal benefit diminishes while resupply requirements and habitat mass increase.
Other findings:
- Resupply every two weeks is ideal; monthly resupply degrades performance significantly
- Six-month missions cause a 27 percent drop in coping capacity over time, even if the synthetic TLX score holds steady
- Bi-weekly crew rotations between the base and the Lunar Gateway reduce tension by 10 percent
- The death of a crew member in the final third of a mission causes a measurable decline in the remaining team’s performance
“The team is more than the sum of its people,” Berea said. “We need to pay attention not only to the astronauts, but the team as a whole, and each team and space mission are unique.”
Built on Antarctic analog data
The model’s psychological parameters were calibrated against data from the EDEN ISS Antarctic greenhouse mission and historic Antarctic expedition records, which provide the closest Earth analog to the isolation, confinement, and environmental stress of a lunar base. Tension levels in the model matched 1993/1994 Antarctic expedition data closely, lending credibility to the projections.
The model explicitly does not yet include physiological effects such as radiation exposure and microgravity bone loss, nor does it simulate the communications delays that would affect deep-space missions beyond the Moon. The authors acknowledge these limitations and frame the study as a benchmark: a starting point for increasingly sophisticated simulations as NASA’s Moon Base plans take shape.
The code is open-source and available on GitHub (github.com/rvera-gmu/Lunar-Base-ABM), allowing other researchers to extend the model with additional variables and mission scenarios.
Why this matters now
NASA’s Moon Base Program envisions a three-phase buildout: Phase 1 (25 launches and 21 landings with approximately 4,000 kilograms of delivered payload), Phase 2 (27 launches, 60,000 kilograms, semi-annual crewed missions), and Phase 3 (29 launches, 150,000 kilograms, continuous crew presence). The entire campaign runs from 2026 through at least 2036, encompassing 81 planned missions.
The crew size question is not academic. Every additional astronaut multiplies the mass of life support consumables, habitat volume, and emergency supplies. An optimal crew of six, rather than the four-person minimum, represents a significant but manageable increase in mission costs: and, according to the model, a substantial improvement in the probability that the base will function as intended.
“The worst-case scenario consists of four astronauts on the moon at one time, only one month resupply window between Earth and Moon, and moderate to high adverse environmental probabilities,” the study states. Good mission design, the researchers argue, can prevent that worst case from ever becoming the default.