Researchers used thousands of computer simulations to study and test how spacecraft can safely position themselves around NASA’s Gateway, balancing precise navigation, fuel efficiency and the challenges of future lunar missions. (Credit: Texas A&M University College of Engineering/Dr. Diane C. Davis)
The Tricky Math Behind Parking Spacecraft Near NASA’s Future Moon Outpost
In A Nutshell
- NASA’s new Gateway station will orbit the Moon on an unstable path that requires constant course corrections just to stay on track.
- Researchers tested three levels of navigation precision and found tighter control cuts drift to a few minutes, at the cost of noticeably more fuel.
- Spacecraft waiting to dock with Gateway can safely park much closer together by tracking Gateway’s actual position instead of following a fixed flight plan.
- The tighter parking strategy held up even when researchers simulated the chaos of a real Artemis mission, including docking, undocking, and venting.
Several spacecraft trying to park, dock, and maneuver around a station orbiting the Moon, each one a potential collision waiting to happen. That’s the reality NASA is preparing for with Gateway, its planned outpost near the Moon. A new study published in Acta Astronautica tackles an unappreciated challenge of the Artemis program: how to safely manage multiple spacecraft flying near each other in a complex gravitational environment.
Gateway won’t sit in a simple circular orbit like the International Space Station. Instead, it will travel an elongated, looping path pulled by both Earth’s and the Moon’s gravity at once, tracing a shape closer to a stretched-out oval than a neat circle. That tug of war between two massive bodies makes the orbit slightly unstable, meaning spacecraft won’t naturally stay on course without regular small rocket firings. Multiply that problem by four or more vehicles flying in close proximity, and a miscalculation could mean a dangerous close encounter hundreds of thousands of miles from home.
Researchers set out to determine how tightly each spacecraft needs to stick to its planned path, how much that precision costs in fuel, and how close together vehicles can safely wait while preparing to dock.
Gateway’s Unstable Orbit Forces Constant Course Corrections
Gateway is planned to travel a specific looping Moon orbit with a roughly 6.56-day period, chosen partly because it avoids long shadow periods that would cut off solar power and lets visiting spacecraft arrive on a predictable schedule. Left alone, a spacecraft there will slowly drift off its intended path, so it has to fire small thrusters roughly once per orbit to nudge itself back on course, a process mission planners call stationkeeping. That gets a lot harder once Orion, a lunar lander, and a cargo ship all need to do it nearby at the same time.
To figure out how tight the leash needs to be, researchers ran computer simulations testing three levels of precision, from loose to razor-sharp, across 100 trials each and up to seven years of simulated flight. They built in the kinds of small, unavoidable errors a real spacecraft deals with: slightly imprecise navigation, the gentle push of sunlight on the hull, and burns that never fire with perfect accuracy. Loose control let Gateway’s timing drift by close to an hour in the worst cases.
Tightening the controls brought that down to just a few minutes, but it cost noticeably more fuel and stopped working reliably during part of the orbit. Mission controllers can also switch between loose and tight control mid-mission depending on what a given phase of the mission calls for, with the spacecraft settling into the new rhythm within about a month.
Waiting Spacecraft Can Safely Park Much Closer Than Expected
For Artemis missions, the most pressing traffic challenge involves spacecraft waiting their turn to dock with Gateway. A lunar lander might arrive weeks before its crew does. A cargo ship might hold position while Orion completes a docking. These vehicles need a safe place to wait: close enough that docking doesn’t cost too much fuel, but far enough away that a navigation error doesn’t cause a collision. Researchers looked specifically at waits of up to 90 days, with the spacecraft lined up nose to tail along the same orbital path like a string of pearls, spaced apart not by miles but by minutes of travel time.
If each waiting spacecraft simply follows its own pre-planned path, small drifts can add up unpredictably. In the worst simulated cases, two vehicles meant to stay 30 minutes apart ended up within 4 kilometers of each other near apolune, the slow-moving far side of the orbit where gaps close in fastest, and less than 1 kilometer apart near Gateway’s closest pass by the Moon.
Researchers found a smarter fix: instead of each spacecraft tracking its own separate flight plan, the waiting vehicle can track Gateway itself and drift along with it. That single change kept the two vehicles hundreds of kilometers apart even when parked as close as an hour’s travel time from each other, and still tens to a few hundred kilometers apart at tighter spacing.
Only the very closest setting tested, about 10 minutes apart, saw some trials close to just a handful of kilometers, a distance the researchers flagged as too risky for a multi-week wait. The fix cost the waiting spacecraft a bit more fuel, but Gateway’s own fuel budget didn’t change at all, and the same tighter settings held up even when the researchers threw in the mess of a real mission: docking, undocking, and venting included.
That single tweak, tracking Gateway instead of a fixed plan, is what makes it realistic for a lander, a cargo ship, and a crew capsule to all wait their turn safely in the same patch of lunar sky.
Disclaimer: This article summarizes findings from a peer-reviewed engineering study and is intended for general informational purposes. It does not represent official NASA mission plans or operational guidance.
Paper Notes
Limitations
This study focuses on the difference between each spacecraft’s planned path and its actual simulated path, and explicitly notes that full navigation analysis, meaning the gap between where a spacecraft actually is and where its onboard systems believe it to be, is beyond the scope of this work. Authors acknowledge that classifying an exact separation distance as safe or unsafe is not straightforward, since uncertainties from navigation, spacecraft modeling, and maneuver execution interact in complex ways, particularly near the closest point in the lunar orbit. They note that distances on the order of tens of kilometers are likely unsafe for medium-term loitering, though no single hard threshold is established. The analysis of the tightest targeting case found that the approach loses effectiveness for maneuvers executed during certain portions of the orbit, and associated correction strategies were not fully verified for those locations. Higher-fidelity mission simulations tested whether the Gateway targeting settings held up under realistic Artemis mission disturbances, but represent specific assumed mission configurations rather than a full retest of the multi-vehicle loitering strategy.
Funding and Disclosures
No standalone funding statement appears in the paper. Authors note that trade names used in the paper are for identification purposes only and do not constitute an official endorsement by NASA. Authors declare no known competing financial interests or personal relationships that could have influenced the work. Acknowledgments credit Nathan Ré for a question at the 2025 International Astronautical Congress that led to an algorithm improvement, and Brian Robinson, Joe Scalora, and Alexander Ghosh for ongoing discussions about NRHO traffic management.
Publication Details
Authors: Diane C. Davis (Texas A&M University), Brian P. McCarthy (a.i. solutions), Stephen T. Scheuerle (NASA Johnson Space Center), Emily M. Zimovan-Spreen (NASA Johnson Space Center), Kathleen C. Howell (Purdue University) | Paper Title: “Cislunar traffic management: Orbit maintenance and loitering in the Gateway NRHO” | Journal: Acta Astronautica, Volume 248 (2026), Pages 185–196 | DOI: https://doi.org/10.1016/j.actaastro.2026.06.006 | Published: Available online June 3, 2026. This article is part of a special issue connected to the 76th International Astronautical Congress.







