Grace Kim, Stanford University; Duncan Eddy, Stanford University; Mykel Kochenderfer, Stanford University
https://doi.org/10.64861/IPHU7563
Keywords: Conjunction avoidance, probability of collision, belief-space planning, chance constraints, Monte Carlo tree search, autonomous maneuvering, space traffic management, probability of collision
Abstract:
Increasing conjunction frequency in Low Earth Orbit places growing pressure on spacecraft operators to determine not only whether an encounter requires mitigation, but whether sufficient information is available to commit to a maneuver. Delaying action may provide additional tracking information that reduces state uncertainty and resolves an apparent conjunction. However, waiting also reduces the time and maneuver authority available if intervention ultimately becomes necessary, potentially increasing the required propellant expenditure.
This work formulates this information-action tradeoff as a belief-space planning problem for conjunctions between a maneuverable spacecraft and an unmaneuverable secondary object. The planner represents the uncertain orbital states as Gaussian beliefs and uses a chance-constrained belief-space Monte Carlo tree search framework to reason over possible future tracking updates before time of closest approach (TCA). Because these observations are not known in advance, the resulting belief at TCA is uncertain, as different possible observation sequences may produce different terminal belief states.
A terminal chance constraint limits the probability of reaching TCA with a belief whose probability of collision exceeds a prescribed threshold, allowing the planner to wait when future tracking is likely to resolve the conjunction while requiring intervention when continued deferral becomes too risky.
We evaluate the approach on eight historical conjunctions from NASA’s Conjunction Assessment Risk Analysis dataset. By varying the secondary-object measurement quality and tracking cadence, we generate a total of 96 distinct evaluation scenarios. Across the evaluated conditions, the planner reaches TCA without maneuvering in approximately 40% of episodes while maintaining no terminal collision-risk violations.
In contrast, fixed-time rule-based maneuver policies resolve more encounters without maneuvering when intervention is deferred closer to TCA, but at the expense of increasing terminal risk violations. The ability to delay maneuver decisions depends strongly on tracking quality and measurement cadence, ranging from 76% under accurate, frequent measurements to approximately 18%-20% under the poorest tracking conditions.
For the same physical conjunction, changing only the expected tracking quality can shift the preferred decision from continued waiting to delayed maneuver commitment or immediate intervention. These results show that tracking quality and frequency are not only inputs to collision-risk estimation; they can determine when intervention becomes necessary. Explicitly valuing future information therefore provides a principled mechanism for deciding when an avoidance decision can safely remain open and when uncertainty itself justifies action.
Date of Conference: September 15-18, 2026
Track: Conjunction/RPO