Orbital Debris and Kessler Mitigation
Low Earth orbit contains tens of thousands of tracked objects larger than 10 cm, and on the order of a million untracked fragments between 1 mm and 10 cm. At orbital velocity a 1 cm fragment carries the kinetic energy of a hand grenade, and every collision produces more fragments.
Kessler syndrome , proposed by Donald Kessler and Burton Cour-Palais in 1978, names the threshold past which collisions generate fragments faster than atmospheric drag removes them. Beyond it, the debris population grows without further launches, and bands of low Earth orbit become unusable for decades. It is a cascade with a critical point, and it is not obviously far away.
The cascade is a branching process
The framing that makes this tractable is the one from criticality. Define the branching ratio as the mean number of subsequent collisions caused by one collision’s fragments. Below cascades die out. Above it they run away. At the distribution of cascade sizes follows a power law, meaning a small event and a catastrophic one are drawn from the same process and cannot be distinguished in advance by their causes.
This is not a loose analogy. It is the same self-organized-criticality machinery used for avalanches and forest fires, and it implies that the useful question is not “will there be a collision” but “what is in this orbital shell, and how far below 1 does it sit.”
Why a swarm
Debris is distributed across orbital shells, and observing it well requires being in many places at once. A single bespoke servicer is expensive, specialized, slow to reach any particular target, and a single point of failure. A constellation of small sensing and servicing spacecraft buys geographic coverage, redundancy against the loss of any node, and per-orbit access to far more of the population, at a much lower marginal cost per target.
Coordination is the difficulty, and it is the same difficulty as everywhere else in orbit. Line of sight comes and goes; inter-satellite links exist on a schedule dictated by orbital mechanics rather than by a network topology; radiation flips bits. The constellation must agree on what it has seen and what it will do without a central authority it can reach on demand — the situation described at orbital compute constellations.
Open research questions
A distributed debris catalog. Each satellite observes fragments; the constellation must maintain a shared catalog that converges despite intermittent links and no central authority. Representing the catalog as a conflict-free replicated data type gives convergence under partition for free, provided every update is expressed as a commutative, idempotent merge. This is the same construction our intent-propagation work uses, and it is the reason that machinery generalizes: what is being replicated changes; the convergence argument does not.
Target selection under a delta-v budget. Removal is propellant-limited and propellant is unreplenishable. Which targets maximize Kessler-risk reduction per joule spent? This is a decentralized allocation problem with heterogeneous agent capabilities and a hard energy constraint, which is exactly what an energy-aware market mechanism addresses — agents bid according to their reachability and remaining propellant, and the allocation emerges without a planner that must be reachable.
Forecast horizon. How far in advance can a distributed observer detect that its local orbital shell is approaching ? Estimating a branching ratio requires observing enough events, and the events are rare precisely while the system is subcritical — which is the only regime in which the answer is useful.
Radiation-induced Byzantine faults. Bit-flips corrupt the state a node uses to participate in the catalog protocol. A node with corrupted state does not fail cleanly; it contributes plausible-looking garbage. Most swarm coordination tolerates fail-stop agents and is defeated by Byzantine ones, and CRDT merge semantics — which are designed to accept every update — are particularly exposed. This is an open direction and, we think, the hard one.
Dual use
Debris-tracking constellations run the same coordination algorithms as orbital surveillance constellations, and a servicer that can rendezvous with a dead satellite can rendezvous with a live one. Arboria does not contribute directly to weaponizable orbital inspection or intercept, and we gate publications that would materially accelerate those capabilities. See our principles.
Adjacent work
Active-debris-removal programmes include ESA Clean Space , NASA’s Orbital Debris Program Office , and JAXA’s CRD2. Ground-based and hosted tracking networks include LeoLabs and Slingshot Aerospace. The cascade-modeling literature runs from Kessler and Cour-Palais (1978) through Liou and Johnson’s long-term evolution studies.
Related
Orbital compute constellations share the geometry and the coordination constraints. Communication in harsh environments sets out the underlying obstacle, and Phase Diagram of Coordination Under Delay quantifies what it costs.