Book a demo

Research into orbital data and collision-risk decision support.

Ramerion is exploring data-driven approaches to orbital awareness, collision-risk analysis and satellite decision support, with a long-term interest in more autonomous and resilient space operations.

This page sets out a research vision and the problem behind it. It does not describe a deployed system, a tracking capability or a collision-avoidance product.

Discuss a research collaborationThe collision-risk workflow

01The orbital environment

Working satellites share their orbits with debris.

Orbital debris is the human-made material left in orbit that no longer serves a purpose: defunct satellites, spent rocket stages and fragments from break-ups and collisions. Objects in low Earth orbit travel at kilometres per second, so even a small fragment can damage a spacecraft.

Operators manage that risk with what is known about where objects are and where they are going. That knowledge is imperfect: not every object is tracked, and every prediction carries an uncertainty that grows the further ahead it looks.

Schematic of objects in orbit, with one predicted close approach magnifiedLeft: Earth surrounded by low, medium and geostationary orbits, with active satellites and debris objects sharing the low orbits. Right: a magnified view of one predicted close approach, showing a satellite and a debris object on crossing tracks, each predicted position surrounded by an uncertainty region that is longest along the direction of travel, and the point of closest approach between them. Not to scale.LEOMEOGEOClose approach
  • Active satellite
  • Debris object
  • Position uncertainty, longest along-track
  • Point of closest approach
  • LEO · MEO · GEO: low, medium and geostationary orbit
Concept visualisationSchematic and not to scale: orbit sizes, object counts and spacing are illustrative. The highlighted pair shows what a predicted close approach looks like — it is not a real event.

02Five different problems

Detecting, tracking, predicting, deciding and removing are separate tasks.

They are often blurred together. Separating them shows where data, analysis and decision support can help — and where Ramerion is not working at all.

ProblemWhat it involvesRamerion's research interest
Detecting objectsSensing objects with radars, optical telescopes and other instruments.Not pursued. Detection depends on observation systems operated by agencies and specialist providers.
Tracking and orbit determinationAssociating observations with known objects and estimating each object's orbit and its uncertainty.Handling and validating orbital data, and making its uncertainty understandable.
Predicting a close approachPropagating orbits forward and screening for pairs of objects that may pass close to each other.Presenting conjunction events, their uncertainty and how they change as new data arrives.
Deciding whether to manoeuvreWeighing collision probability against mission objectives, fuel and the risk a manoeuvre creates.Decision support for the operator — never the decision itself.
Removing debrisCapturing or de-orbiting objects with dedicated missions.Not pursued.

03Concepts, in plain terms

The vocabulary of collision risk.

The terms below recur in every serious discussion of space safety, including the research directions on this page.

Orbital debris
Human-made objects in orbit that no longer serve a useful purpose — defunct satellites, spent upper stages and fragments.
Space situational awareness (SSA)
The combined picture of what is in orbit, where it is and how it is behaving, built from many observations and data sources.
Space surveillance and tracking (SST)
The observation systems — radars, telescopes, laser ranging — and the tracking methods that estimate objects' positions and trajectories.
Orbit determination
Estimating an object's orbital state, its position and velocity, from observations, together with how uncertain that estimate is.
Prediction uncertainty
Every predicted position comes with an uncertainty region. It generally grows with the time span of the prediction and is usually largest along the direction of travel.
Conjunction assessment
Screening predicted trajectories for pairs of objects that may come close, then examining each such event in detail.
Collision probability
An estimate that depends on the predicted relative trajectories, the combined uncertainty of both objects and the geometry of the encounter. It informs a decision; it does not make one.
Operator review
A flagged event may call for additional observations, updated calculations and closer analysis before anything is decided.
Manoeuvre planning
A manoeuvre must respect orbital mechanics, mission objectives, fuel and operational constraints — and must not create a new close approach.
Operator responsibility
The spacecraft's operator remains responsible for the decision and for every command sent to the satellite.

04The collision-risk workflow

From observation to a reviewed decision.

A conceptual research architecture showing where Ramerion’s interests sit in the chain. It is not an implemented system.

Observe and estimate

  1. InputObservation sourcesRadar, optical and other tracking data
  2. ToolData ingestionCatalogue and operator data, with provenance
  3. CheckQuality & time checksStale, inconsistent or misaligned records flagged
  4. ToolOrbit determinationOrbital state with its uncertainty

Assess and decide

  1. ToolConjunction screeningPairs that may pass close
  2. ToolRisk assessmentCollision probability and encounter geometry
  3. PersonDecision supportAnalyst reviews options and uncertainty
  4. PersonManoeuvre planningOperator plans and authorises; mission systems command↳ no manoeuvre: keep monitoring
  5. CheckReassessmentThe new orbit is screened again
  • Input
  • Tool
  • Check
  • Person
Conceptual research architectureConceptual research architecture — not an implemented Ramerion system. Software in this chain informs people; it does not command spacecraft.

No AI agent issues satellite commands in this architecture. Safety-critical operations need validated systems, proper interfaces to the operator’s mission control, operator authorisation and mission-specific controls. AI assistance, where it is useful at all, sits beside the analyst.

05Research directions

Five directions, from data quality to future sensing.

  1. Orbital data intelligence

    Exploratory research

    Methods for ingesting and validating orbital data: spotting stale or inconsistent records, tracking where each value came from, and helping operators read the uncertainty attached to it.

  2. Collision-risk decision support

    Exploratory research

    Ways to present conjunction events clearly: how close, how confident, how the picture changes with new data, the potential consequences, and the alternative manoeuvre scenarios an operator might weigh.

  3. AI-assisted operations

    Exploratory research

    How AI agents and retrieval systems might help engineers search technical documentation, summarise operational context and explain analytical results.

    AI does not replace validated orbital mechanics or authoritative collision-probability calculations.

  4. Autonomous and resilient operations

    Future possibility

    Longer-term opportunities for more automated planning and decision support, with human authorisation and safety assurance appropriate to each mission.

  5. Quantum technologies for space systems

    Future possibility

    Quantum sensing, precision timing and inertial-navigation research that could one day help spacecraft navigate when satellite-navigation signals are unavailable or degraded.

    Quantum inertial sensors are navigation instruments, not debris detectors, and are distinct from the classical methods used for tracking and collision analysis.

06Scientific limits

What this research does not claim.

Not claimed

  • That quantum radar can detect orbital debris at operationally useful distances.
  • That Ramerion tracks orbital objects, let alone every one of them.
  • That any satellite can be guaranteed never to collide with debris.
  • That quantum computing is required for, or better at, collision prediction.
  • That Ramerion has an operational collision-avoidance product.
  • That Ramerion can remove orbital debris.

Where things stand

  • Practical debris tracking relies primarily on established observation systems, orbital data and tracking techniques.
  • Quantum technologies are potential future enablers in specific roles, such as navigation and timing — not a general answer to space safety.
  • Before any capability could be claimed, a prototype would need validation against established methods and real operational data.

07Further reading

Technical references.

08Related

Collaborate

Working on space safety? We would value the conversation.

We welcome researchers, engineers and operators interested in orbital data, decision support and the long-term sustainability of space activities.

Discuss a research collaborationAll research