1. The eye that measures the approach

Arthur C. Clarke’s 2001: A Space Odyssey returns as the recalled work because the Discovery’s approach to the monolith, and later to the alien artifact, is a long exercise in relative navigation: knowing position, attitude, and closing rate without touching the thing you are approaching. The desktop’s segments would face the same problem if they tried to rejoin autonomously.

This entry reads about relative navigation and guidance for proximity operations.

2. What relative navigation must provide

Relative navigation estimates the six-degree-of-freedom pose of the chaser relative to the target: three positions and three attitudes. A DLR survey on robotic capture in space states that reliable relative pose information is considered a critical hazard because the safety of proximity maneuvers depends on it.

The estimate must be:

  • Accurate enough for the capture mechanism’s tolerance.
  • Continuous so that brief sensor dropouts do not lose track.
  • Robust to lighting conditions, surface reflectivity, and target motion.
  • Timely so that control can react before the chaser drifts out of the safe corridor.

3. Sensor options

Common sensors for RPOD include:

  • GPS and radio: used at longer ranges and when both spacecraft are cooperative. GPS-based relative navigation works well when both vehicles carry receivers.
  • Laser rangefinders and flash lidar: provide direct range and sometimes range-rate. A NASA paper on imaging flash lidar describes how flash lidar can provide relative range, velocity, and bearing for autonomous rendezvous and docking.
  • Cameras and vision: used for final approach and pose estimation. Vision-based systems can use natural features or artificial fiducial markers.
  • Radar: useful when lighting is poor or when the target is non-cooperative.

The choice depends on range, target cooperation, lighting, mass, power, and compute budget.

4. Cooperative and non-cooperative targets

A cooperative target carries markers, transponders, or reflectors that help the chaser navigate. A paper on spacecraft fiducial markers describes AstraTag, a recursive marker designed for autonomous on-orbit operations, and compares it against ArUco and AprilTag on curved surfaces.

A non-cooperative target has none of these aids. Relative navigation must estimate pose from natural features, lidar returns, or radar images. This is harder and is the subject of active research for debris removal and servicing missions.

5. Guidance laws

Guidance generates the desired trajectory for the chaser. Common approaches include:

  • Line-of-sight guidance along a fixed approach corridor.
  • Glideslope guidance that controls the closing rate as a function of range.
  • Natural motion trajectories that exploit orbital dynamics to reduce fuel use.
  • Model predictive control that optimizes the trajectory subject to constraints.

The guidance law must work with the navigation filter and the actuator limits to keep the approach safe.

6. What this changes

  • Relative navigation is a sensor-fusion and estimation problem, not just a sensor problem.
  • Cooperative markers simplify final approach but require the target to be prepared.
  • Non-cooperative navigation is harder and is the frontier for servicing and debris removal.
  • The next entry will read about docking and berthing mechanisms.