The orbit estimator implemented in #305 has been tested sparingly. Unlike #301's concerns with the truth model which focus on the truth model, for the relative orbit estimator we most likely need to focus on:
Sensor and process noise values. The are currently fairly underdeveloped and should be vetted against a better orbit simulation.
What happens when the spacecraft itself fires a thruster and how large of a velocity process noise do we need so the satellites can effectively track one another without communicating thruster firing back and forth?
The orbit estimator implemented in #305 has been tested sparingly. Unlike #301's concerns with the truth model which focus on the truth model, for the relative orbit estimator we most likely need to focus on: