Experimental Performance Assessment of a CubeSat-Based Multisensor Pose Estimation Module

The increasing number of small-satellite missions motivates efforts to demonstrate in-orbit servicing (IOS) capabilities using compact, low-cost solutions to improve the sustainability of the Earth orbit environment. In this context, this paper presents a relative positioning (RELPOS) module for relative navigation, integrating a monocular camera and a laser range finder operating in the near-infrared band within two CubeSat units. The paper presents the functional and software architecture of the module and discusses the design, development, and calibration of an experimental setup to assess its performance. Supervised and unsupervised hardware-in-the-loop tests show that the pose determination algorithm of the RELPOS module can perform both pose acquisition and tracking of a cooperative target equipped with retroreflective markers. The supervised tests show that pose acquisition can be performed with 3σ values on relative positioning errors of 0.12 cm cross-boresight and 1.11 cm along-boresight; attitude errors show 3σ of 4.29 deg cross-boresight and 0.33 deg along-boresight. The unsupervised tests confirm the results of the supervised ones, showing RMSE on marker reprojection error between 0.68 and 1.43 pixels over target-chaser separations spanning from 6.01 to 0.80 m. The module can perform target acquisition within seconds and object tracking at a minimum rate of 2 Hz.