Robotic Testbed for Rendezvous and Optical Navigation: Multi-Source Calibration and Machine Learning Use Cases
This work addresses the need for high-fidelity validation of machine learning in space navigation, representing an incremental advancement by applying existing calibration and CNN methods to a new robotic setup.
The researchers tackled the problem of validating machine learning algorithms for spaceborne optical navigation by developing the TRON robotic testbed, which achieved millimeter-level position and millidegree-level orientation accuracies in pose estimation and revealed a performance gap in a CNN when tested on simulated versus synthetic images.
This work presents the most recent advances of the Robotic Testbed for Rendezvous and Optical Navigation (TRON) at Stanford University - the first robotic testbed capable of validating machine learning algorithms for spaceborne optical navigation. The TRON facility consists of two 6 degrees-of-freedom KUKA robot arms and a set of Vicon motion track cameras to reconfigure an arbitrary relative pose between a camera and a target mockup model. The facility includes multiple Earth albedo light boxes and a sun lamp to recreate the high-fidelity spaceborne illumination conditions. After the overview of the facility, this work details the multi-source calibration procedure which enables the estimation of the relative pose between the object and the camera with millimeter-level position and millidegree-level orientation accuracies. Finally, a comparative analysis of the synthetic and TRON simulated imageries is performed using a Convolutional Neural Network (CNN) pre-trained on the synthetic images. The result shows a considerable gap in the CNN's performance, suggesting the TRON simulated images can be used to validate the robustness of any machine learning algorithms trained on more easily accessible synthetic imagery from computer graphics.