The SPHERES Navigation System: from Early Development to On-Orbit Testing

The MIT Space Systems Laboratory has developed the Synchronized Position Hold Engage and Reorient Experimental Satellites (SPHERES) facility for the testing of formation flight and autonomous docking algorithms inside the International Space Station (ISS), in NASA’s reduced gravity aircraft and in a 1-g laboratory environment. To provide SPHERES with reliable and accurate position, velocity, attitude and angular rate estimation, an innovative state estimation system based on ultrasound transmission has been developed. An extended Kalman filter (EKF) processes time-of-flight data collected by ultrasonic receivers, as well as angular rate measurements provided by gyroscopes, to compute the state estimates required by the satellites when maneuvering. To increase the robustness of the system, the EKF has been augmented with a fault detection capability that uses the filter innovation (residual) to diagnose measurement errors. Two versions of the algorithm were successfully implemented and used on the SPHERES facility onboard the ISS during a series of five test sessions, from May 2006 to November 2006. This paper describes both versions in detail, along with difficulties encountered during the implementation on the hardware and their solution. Results from experiments performed in the ISS to validate the algorithms are also presented.