Autonomous vision in space, based on Advanced Stellar Compass platform.

The Ørsted Star Imager, comprises the functionality of an Advanced Stellar Compass (ASC). I.e. it is able to, autonomously solve "the lost in space" attitude problem, as well as determine the attitude with high precision in the matter of seconds. The autonomy makes for a high capability for error rejection and fault recovery, as well as graceful degradation at radiation, false object or thermal loads. The instrument was developed from Concept to Flight Model within 3 Years. The instrument surpasses the initial specifications for all parameters. For Precision, Computational speed and Fault detection and recovery by orders of magnitude. This was accomplished, by the use of advanced high level integrated chips in the design, along with a design philosophy of maximum autonomy at all levels. The instrument tracks ALL stars in the field of view, which enables a variety of applications not normally associated with conventional star trackers. This paper starts by giving a generel decribtion of the Advanced stellar compass. Including it's primary specifications and performance levels. Some of the more promising of the advanced applications are then discussed, along with test-results and methodologies. The diversity of the advanced applications are vast, as depicted by the topics adressed, namely: 1) Detection and Tracking of distant non-stellar objects (e.g. meteors). 2) Delta-V correction, for encounter phases. 3) Tracking of selected Objects (e.g. guidance for other instruments). 4) Mass Estimation via pellet ejection. 5) Complex Object surface tracking (e.g. space docking, planetary terrain tracking). All the above topics, has been realized in the past. Either by open loop, or by man-in-loop systems. By implementing these methods or function in the onboard autonomy, a superior system performance could be acheived by means of the minimal loop delay. But also reduced operations cost should be expected.