Configuring the Deep Impact AutoNav System for Lunar, Comet and Mars Landing

JPL’s autonomous onboard optical navigation system – AutoNav which has been responsible for obtaining all of NASA’s close-up images of comet nuclei, is being expanded in capability to accomplish a host of much more advanced missions than even the spectacular impact with comet Tempel-1 on July 4, 2005 (Figure 1). Among the most important of these mission scenarios are those that require precision landing on the Moon, a comet, and Mars. Lunar missions, in particular, are high on NASA’s agenda; both manned and unmanned. AutoNav is being utilized at this stage as a platform for evaluating lunar landing navigation strategies, and for testing algorithms, and indeed is being extended to become an onboard AutoGNC (Guidance Navigation and Control) system, that is an integrated inertial position and attitude determination function. At the core of these strategies is the nature and form of the optical navigation techniques to be used. AutoNav is being integrated with a comprehensive system for surface modeling and landmark tracking called OBIRON (OnBoard Image Registration for Optical Navigation). OBIRON provides both the means of simulation and onboard processing for purposes of estimating spacecraft position and attitude. This paper will describe the system architecture of the AutoGNC landing configuration, and also the simulation platform that provides simulated sensor inputs to the AutoGNC system. During the landing, performance of the system was excellent, with meter-level knowledge and control achieved. This paper will provide an evaluation of these results, and the associated system and process, from trial simulated operation of the system during landings at the Moon and on a small body.