Autonomous navigation for deep space missions is a unique capability that was developed at JPL as a technology demonstration on the Deep Space 1 mission in the late 1990s.  This system, called AutoNav, enabled a spacecraft to navigate itself without the need for ground radio contact collecting radiometric Doppler and range measurements, which is the standard method of spacecraft navigation.  AutoNav used optical bearing measurements to natural bodies, specifically asteroids, to perform orbit determination, from which it computed low-thrust control parameters to steer the ion engines on the spacecraft for orbit control.   For several months, the onboard system demonstrated the ability to determine its own position and steer the spacecraft.  In September 2001, it was also used during the fast flyby of comet Borelly to track the nucleus of comet. AutoNav was then flown successfully on two subsequent comet intercept missions, Stardust and Deep Impact; the former to maintain visual lock on the nucleus on its flybys of  comets Wild-2 and Tempel 1, the latter for terminal guidance to impact the nucleus of Tempel 1 on one spacecraft, as well as maintain visual lock on the nucleus for the flyby spacecraft (and later used on the encounter with comet Hartley 2).    

Although AutoNav was executed successfully multiple times, the software is now over 25 years old, and there is a strong desire to develop the next generation system that incorporates the latest performance capabilities of onboard processors and improvements in flight software practices and principles.  Over the past several years, we have been developing this next generation system, borrowing core lessons learned from the heritage system but using algorithms and practices from the gold-standard ground navigation system called MONTE.  This system, now renamed AstroNav, was built from the ground up to be more modular and adaptable to a wide range of scenarios encountered by deep space missions.  In addition to using optical data, AstroNav will also be able to process one-way uplinked radiometric Doppler and range data when coupled with a high accuracy atomic clock.  Furthermore, any potential other sources of data, such as pulsars, or lidars when in close proximity to a small body, will simply be plug-in modules that can be incorporated into AstroNav.  The core computational elements are merged with a higher level layer that forms the interface between these elements and the other elements of flight software, such as attitude and thruster information from the attitude control system, and telemetry.  This is being built on the F’ flight software layer developed at JPL, and used on a number of smallsats and cubesats.  The intention is to have high portability for the core functional capabilities without changes, thus building up heritage for its use as more and more missions adopt it, and customize the interface layer for any specific spacecraft.

The core of AstroNav is the software which performs navigation, but AstroNav has options to include hardware as well.  Currently, we are building the system on a PolarFire Icycle board with VxWorks operating system.  The sensor suite includes two cameras, a narrow angle camera (NAC) with roughly 1-2 degrees field-of-view (FOV), and a wide angle camera (WAC) with 20-30 deg FOV.  Radiometric tracking data is obtained through an IRIS radio, with the clock driven by a Miniature Atomic Clock (MAC).  In the full configuration of AstroNav, this entire software and hardware package would be a payload with approximately 9 kg of mass, drawing a peak of about 60 W of power. 

The current plan is to perform a technology demonstration of the capability on an upcoming commercial mission, with the prime option being the CAPSTONE-2 spacecraft, scheduled to be launched in mid 2027.  We are also pursuing other options with other commercial companies, as well as responding to technology calls from NASA. 

AstroNav is an enabling capability for many of NASA’s science driven missions exploring the outer planets; one example is a mission to fly through the plumes of Enceladus which requires rapid turnaround of navigation solutions.  It is also necessary for planetary defense applications, such as ion beam deflection or gravity tractoring.  Commercial spacecraft can also use it to minimize the use of ground antennas.