In general, autonomous rendezvous and docking requires that two spacecraft start at a remote distance (i.e., out of sight of each other), come together into a common orbit, rendezvous, dock, and control the new combined spacecraft in both orbit and attitude. Doing this requires developing and testing a variety of new technologies including absolute and relative autonomous navigation, autonomous rendezvous and docking hardware and software (both sensors and actuators), and autonomous control of a "new" spacecraft with different mass and inertia properties than either of the two original spacecraft. While these are very workable technologies, they do require a significant change in mindset -- turning over control of thrusters and other actuators to an on-board computer. While there is substantial potential for cost savings, risk reduction, and new mission modes by use of these technologies, there is a very strong reticence to allowing operational spacecraft to control their own destiny, particularly in firing thrusters.
This paper summarizes work at Microcosm and elsewhere in each of the above technologies. Autonomous navigation and absolute orbit control have been demonstrated on orbit. In conjunction with Michigan Aerospace, autonomous rendezvous and docking hardware and algorithms have been demonstrated in parabolic flights and zero-g simulations. Approaches have been proposed for more precise and robust autonomous navigation and autonomous on-orbit estimation of combined mass and inertia properties, leading to efficient orbit and attitude control of the combined spacecraft. Many of these technologies can be tested at low cost in parabolic flights, suborbital flights, and evaluation of data from existing or planned missions. Thus, a "coordinated attack" on the complete problem of fully autonomous rendezvous and docking is both feasible and potentially very low cost.