Working notes on a developmental approach to autonomous spacecraft and the recovery of the Hubble Space Telescope

At NASA's Goddard Space Flight Center (GSFC), we have been developing the means to create space and surface systems that are active participants in their environment rather than being merely visitors that withstand space's hazards as our extended remotely controlled tools. Central to this work has been the development of the Autonomous Nano-Technology Swarm (ANTS) mission architecture and the Neural Basis Function/Synthetic Neural System (NBF/SNS) which are included among the subjects of several GSFC provisional patent applications. These are scalable systems with non-linear dynamics built in to deal with irregularity, uncertainty, and unpredictability in their environments. NBFs are a synthesis of low- and high-level approaches to system control in which specialized components are embedded in an Evolvable Neural Interface (ENI) that mediates information flow through the system. Individual system behaviors are driven by individual NBFs, and more complex behaviors are composed by linking multiple NBFs through the ENI. Conflict resolution and ontological mapping between NBF components occurs primarily within the ENI which is genetically evolved and trained to perform these resolutions and mappings. This built-in flexibility is especially important for space systems which currently operate only within very narrow margins of safety. The first NBF is being developed to demonstrate that independent reactive low-level attitude and propulsion control and a heuristic high-level navigation executive can