A Hybrid End-to-End Architecture that uses Model- and Rule-Based Inferencing

As a key component within Remote Agent, Livingstone has the most maturity and potential for commercialization. ICS is proposing to integrate Livingstone into an existing end-to-end architecture for real-time command and control. This flight proven architecture has been demonstrated recently with the NASA FUSE program. Adding the Livingstone component would fill a void in the SCL architecture and would significantly augment the capability of the spacecraft or rover for onboard autonomy. A significant deviation from the expected behavior of the model would be flagged to the inference engine. This would add an enhanced flexibility and versatility of the system in allowing the user or operator to use dynamic scripts and rules in order to complement a model and support progressive autonomy as necessary.ICS believes that the two technologies are highly complementary and would greatly benefit autonomous operations. The key to success is in the tight integration of those technologies into a flight proven environment and supported by an integrated ground development system. This includes the insertion of Livingstone in the flight architecture but also addresses the modifications of the ground development tools to merge the model-based and rule-based languages.