Synaptech LLC — Department of Defense SBIR Phase I: AF203-CSO1

Description

Modeling the behaviors of space threats, new space concepts, and TTPs is difficult, time consuming, error prone, and challenging to understand and communicate.  The current AI modeling and simulation timelines are not consistent with the dynamic space environment.  Synaptech proposes a collaborative customer discovery project to understand the need for AI models in the space domain, identify common pain points amongst stakeholders in the development and execution of their AI models, as well as identify desired software features and capabilities to mitigate the concern. Synaptech also proposes exploring how the gaming industry tackles AI modeling, the lessons learned in game development, and the technologies that may be applied to military applications. Finally, Synaptech will use its domain and simulation experience to propose innovative ideas for solutions to AI modeling problems. For instance, a state machine or behavior tree creation tool that allows a user to build complex satellite behavior graphically and intuitively. The states or actions would be backed by pre-built Python code in Synaptech’s RAPTR simulation and driven by user inputs from the graphical tool. Simplicity is key for understanding the behavior, speeding up the development timeline, and validating the behaviors.