An AI-Enhanced Architecture as Code Capability for Automatically Generating SysML Models of Advanced Space Systems

This project aims to develop an AI-enhanced architecture as code capability, Icarus, to automatically generate SysML models from system documentation. Icarus will use natural language processing and machine learning to ingest text-based requirements and architecture definition documents and produce comprehensive SysML diagrams. This will significantly enhance the efficiency, accuracy, and accessibility of model-based systems engineering. Icarus will be validated through testing in internal and Air Force environments, including integration with AFSIM. The dual-use tool is designed for versatility across defense and commercial aerospace applications. Key anticipated benefits are 60x faster model creation versus manual methods, 90% accuracy in capturing system design intent, and 40% gains in engineering productivity. Successful Phase II development will pave the path for expanded Phase III deployment. This project ultimately aims to modernize digital engineering capabilities, enabling more agile development and integration of complex defense systems essential for strategic readiness.