Jason Jamerson

The NASA Michoud Assembly Facility Digital Twin is a major industrial simulation and advanced-manufacturing initiative centered on NASA’s two-million-square-foot rocket factory in New Orleans. Supported by $7.5 million in congressional appropriations to Louisiana State University and developed in partnership with NASA and aerospace manufacturing stakeholders, the project establishes a digital foundation for understanding, operating, and transforming one of the nation’s most consequential manufacturing facilities.

The project encompasses two distinct but interconnected digital environments. A controlled unclassified information (CUI) environment incorporates high-resolution LiDAR scanning, building information models, equipment data, and operational information for internal use. A parallel non-CUI environment is constructed from publicly available references and hand-authored digital assets, allowing the team to develop interfaces, demonstrate capabilities, conduct research, and share selected aspects of the work without exposing protected facility information.

Together, these environments represent the architecture, tooling, production systems, aerospace vehicles, infrastructure, and workflows of Michoud. The digital twin connects spatial models with asset catalogs, engineering metadata, sensor feeds, real-time location systems, and operational records. This creates a navigable, interactive system through which users can locate equipment, inspect conditions, understand relationships among systems, review process information, and evaluate proposed changes within the context of the factory.

The platform is built around Unreal Engine and a secure cloud-based architecture that supports high-fidelity visualization across desktop, mobile, immersive, and streamed interfaces. Users can move through the facility, access associated BIM and engineering information, review live and historical data, and interact with complex manufacturing information through a unified spatial interface. Augmented reality, mobile access, and immersive visualization extend the digital twin beyond a conventional facility model and make it available as an operational, training, planning, and decision-support environment.

Artificial intelligence provides an additional layer of access to the system. An integrated language-model interface allows users to query and cross-reference information from the asset catalog, sensor network, facility metadata, and associated documentation. The system can identify relevant assets, surface contextual information, and transport users directly to corresponding locations within the three-dimensional environment. This approach turns a large collection of models and databases into an accessible institutional knowledge system grounded in the physical organization of the factory.

The project also establishes a scalable method for managing digital-twin fidelity. Concept-level environments support rapid exploration and communication. Review-level models allow stakeholders to examine workflows, layouts, and operational scenarios. Decision-grade environments bring together validated geometry, engineering information, and process data for higher-consequence planning and analysis. This tiered approach directs resources toward the level of precision required by each decision while preserving a coherent facility-wide platform.

The next phase of the work advances beyond general facility representation toward targeted production acceleration and the factory of the future. Current development centers on core-stage assembly, friction-stir-welding support workflows, tooling and material movement, inspection, workforce training, and human-robot collaboration. These scenarios position the digital twin as an environment for rehearsing future production systems before they are introduced into the active factory.

This work supports NASA’s transition toward a more digitally integrated, adaptable, and robot-collaborative manufacturing model. It provides a platform for evaluating how people, machines, facilities, data, and intelligent systems operate together. The long-term objective is to reduce production friction, strengthen institutional knowledge, accelerate training and planning, and create a repeatable digital infrastructure that can support future NASA and aerospace manufacturing programs.

The Michoud project has become a foundation for LSU’s broader leadership in digital twins, industrial simulation, embodied robotics, and advanced visualization. The methods developed through the project now inform work across aerospace, chemical manufacturing, infrastructure, coastal systems, medicine, cultural heritage, and human performance.

All visual material presented on this page is derived from publicly available information and the project’s non-CUI development environment.