Timeline
The project began in August 2022 and continues to progress, leveraging cutting-edge technology to optimize aerospace engineering and education.
Background
The Digital Twin Project initially focused on creating an intricate 3D environment of NASA’s Michoud Assembly Facility using Unreal Engine. When I joined the project in August 2023, the team had already laid the groundwork for a realistic simulation of the facility. For the first year, the primary effort was on building a detailed virtual environment and integrating simulations to mirror real-world conditions.
The next phase involved expanding the project to include user interaction capabilities, as originally planned. The initial approach was to design the user interface directly within the simulation, allowing users to engage with the digital twin seamlessly. However, after multiple rounds of user testing and feedback, we identified limitations in accessibility and user experience. This insight led us to pivot towards developing a web application that would provide an intuitive interface for interacting with the digital twin while also offering access to additional features and functionalities beyond the simulation.
This transition enabled a more flexible and scalable platform, allowing users to explore the virtual twin, analyze data, and interact with the system remotely.
I joined the project on August 9, 2023, as part of my Graduate Assistantship at LSU. At present Leading the product design team and contributing to the ongoing efforts in virtual simulation development.
User Research
Competitive Analysis, User Interviews, Journey Mapping
UX Design
Sketching, Low & high fidelity wireframes, Interaction Design, Prototyping, Usability Testing
Product Management
Project Coordination, Requirements Gathering, Stakeholder Communication, Sprint Planning, Cross-Functional Collaboration, Prioritizing Features, Quality Assurance
Planning & Research
The initial plan was to develop the user interface within the simulation using Unreal Engine, enabling users to engage intuitively with the digital twin. To validate this approach, we conducted usability tests with a small sample of users in a controlled test environment. The early feedback was promising, so we moved forward with a production deployment. However, once the platform was opened to a broader audience, we identified significant limitations in accessibility and overall user experience.
These findings prompted us to rethink our strategy, leading to a comprehensive UI redesign. We embarked on user research, conducting detailed interviews and observational studies to gather insights into user behaviors, pain points and requirements. The collected data was synthesized to create a detailed user journey map, outlining key stages, actions, and opportunities for enhancement. To brainstorm innovative solutions, we held collaborative ideation sessions, generating a high volume of concepts. This was followed by creating quick sketches and low-fidelity wireframes to visualize and refine our ideas before moving into the design phase.
Design & Prototyping
The user insights gathered during our research phase played a crucial role prompting us to pivot from designing the UI within the simulation environment to developing a web-based application. This change was driven by the realization that users found it challenging to interact comfortably with the gamification-style interface within the Unreal Engine simulation. To address this, we transitioned to mid-fidelity prototypes to translate the wireframes into a more interactive, user-friendly web interface.
To ensure that our design was truly aligned with user needs, we established a focus group consisting of potential core users of the digital twin. This group provided continuous, real-world feedback, which was invaluable in refining our prototypes. We adopted an iterative design approach, conducting this process in three distinct phases. Each phase involved gathering feedback, identifying pain points, and incorporating enhancements to create a more robust and intuitive user experience. Through this approach, we ensured that our prototypes were thoroughly tested and aligned with user expectations before moving into the Testing phase.
Testing & Optimization
After finalizing the design, the implementation was rolled out in the User Acceptance Testing (UAT) environment to perform extensive testing across multiple devices and platforms, ensuring optimal compatibility and performance. We executed a series of test cases designed to evaluate functionality, usability, and system responsiveness. Testing was conducted with our focus group and project team members to gather diverse perspectives, identify potential functional gaps, and validate the design against real-world scenarios.
By leveraging comprehensive feedback from these tests, we were able to pinpoint critical issues, resolve identified bugs and refine the application to enhance user experience. This iterative process was essential in optimizing the platform before its deployment to production, ensuring that any performance bottlenecks or usability challenges were resolved. The feedback loop allowed us to deliver a robust, polished product that meets user expectations and project goals.
Deployment
Following extensive testing and optimization, we moved forward with the deployment phase, transitioning the web application from the UAT environment to the live production server. This involved performing a series of pre-launch checks, including system integration, data migration, and server performance optimizations to ensure a seamless transition.
Prior to the full-scale launch, we conducted a controlled rollout to a select group of users, allowing us to monitor system behavior under real-world conditions and gather final feedback. This phased approach ensured that any last-minute issues were promptly addressed. Once stability was confirmed, the application was fully deployed, delivering a robust and user-friendly application. The successful launch marked a significant milestone in delivering a scalable solution that aligns with the project’s long-term vision.
Measurement Mode
This feature enables users to precisely measure distances within the digital twin environment. It serves a critical function in assessing whether large equipment can be maneuvered or placed in designated areas within the NASA Michoud Assembly Facility. Users can measure the dimensions of rooms and compartments to confirm whether large equipment can be transported through specific passages or if it will fit into predefined spaces within the facility. This capability allows for virtual trial runs, reducing risks before executing real-world installations.
Interactive Mode
This mode allows users to engage dynamically with 3D assets within the digital twin, visualizing various simulations of equipment performance in real-life scenarios. By interacting with the 3D assets, users can explore different operational states and behaviors, enhancing their understanding of machinery capabilities. This helps in scenario planning and optimizing equipment usage within the facility.
Asset Relocation Mode
This newly implemented feature facilitates the repositioning of 3D assets within the simulation. Users can replicate the actual tasks of moving equipment around the facility, providing a risk-free way to plan logistics and spatial management. By virtually simulating the relocation process, users can optimize workflows and prevent disruptions during physical implementations.
Teleport Mode
Given the vast 2-million-square-foot footprint of the Michoud Assembly Facility, the Teleport Mode enables users to swiftly navigate the digital twin environment. Users can either input specific aisle numbers or utilize an integrated map to instantly transport to desired locations, thereby streamlining accessibility and enhancing productivity.
Data Catalog
This feature provides access to a comprehensive repository of metadata, allowing users to retrieve detailed information on any asset within the simulation. Users can either input specific aisle numbers or utilize an integrated map to instantly transport to desired locations, thereby streamlining accessibility and enhancing productivity.
AI Assistant
The AI Assistant enables users to interact with both the Digital Twin and Digital Thread components. Users can locate, inspect, and retrieve data on any asset simply by typing its name. The assistant also allows users to explore the digital thread by specifying the artifact of interest. Beyond data retrieval, it can analyze diverse artifacts, including images, videos, and documents. Additionally, it supports predictive analytics by estimating potential losses in production and providing real-time insights for process optimization.
Digital Thread
The Digital Thread offers a visual representation of the interconnections between various components within the facility. It links each asset to its constituent parts and related materials. Users can select an asset (e.g., a golf cart) and view its entire hierarchy, including individual parts like bolts, wires, and coatings, providing a comprehensive understanding of its composition and dependencies. This is particularly useful for maintenance and quality assurance.
Sensor Mode
This mode provides real-time environmental monitoring, displaying metrics such as temperature, humidity, and dew point across different areas of the facility. Users can assess environmental conditions to ensure they meet the operational standards required for sensitive equipment, minimizing risks associated with fluctuations in ambient conditions.
The implementation of the digital twin has significantly enhanced the user experience, transforming how users interact with the Michoud Assembly Facility virtually. By shifting from game simulation UI to a web-based platform, we improved accessibility and increased application usage by 45%. The introduction of specialized features, such as AI-driven asset search, teleportation, and real-time environmental sensors, resulted in a 40% increase in operational efficiency during virtual planning sessions. User feedback demonstrated a 90% satisfaction rate citing smoother navigation, better spatial awareness, and greater confidence in conducting simulations before executing physical tasks. These metrics reflect our commitment to refining digital interactions for real-world applications, ensuring precision and reducing trial-and-error efforts in critical aerospace operations.