Enhancing the MAIT of Aerospace Systems Through AI-Based Immersive Technologies

The aerospace industry is undergoing a significant transformation through the adoption of augmented reality (AR) technologies that overlay digital information onto the physical world, enhancing human-machine interactions in complex assembly processes. This approach provides real-time visualizations, step-by-step instructions, and interactive overlays, which can significantly reduce errors, shorten learning curves, and improve the efficiency and quality of assembly tasks in the Manufacturing-Assembly-Integration-Testing (MAIT) loop. Despite the promising potential of these technologies, existing solutions are often developed for specific applications and lack a standardized, adaptable framework that can be easily deployed across different scenarios. Driven by advancements in immersive technologies and AI-based methods, AR integration in assembly processes has gained substantial attention for its ability to enhance operator performance and task accuracy. Several studies highlight the benefits of AR, including its capabilities to provide interactive guidance during complex tasks, which expedites the learning curve for shopfloor operators and improves overall task efficiency. However, most tools implemented to date are ad-hoc solutions tailored to specific applications, and the development of a generic, adaptable framework remains a challenge. To address this gap, our approach includes a Domain-Specific Language (DSL) designed to streamline the configuration and customization of AR-assisted assembly processes. The DSL facilitates the definition of tasks, components, and tools, enabling rapid adaptation to various assembly scenarios without extensive reprogramming. This approach emphasizes reducing computational overhead on AR headsets by distributing processing tasks across different system components, thereby enhancing the quality of life for operators by optimizing the interaction between virtual and physical environments. This work aims to demonstrate a functional architecture and related methodologies that could be implemented to enhance the integration of AR technologies in assembly tasks. One of the key elements is the use of an orchestrator that manages the data flow between AR interfaces, 3D model management, and sensor integration. By focusing on these aspects, the framework seeks to provide a flexible and adaptable approach that supports operators in performing complex assembly tasks with greater accuracy and efficiency. The goal is to highlight an example of a functional architecture and the methodologies behind it, underscoring the potential for AR technologies to revolutionize aerospace assembly processes through thoughtful design and integration strategies. Preliminary tests using a CubeSat engineering model demonstrate the system’s effectiveness, and future work will explore leveraging AI and large language models to further automate the configuration of assembly processes, enhancing scalability and adaptability in various aerospace applications.