AI-Driven Smart Manufacturing

A novel approach to enhance Additive Friction Stir Deposition (AFSD) is presented through the integration of Multimodal Retrieval Augmented Generation (RAG). AFSD which is a solid-state metal additive manufacturing process characterized by its complexity, necessitating advanced analytical tools. A multimodal RAG system is implemented, integrating textual, and  visual data. The methodology involves text extraction using LangChain's PyPDFLoader, followed by chunking and embedding generation via Google's Generative AI model. ChromaDB is utilized for vector storage and efficient information retrieval. The system, powered by Google's Gemini large language model, demonstrates proficiency in generating detailed explanations of friction-based deposition processes, identifying manufacturing techniques from visual data, and evaluating material microstructures. Information from multiple sources is synthesized to produce context-aware responses. Process control, quality assurance, and decision-making in AFSD are significantly enhanced by this approach. The system's capabilities are illustrated through the analysis of various friction-based deposition techniques and microstructure evaluation.

Paper Link: https://www.preprints.org/manuscript/202410.0099/v1

We are creating an innovative manufacturing process optimization framework with deep learning approaches. Our collaborative research focuses on surface properties and material removal rates in precise manufacturing. The project uses unique neural network architectures to predict and optimize process parameters, while wear maps help to visualize and understand the complicated correlations between numerous machining factors and surface quality. Through the application of contour-based visualization approaches with dynamic transitions, we can gain a better understanding of parameter interactions and their consequences on machining processes.

Stay tuned for more updates! 

Machine learning is becoming increasingly important in the numerical modeling of solid-state additive manufacturing (SSAM) processes like Wire Arc Additive Manufacturing (WAAM) and Friction Stir Based Additive Manufacturing. This is because SSAM involves complex, dynamic interactions between various process parameters (e.g., tool speed, force, geometry), material properties (e.g., flow stress, thermal conductivity), and the resulting microstructure and mechanical properties. Traditional numerical models often struggle to accurately capture these interactions due to the challenges in modeling material behavior at high strain rates and temperatures, complex geometries, and evolving microstructures.

We are currently working to enable more accurate predictions of key outcomes like residual stress, distortion, and microstructure, leading to improved process optimization, reduced experimental costs, and accelerated development of new SSAM processes and materials. 

Stay tuned for more updates!