Principal investigator: Hui Yang
University: The Pennsylvania State University
Industry partners: Argolytics, LLC
Modern manufacturing enterprises are investing in advanced sensing modalities such as optical metrology and computed tomography to cope with system complexity and increase information visibility. Real-time sensing now brings the increasing availability of imaging data for quality inspection and process improvement. For example, the quality of a machined surface profile is pertinent to cutting tool set up and conditions. In addition, layerwise imaging is critical to reducing the variations in the laser powder-bed-fusion process. However, traditional statistical quality control (SQC) is not concerned with imaging data but instead with key product or process characteristics, and is limited in its ability to readily address complex data structures. Dealing with 2D and 3D imaging data is a general problem facing both traditional and next-generation innovation practices in low-volume and high-mix manufacturing. Realizing the full potential of imaging data for quality control requires fundamentally new methodologies to harness and exploit complexity. The objective of this research project is to design and develop a new framework of image-guided quality control for advanced manufacturing. The developed methodology will be evaluated and validated with real-world data from additive manufacturing, biomanufacturing, and precision machining. This project addresses the complex structures in the high-dimensional image streams for in-situ monitoring and control of manufacturing processes, which enables real-time quality inspection, defect mitigation, and process improvement. The developed image-guided SQC toolbox will increase the competitive advantages of Argolytics in the market of quality management software. In turn, this project will help PA manufacturers minimize efforts required towards obtaining the best quality product, thereby spurring the growth of advanced manufacturing in the state of Pennsylvania and the nation.