MB.CI: Model-Based Point Cloud Inspection

NAVAIR has created a custom process for managing model-based definition (MBD) parts, linking manufacturing and inspection data to the part model using the Quality Information Framework (QIF) Standard. However, this workflow has three capability gaps that need to be addressed: 1- criticality-based guard banding (CBGB), 2- measurement uncertainty quantification (MUQ) for non-contact articulating arms (NCAA), and 3- optimizing inspection planning to reduce the time and cost of inspecting parts based on their criticality. Global Technology Connection (GTC), in collaboration with the University of Georgia (UGA) and Metrologic Group (MG), is proposing the development of a model-based enterprise workflow (MBEW) that aligns with the QIF MBD. In Phase I, the objective is to address the above mentioned three categories and demonstrate the feasibility of the proposed workflow in a laboratory setting. Furthermore, as part of the Phase I Option, the team will work on the interoperability of the proposed tool chain, alignment with the relevant standards, improvement of the selected ML model in Phase I Base, and integration of the developed software into the MBD tools. In Phase II, the team will improve the new proposed process workflow (MBEW) for the organic industrial base by building upon Phase I solutions. This workflow will integrate with existing manufacturing practices, minimizing deployment challenges. It includes repackaging and deploying the workflow directly on Navy databases. Operations will be integrated into a Digital Enterprise Tool and the workflow will be demonstrated and validated using real Navy data in a prototype manufacturing environment. The team proposes the integration of QIF Model-Based Quality Workflow with the criticality assessment of a product within a MBEW. The innovations of the proposed solution are listed below: Modeling the criticality of a product using FMECA (Failure mode, effects, and criticality analysis) in a unified architecture framework (UAF), and subsequently defining an acceptable level of PoF (probability of failure) for that product Performing an uncertainty-back-propagation from stress level to the geometries and defining a CBGB based on the new acceptable PoF for each measure and Using Bayesian Inference and machine learning (ML) to quantify MU of NCAA Optimizing scanning path and configurations to achieve expected measurement quality considering the feature criticality and measurement uncertainty. Demonstrating feasibility of the proposed solution in a lab environment