Co-authors: Kyle Devlugt, Brian Au, Joe Nicolino, Alex Caviness, Sam Ludwig, Alex Blum, and Adam Jaycox
Manufacturing design, which includes fabrication and inspection of parts, is traditionally manual, disjointed, and results in rigid process workflows. Often different processing steps and measurements are executed in a non-collocated and non-automated fashion with data that is equally scattered, if collected. Furthermore, these activities may require sustained manual oversight and intervention. Thus, the development cycle (e.g. part specification, fabrication, and qualification) is subject to bottlenecks, making part repeatability difficult and costly to achieve, quantify, and optimize. To address these issues, our team is standing up hardware and software to achieve Autonomous Multimodal Manufacturing Optimization, where we take an object-oriented approach to manufacturing development and boutique manufacturing. We use our modular manufacturing cell wherein cradle-to-grave data is collected from sensing, digital twins, and inspection. In addition to some early stage work, we present some of our modules, in terms of data archiving analyzing part files to leverage experiential knowledge in a formulaic and process general way, digital twins of fabrication and inspection of fused deposition modeling parts, and applied machine learning use cases that leverages these data streams.
This work was performed under the auspices of the U.S. Department of Energy by LLNL under Contract DE-AC52-07NA27344, LLNL-ABS-823029.
Speaker Biography: Dr. Giera is a member of the technical staff in the Materials Engineering Division under the Engineering directorate at LLNL. He is currently a Principal Investigator for the Autonomous Multimodal Manufacturing Optimization project and technical lead on several advanced analytics and additive manufacturing projects. This work focuses on developing and applying physics-based and machine learning models to a variety of advanced manufacturing systems (e.g., extrusion, photopolymerization-based, powder bed fusion, electrophoretic deposition, etc.) and complementary process monitoring and part inspection platforms. Collectively, these capabilities are aimed at achieving automation, scale up, defect minimization, and process understanding.
Panel Members:
Dr. Stephen Welch, VP of Data Science, Mariner
Dr. Trichy Pasupathy, Senior Manager, Metrology, Confluent Medical Technologies
Dr. Michael Sharp, Reliability Engineer, Systems Integration Division, NIST
Moderator: Harish Cherukuri
2020 Annual Meeting (Canceled)