REWAS 2022: Automation and Digitalization for Advanced Manufacturing: Use of Artificial Intelligence for Improved Process Control & Optimization
Sponsored by: TMS Extraction and Processing Division, TMS: Recycling and Environmental Technologies Committee, TMS: Process Technology and Modeling Committee
Program Organizers: Elsa Olivetti, Massachusetts Institute of Technology; Alexandra Anderson, Gopher Resource; Mertol Gokelma, Izmir Institute of Technology; Camille Fleuriault, Eramet Norway; Kaka Ma, Colorado State University
Wednesday 8:30 AM
March 2, 2022
Room: 212A
Location: Anaheim Convention Center
Session Chair: Alexandra Anderson, Gopher Resourece; Luca Montanelli, Massachusetts Institute Of Technology
8:30 AM Introductory Comments
8:35 AM Keynote
Digitalizing the Circular Economy (CE): From Reactor Simulation to System Models of the CE: Markus Reuter1; Neill Bartie1; 1SMS Group
Embracing the circular economy, this paper will discuss recent work that scales reactor technology to system models with relevant digital twins.By the combination of different tools and methods (from AI, CFD, mass- and heat transfer, kinetics, industrial experience etc.) and integrating these into suitable digital platforms this paper will analyze different circular economy systems also in terms of complete supply chains. With a focus on for example thermoeconomics, exergy dissipation of the systems will be quantified by suitable digital twins for PV cell manufacture, battery technology etc. Integration with impact assessment approaches will show how to minimize the impact of complete supply chains and show which systems produce the lowest footprint products. In addition the link of the digital twins to the sustainability development goals (SDGs) of the United Nations will be elaborated on.
9:15 AM
AI/Data Mining in Materials Manufacturing: Elsa Olivetti1; 1Massachusetts Institute of Technology
Presently commercialized technologies, silicon photovoltaics and lithium-ion batteries, have required decades to reach even modest levels of global adoption. One among many factors contributing to this protracted transition to large scale manufacturing is that the processes and equipment used in lab diverge from those which are available or required at production scales. When novel materials or devices proceed from the laboratory to manufacturing, unforeseen and unfamiliar processing challenges can easily arise. Fundamental scientific and design problems that could be solved during early-stage research with comparatively little time and expense must instead be addressed during scale up, compelling the use of significantly more capital at a time when such delays are ill-afforded and in an industry in which low-cost incumbents drive perilously thin margins for new entrants. This presentation will focus on data-driven insight and tools for early research to design new technologies explicitly for manufacturing and scale up from inception.
9:35 AM
Factors to Consider when Designing Aluminium Alloys for Increased Scrap Usage: Luca Montanelli1; 1Massachusetts Institute of Technology
For a significant shift in alloy design to happen, the aluminium alloy industry needs to explore a broader range of compositional and processing dimensions. The proposed project will investigate unexplored regions of the compositional space to guide the design of new alloys, especially where opportunities are present to improve recyclability. To achieve this, a blending model will be optimised over compositional space to inform alloy compositions that enable higher quantities of scrap use. Blending models inform on the scrap usage of a candidate alloy when it is set in a predetermined market landscape. Due to their computational cost, machine learning optimisation methods such as Bayesian optimisation will be employed to focus the design process. The optimisation will be subject to constraints that are based on compositions, phase combinations, and relevant properties to ensure that the alloys not only maximise amount of scrap use but also meet technical requirements.
9:55 AM Break
10:15 AM Invited
NOW ON-DEMAND ONLY - An Automated Recycling Process of End-of-life Lithium-ion Batteries Enhanced by Online Sensing and Machine Learning Techniques: Liurui Li1; Maede Maftouni1; Zhenyu Kong1; Zheng Li1; 1Virginia Polytechnic Institute and State University
The End-of-life (EOL) lithium-ion batteries (LIBs) are hazardous and flammable with various sizes and shapes, which create significant challenges to automate a few unit operations (e.g., disassembly at the cell level) of the recycling process. In this work, we demonstrate an automatic battery disassembly platform enhanced by online sensing and machine learning technologies. Computer vision is used to classify different types of batteries based on their brands and sizes. The real-time temperature data is captured from a thermal camera. A data-driven model is built to predict the cutting temperature pattern and the temperature spike can be mitigated by the close-loop control system. Furthermore, quality control is conducted using a neural network model to detect and mitigate the cutting defects. The integrated disassembly platform can realize the real-time diagnosis and closed-loop control of the cutting process to optimize the cutting quality and improve safety.
10:45 AM
Steel Production Efficiency Improvements by Digitalization: Markus Schulte1; Bill Emling1; 1SMS Group
Digitalization supported by process knowledge drives the holistic improvement process for advanced steel producers. Four digital focus areas have been identified where digital applications are most impactful for an efficient steel production process: Asset Optimization, Product Quality, Energy and Sustainability as well as Production Planning. Applications in those areas are fed with data from sensors and automation systems, which are combined and preprocessed in a centralized data center. AI & ML help to find pattern in the data that allow forecasting of certain events before they occur. This makes it possible to react on undesired upcoming events and avoid their occurrence or mitigate their risk by suggested counter measures which can be directly executed via the automation systems. Several use case in those areas have been realized which increase efficiency and create value for steel producers.
11:05 AM
Development of Virtual Die Casting Simulator for Workforce Development: John Moreland1; Kyle Toth1; John Estrada1; Junyi Chen1; Na Zhu1; Chenn Zhou1; 1Purdue University Northwest
High pressure die casting is a complex manufacturing process that requires a highly developed work force. A virtual die casting machine has been developed for operators to provide a better understanding of how the machine works and how to deal with a variety of practical situations and issues that arise on the shop floor. Computational fluid dynamics (CFD) simulations have also been developed and integrated into the simulator to help die casters understand how parameters such as shot speed can affect the resulting quality of castings being produced. A virtual melter furnace is also being developed to learn and practice maintenance and safety procedures. The simulator was developed for virtual reality (VR) headsets and controllers, but is also usable on standard PC with mouse and keyboard. Development methodology and overview of simulator functionality will be discussed.