Research
Research Thrust 1: System Optimization with Simulation
Today’s manufacturing system is observed to be interdependent and constantly emerging, in fact acting as an ecosystem. To design a smart factory, a translation is required from a physical to a digital domain, where the physical manufacturing system is simulated with a virtual factory model using discrete-event and/or agent-based simulation. The focus of this activity will be to introduce students to the current simulation software (Anylogic/SIMIO) pertaining to the manufacturing domain, in providing a rapid modeling capability for reporting, analyzing, and predicting smart factories.
Students will be able to design smart factories by using the four discrete-event paradigms evolved in simulation (i.e., events, processes, objects and agent-based modeling), and interpret simulation results to identify and solve problems in manufacturing. For an immersive learning experience, the simulation models developed will be developed to facilitate real-time analysis for remorse optimization and allocation.
Research Thrust 2: AR/VR Simulation for Intelligent Manufacturing Environment
AR/VR has evolved and improved over decades and include a paradigm of solutions in many fields and the manufacturing environment. The VR solutions encompass everything the workflow has to learn in order to deal with smart manufacturing environment. The purpose of this research project is to expose the student to exploring the AR/VR technologies to understand the complex activities and components in a manufacturing environment such as design, prototyping review, assembly and disassembly process. First, students will be able to interact with any designed product and prototype in a virtual environment to realize the potential of early design and digital prototyping to avoid risks. Later, they will focus on virtual plant walkthrough to leverage a clear understanding of the assembly and disassembly tasks and plant layout. Third, a quality control system with data analytics applications will be developed to support in training Quality Engineers (QEs) in being fully aware of inspection situation and being able to forecast defective rate at any future time. This will enable in utilizing as platform for educational training and applied research in the field of modeling and simulation.
Research Thrust 3: Machine Learning and Data Analytics for Manufacturing Systems
Machine learning and data analytics are now widely used to model manufacturing data for production planning and prediction. Such application is being very popular in additive manufacturing. Layer-wise monitoring in additive manufacturing (AM) is one of the key aspects for ensuring the quality of final products. Depending on the AM technology, the product quality in 3D printing could deteriorate due to changes of optional parameters such as temperature, print speed, positional distortion, photodiode signals, and others. continuous observation of these parameters provides an effective way to prevent any kind of product distortion. However, the task is not trivial due to the complex pattern and unforeseen changes in operational parameters. Moreover, the layers-wise printing setup generates a voluminous amount of data, which need to be analyzed properly to extract hidden pattern and findings. Nowadays, machine learning approaches have shown promising applications in big data analytics. In this project, we are exploring the machine learning approach, especially the deep learning algorithm, to develop in-situ process monitoring for Laser Power Bed Fusion (L-PBF) AM process.