Virtual Solar Energy Center: A Case Study of the Use of Advanced Visualization Techniques for the Comprehension of Complex Engineering Products and Processes

Industry has a continuing need to train its workforce on recent engineering developments, but many engineering products and processes are hard to explain because of limitations of size, visibility, time scale, cost, and safety. The product or process might be difficult to see because it is either very large or very small, because it is enclosed within an opaque container, or because it happens very fast or very slowly. Some engineering products and processes are also costly or unsafe to use for training purposes, and sometimes the domain expert is not physically available at the training location. All these limitations can potentially be addressed using advanced visualization techniques such as virtual reality. This dissertation describes the development of an immersive virtual reality application using the Six Sigma DMADV process to explain the main equipment and processes used in a concentrating solar power plant. The virtual solar energy center (VEC) application was initially developed and tested in a Cave Automatic Virtual Environment (CAVE) during 2013 and 2014. The software programs used for development were SolidWorks, 3ds Max Design, and Unity 3D. Current hardware and software technologies that could complement this research were analyzed. The NVIDA GRID Visual Computing Appliance (VCA) was chosen as the rendering solution for animating complex CAD models in this application. The MiddleVR software toolkit was selected as the toolkit for VR interactions and CAVE display. A non-immersive 3D version of the VEC application was tested and shown to be an effective training tool in late 2015. An immersive networked version of the VEC allows the user to receive live instruction from a trainer being projected via depth camera imagery from a remote location. Four comparative analysis studies were performed. These studies used the average normalized gain from pre-test scores to determine the effectiveness of the various training methods. With the DMADV approach, solutions were identified and verified during each iteration of the development, which saved valuable time and resulted in better results being achieved in each revision of the application, with the final version having 88% positive responses and same effectiveness as other methods assessed