Mining Asset Development for Virtual Reality

DISCLAIMER The findings and conclusions in this paper are those of the authors and do not necessarily represent the official position of the National Institute for Occupational Safety and Health, Centers for Disease Control and Prevention. Mention of company names or products does not constitute endorsement by NIOSH. ABSTRACT The pervasiveness of high-fidelity video games raises expectations of quality, detail, and lighting in virtual reality (VR) applications. However, high-quality models of mining equipment and environments are often not readily available, and developers may not be familiar with the latest techniques, such as physically based rendering (PBR). Because a consistent aesthetic is critical for maintaining immersion and controlling visual stimuli in research as well as in training, researchers at the National Institute for Occupational Safety and Health (NIOSH) created an asset development workflow. This paper details this workflow, including precision modeling and PBR techniques as well as a rigorous review process. It also describes lessons learned in the creation of photorealistic VR assets. BACKGROUND Virtual reality (VR) describes the simulated experience generated by an assemblage of hardware and software that are used to immerse the user. Over many years VR developers targeted the development of virtual environments (VE) toward various purposes: entertainment, the treatment of phobias [1], rehabilitation [2], and training [3]. A large part of the VE development includes digital assets, where digital assets refer to the non-programmed elements including 3D models, audio, and user interface. Because vision provides nearly 10Mbps of data [4] and accounts for the majority of all sensory input combined [5], a focused effort on asset appearance in VEs is warranted. Furthermore, research has shown that higher-resolution imagery increases search performance and training transfer, necessitating high-fidelity assets [6, 7]. In addition to being visually appealing, it is important that the assets and environment are also dimensionally accurate and consistent in order to elicit appropriate responses from the end users as they interact with and among these items [8, 9]. Researchers at the National Institute for Occupational Safety and Health (NIOSH) have developed VEs and digital assets under several research projects over the course of the last two decades [10, 11, 12, 13]. This work has provided researchers with a strong knowledge base for digital asset development. However, the assets themselves have had limited reuse value. Due to the extended timeframes and limited scope of these projects, the previously developed assets are not consistent in visual quality and dimensional accuracy. Moving forward, researchers wanted to minimize future development time and use of resources as well as adopt new asset development tools and game engine technology. Therefore, the aim of this paper is to document a new asset development workflow including geometry, materials, UV layout, and shader implementation combined with review steps in order to produce high-fidelity simulation assets that are accurate with consistent characteristics, using repeatable methodologies that are well documented.