Multiple Degrees-Of-Freedom Input Devices for Interactive Command and Control within Virtual Reality in Industrial Visualisations

Virtual Reality (VR) applications have often proven practical and effective in offering users an increased sense of presence and engagement when exploring inside a virtual environment (VE) creation. However, numerous user experience problems still occur at every stage within the application design, particularly for industrial visualisations (VR applications created exclusively for the industry). Students or employees need to be familiar with the essential features of these VR applications to complete their tasks, which can be hard when they come from different backgrounds. Industrial visualisations may contain multiple features that make them complex to map to a set of input mechanisms for user interaction, requiring training and learning skills. Training can take hours to months or years, depending on the users’ skills and interfacing with such environments. On the other hand, according to many research studies, having input devices that allow up to six degrees of freedom (6DOF) is a functional minimum to interface with these 3D environments. However, it is asserted that, depending on the task, there are cases in which users may need more DOF. Therefore, this thesis aims to design, build and implement a layered computing framework with a built-in input devices ontology and a strictly defined set of sub-APIs between the layers that intelligently connect multiple input devices to multiple application commands calls, enabling multiple DOF simultaneously. By leveraging a large number of DOF, users can interact with different input devices, allowing them to have more intuitive and natural control and manipulation of 3D objects in industrial visualisations and potentially master these VR applications in a short time. Empirical evaluations and case studies in industrial fields are presented that combine linear and non-linear function transformations with a comparison system. This study set out to demonstrate that by combining human spatial reasoning and computer graphics theory technologies, a framework like the one presented here can improve users' ability to understand, test and evaluate, reengineer, and then communicate better virtual behaviour. / A demo of this framework, as well as the datasets used for the experiment, can be found at: https://data.mendeley.com/datasets/k9wk9ybd64/1. - All the work was conducted on a PC within Windows 10 Pro, Dell Optiplex 7010, with an Intel Core i7-3770S processor, clocked at 3.10 GHz. - The padlock mechanism CT datasets were provided by the Manchester X-ray Imaging Facility (http://www.mxif.manchester.ac.uk/). - The Ketton carbonate core CT datasets were obtained from the British Geological Survey (BGS) database (https://metadata.bgs.ac.uk/geonetwork/srv/api/records/7315b790-333e-4e5b-e054-002128a47908/). - The Submillimetre mechanistic designs of termite-built structures CT datasets were obtained from the Zenodo database (https://zenodo.org/record/4792633). - Programming was done in Python (v.3.3), linking to the API of ANU Drishti version 2.6.4, compiled on Windows 10 using Qt 5.4.1 and libQGLViewer 2.6.1. - Controller setups were selected to be cross-evaluated; an Oculus Go Standalone Virtual Reality Headset - 32 GB, a SpeedLink SL-6638 Phantom Hawk Flightstick joystick, a Worthington Sharpe's Wing V.2, a Microsoft Xbox 360 controller, a Sony PS4 DualShock 4 V2 Wireless Controller, a custom setup consisting of a Keyboard + Mouse, and two 3DConnexion SpaceNavigators.