Emergencies that require immediate evacuation should be encountered with effective preparedness. With over 14 billion dollars in damages and 3,000 people killed each year, fire emergency preparedness is of critical importance. Fire drills aim to prepare and educate people on how to react properly, in order to avoid as many casualties possible. Fire drills can be expensive and time consuming to conduct, and in most cases lack the level of realism to properly educate the trainees. In this thesis, a virtual reality (VR) emergency evacuation training platform is presented. With VR, we aim to eliminate the real life constraints that exist, while succesfully training individuals. The application operator can spawn fires in any desired location, and at the same time the user being trained is getting informed about the safest and fastest path available, while being provided with constant feedback. We generate a grid graph on a given floor plan to run a pathfinding algorithm. We use Linear Temporal Logic (LTL) to formulate the existing constraints in our approach.

Currently, games with rich environments allowing a wide range of possible interactions and supporting a large number of physical simulations make use of a large number of scripts and bespoke physical simulations, adapted to fit the needs of the game. This thesis proposes a methodology that can be used to tie together various different physical simulations, both off-the-shelf and bespoke, such as rigid body physics, electrical and magnetic simulations to give something greater than the sum of the individual parts. We present a notation for designing the overall physical simulation and a means for the different parts to interact. Experiments using an implementation of the methodology containing electricity, rigid body simulation, magnetics (including electro-magnetics), buoyancy and sound show that it is possible to model everyday objects such an electric motor or a doorbell. These object work ‘as expected’, without the need for special scripts and new, originally unexpected, interactions are possible without further modification of the experiment setup.

Virtual reality (VR) headsets are becoming more common and will require evolving input mechanisms to support a growing range of applications. Because VR devices require users to wear head-mounted displays, there are accomodations that must be made in order to support specific input devices. One such device, a keyboard, serves as a useful tool for text entry. Many users will require assistance towards using a keyboard when wearing a head-mounted display. Developers have explored new mechanisms to overcome the challenges of text-entry for virtual reality. Several games have toyed with the idea of using motion controllers to provide a text entry mechanism, however few investigations have made on how to assist users in using a physical keyboard while wearing a head-mounted display. As an alternative to controller based text input, I propose that a software tool could facilitate the use of a physical keyboard in virtual reality. Using computer vision, a user€™s hands could be projected into the virtual world. With the ability to see the location of their hands relative to the keyboard, users will be able to type despite the obstruction caused by the head-mounted display (HMD). The viability of this approach was tested and the tool released as a plugin for the Unity development platform. The potential uses for the plugin go beyond text entry, and the project can be expanded to include many physical input devices.

Sound is a crucial part of our life. Well-designed acoustic behaviors can lead to significant improvement in both physical and virtual interactions. In computer graphics, most existing methods focused primarily on improving the accuracy. It remained underexplored on how to develop efficient acoustic simulation algorithms for interactive practical applications. The challenges arise from the dilemma between expensive accurate simulations and fast feedback demanded by intuitive user interaction: traditional physics-based acoustic simulations are computationally expensive; yet, for end users to benefit from the simulations, it is crucial to give prompt feedback during interactions. In this thesis, I investigate how to develop efficient acoustic simulations for real-world applications such as immersive media and digital fabrication. To address the above-mentioned challenges, I leverage precomputation and optimization to significantly improve the speed while preserving the accuracy of complex acoustic phenomena. This work discusses three efforts along this research direction: First, to ease sound designer's workflow, we developed a fast keypoint-based precomputation algorithm to enable interactive acoustic transfer values in virtual sound simulations. Second, for realistic audio editing in 360° videos, we proposed an inverse material optimization based on fast sound simulation and a hybrid ambisonic audio synthesis that exploits the directional isotropy in spatial audios. Third, we devised a modular approach to efficiently simulate and optimize fabrication-ready acoustic filters, achieving orders of magnitudes speedup while maintaining the simulation accuracy. Through this series of projects, I demonstrate a wide range of applications made possible by efficient acoustic simulations.

With the increasingly rapid adoption of vehicles with autonomous features, concerns over human driver and passenger safety in such vehicles have greatly increased, especially in regards to autonomous driving features such as Tesla’s Autopilot. In order to improve current

In the so-called postmodern era, when networked media are increasingly ubiquitous in everyday life, where the ‘real’ and the ‘simulation’ become ever more indistinguishable; the physical and virtual intertwine; machines and man merge, and audience and stars transpose. To understand consumption in a time when realness and authenticity are no longer relevant, this thesis draws attention to the consumption and production of media content through case studies of consumer participation and social trends in Japan. The work begins in a themed shopping mall, Venus Fort in Tokyo Bay; continues with the reproduction of Audrey Hepburn‘s image; expands to the dramatised ‘realness’ of television; and finally moves to the omnipresent mobile phone and the impact of networked personal media on our idea of the ‘real’. First, through an analysis of a themed consumption environment, it is suggested that a transition is taking place in consumption from objects to experiences, services and spectacle. Secondly, by showing Audrey Hepburn‘s transition from a Hollywood star to a virtualised idol, technologically-aided illusions are shown to make hierarchical realness irrelevant. Thirdly, via Reality TV dating programs, the focus shifts to the role of audience participation in the consumption of media content. These themes are demonstrated individually, then merged into the last example – the social and cultural evolution induced by the mass consumption of networked media, that promise to revolutionise the way we consume, communicate and connect between people, machines and consumer goods.The thesis grounds its analysis of contemporary trends in the culture of consumption in Japan in theories of commodity and culture, the real and the simulation, speed and reality, the spectacle and the self in mediated spaces, and probes further into the collapse of demarcations between the virtual and the real, the event and the everyday and media and the self in the network society / Doctor of Philosophy (PhD)

University of Technology, Sydney. Faculty of Information Technology. / This thesis establishes Virtual Institutions as a comprehensive software engineering technology for the development of 3D Virtual Worlds that require normative regulation of participants’ interactions (such as the commercially-oriented Virtual Worlds). 3D Virtual Worlds technology currently offers somewhat unregulated environments without means to enforce norms of behavior and interaction rules on their inhabitants. Furthermore, existing methodologies for Virtual Worlds development focus primarily on the design side of the “look-and-feel” of the inhabited space. Consequently, in current 3D Virtual Worlds it is difficult to keep track of the deviant behavior of participants and to guarantee a high level of security and predictable overall behavior of the system. The Virtual Institutions Methodology proposed by this dissertation is focused on designing highly secure heterogeneous Virtual Worlds (with humans and autonomous agents participating in them), where the participants behave autonomously and make their decisions freely within the limits imposed by the set of norms of the institution. It is supported by a multilayer model and representational formalisms, and the corresponding tools that facilitate rapid development of norm-governed Virtual Worlds and offer full control over stability and security issues. An important part of the Virtual Institutions Methodology is concerned with the relationship between humans and autonomous agents. In particular, the ways to achieve human-like behavior by learning such behavior from the humans themselves are investigated. It is explained how formal description of the interaction rules together with full observation of the users’ actions help to improve the human-like believability of autonomous agents in Virtual Institutions. The thesis proposes the concept of implicit training, which enables the process of teaching autonomous agents human characteristics without any explicit training efforts required from the humans, and develops the computational support for this new learning method. The benefits of using Virtual Institutions are illustrated through applying this technology to the domain of E-Commerce. It is demonstrated that providing shoppers with a normative environment that offers immersive experience and supports important real world attributes like social interaction, location awareness, advanced visualization, collaborative shopping and impulsive purchases can improve existing practices in E-Commerce portals.

Thesis (Ph. D.)--University of Wyoming, 2006. / Title from PDF title page (viewed on Dec. 20, 2007). Includes bibliographical references (p. 78-83).

Thesis (M.S.)--West Virginia University, 2001. / Title from document title page. Document formatted into pages; contains ix, 87 p. : ill. (some col.). Includes abstract. Includes bibliographical references (p. 35-36).

Thesis (Ph. D.)--University of Washington, 2001. / Vita. Includes bibliographical references (leaves 95-104).