A Grammar for 3D Visualization and a Cloud-Distributed Implementation

Abstract

In many real-time control situations, including traffic management centres, the quantity and surge of events can overwhelm the operator’s cognitive ability to understand and react to event notifications. Research has shown that effective visualizations that provide context and personalization to the operator can improve their ability to interpret visualizations. This thesis investigates the efficacy of employing a 3D visualization grammar for prototyping 3D visualizations. It also tries to compare how geo-representation im- proves context for operators in understanding the underlying information in contrast to the traditional text based systems. The literature proposes visualization design techniques for effective use in the traffic sat- uration visualization. Different declarative systems and their system architectures were also investigated for the design of the 3D visualization grammar. In this context, Aloka - a 3D visualization grammar, was designed and developed. It allows users to define 3D visualizations in a declarative specification in a JavaScript Object Notation (JSON) file. It also provides a novel way to customize 3D visualization types and offers a flexible approach towards different data types, display types(VR, HDTV etc) and data mapping to visual representations. The design and development of Aloka is explored, its system architecture dissected and it’s philosophy explained. This is a novel approach to creating complex 3D visualizations for a wide range of systems by defining the specification of a 3D visualization by data-binding it to visual properties for high quality ray traced rendering. Experiments were carried out to find that operators using a custom designed 3D geo- visualization over the existing text based systems UTC (Urban Traffic Control) improved context and significantly speeds up response time in detecting traffic saturation. These 3D visualizations were prototyped using Aloka. This project is sponsored by an Engineering and Physical Sciences Research Council (EPSRC) iCASE award from Siemens PLC, Poole.