Research Proposal Augmented Reality Visualization of Outdoor Environmental Corrosion
Download 3.41 Mb.
|
- Navigate this page:
- Limitations
- Research Design
- Initial Analysis
- Research Methodology
- Expected Outcomes
- User Evaluations and Ethics
- Proposed User Evaluation Form
| Figure 5 Handheld humidity visualization
The application of visualization to large scale environmental monitoring has been examined (Fan and Biagioni, 2004, Yuxi et al., 2009). Despite the use of non AR visualization, Yuxi et al. (2009) demonstrated the use of sensors visualization in monitoring remote wetland environments. The wireless network topology is visualized as a layer on top of a geographic map, using colour codes to indicate the relationships between nodes. However, despite the application of mapping to visualization, the actual visualization of logged data is a traditional line graph. Similarly, Fan and Biagioni (2004) propose the GRASS GIS system, also for large scale environmental monitoring. Using a wireless sensor network, environmental data is logged, with rainfall visualized as colour-coded Voronoi diagrams, with temperature also visualized as line graphs. Whilst not visualizing sensor data, Malkawi and Choudhary (1999) demonstrate the visualization of simulated heat transfer within a structure over time. By adjusting the structure’s physical parameters, the visualization will adjust the visualization to reflect the transfer of heat within the structure. Rad and Khosrowshahi (1997) extend this concept again, by visualizing the impact of parameters on a structure’s maintenance over time. Following a set of mathematical models, the structure visibly decays, representing the required maintenance for up-keep. LimitationsThis section will provide an analysis of the research presented in the literature review. By analysing the limitations of the current systems and methodologies, we can identify areas of opportunity for innovation, development and further research. The section will serve as justification for this research’s contributions. Current systems for AR visualization of sensor data have had the capability to only visualize a single parameter. Despite certain conditions, such as environmental monitoring, containing a number of logged parameters, the visualization’s have only enabled the inclusion of one parameter, thus obscuring any relationships that may be present between the parameters. Despite the use of a single parameter, only one AR system has included the ability to interpolate between values at fixed points. Given the possible application of in-situ visualization across large areas, it would be impossible to log data from every point, inferring that interpolation would be a valuable asset in representing the collected data. Despite work relating to the visualization of building conditions, currently there exists a gap between the use of environmental sensor networks and the application of that data in building visualization. Current systems do not currently inform the user of any abnormalities or rapid changes within a data set (no analysis is performed). For the use of large data sets, one could assume that the automation of detecting and identifying rapid changes and outliers in the data to the user would be a desirable function. Similarly, the ability to retrieve the exact data values in conjunction with its visualization does not seem to be readily supported by current tools. Research DesignThis section outlines the proposed design of the research and is split into three areas; Initial Analysis, Research Methodology and Expected Outcomes. The Initial Analysis provides the foundation for establishing a plan and research method which constitutes the Research Methodology. The section concludes with the expected outcomes of the research. According to Shaw (2002) the classification of this research is a ‘method of development’ with the expected results being a ‘procedure or technique’, which given the research, will be a technique. Initial AnalysisAn initial review of the planned research has been conducted to determine the best approach and methodology for the desired outcome. The platform of development will be the Tinmith (Piekarski and Thomas, 2003) mobile AR system. Despite the integration of my research as part of an existing system, the development of a new ‘plug-in’ model for Tinmith will enable the construction of the system with minimal interaction with Tinmith. A shared header interface (Figure 6) will enable the construction of a separate system without relying on detailed knowledge of the Tinmith system. This header will also enable the use of the visualization subsystem by any OpenGL application that implements the shared header’s interface. The visualization system will contain no Tinmith dependant code. 
Figure 6 Proposed Tinmith shared library architecture Given the nature of the research in creating effective representations for outdoor environmental corrosion, preliminary research will be conducted to provide background knowledge to assist in the effective design and selection of the visualizations. This will serve as part of the first step in the analyse, design, implement and evaluate process. Following a review of the proposed system design, implementation will occur on the Tinmith plug-in, after which the process will be evaluated to determine the continued feasibility of the methodology. Research MethodologyFollowing an initial review of the nature of the research, an iterative analyse, develop, implement and evaluate development cycle is proposed (Figure 7). Given the timeline for the research, two iterations are proposed, with user evaluations performed at the conclusion of each iteration. These evaluations will provide an indication of progress and provide feedback as grounds for future improvement of the system. Both iterations will focus on the development of effective and intuitive visualizations for outdoor environmental corrosion, with each iteration resulting in a running system to ensure the system’s availability for user evaluations. 
Figure 7 Proposed iterative development cycle Expected OutcomesThe outcome of this research is intended to be a set of visualizations that effectively represent outdoor environmental corrosion in mobile AR. For the visualization techniques, full descriptions of the development process, along with a functional visualization system prototype (as a plug-in to the Tinmith mobile AR system) to demonstrate the visualizations will be provided. The system will support the viewing of detailed numerical data in conjunction with its visual representation, along with automatic identifying and directing the user to ‘areas of risk’ on the structure. The visualization system will be able to be implemented by any system that implements the system’s ‘plug-in’ interface. Full user evaluations of the visualizations will be available as justification for their development history, outlining why specific facets of them were designed as they were, along with additional comments relating to effective implementation of them. The requirements for implementing the visualization system in other systems will be given, along with a list of essential and non-essential hardware required to interact with the system. Specifications and working examples for the reading, transmission and storage of data from the wireless sensors will also be provided. Full source code will be provided for the visualizations, the sensor visualization system along with all associated utilities that have been developed. Despite not being a contribution, a side outcome of the research will be a generic plug-in interface and template for the Tinmith system. The interface specification, along with a working example will be available. User Evaluations and EthicsThis section outlines the involvement of user evaluations as part of the research, along with the associated ethics considerations. The requirement for user involvement in the evaluation of visualization techniques leads to the requirement for ethics approval by the university. Given the nature of the data required, it is expected that the ethics proposal will fall under a ‘low risk’ category. The research’s ethics application is already in progress, and is expected to be completed mid-June. Given the reasoning behind using a mix of qualitative and quantitative methodologies for evaluating visualizations (North, 2006), the user survey will involve a range of questions serving both methods of measurement. The remainder of section provides a sample of the expected user evaluation form that will be used to gauge the effectiveness of the proposed visualizations.
Name:
Age: Gender: M F (Please circle) Have you used an augmented reality system before? No Yes Extensively
Have you used a wearable computer system before? No Yes Extensively (Please circle) Please answer the following questions based on your interpretations of the system: Which sensor had the most corrosion? (Please circle) Left-Near Left-Far Right-Near Right-Far Which sensor was the first to draw your attention? (Please circle) Left-Near Left-Far Right-Near Right-Far What was the inside temperature at the left-near sensor? ________oC What was the outside temperature at the left-far sensor? ________oC Which sensor requires the most urgent attention? (Please circle) Left-Near Left-Far Right-Near Right-Far Did you prefer the ‘plasma’ or ‘crystal’ effect for showing corrosion? (Please circle) Neither Plasma Crystal Both Why? ___________________________________________________________________________ Did you prefer the ‘box’ or ‘gauge’ sensor visualization for representing the sensor? (Please circle) Neither Plasma Crystal Both Why? ___________________________________________________________________________ How intuitive was the ‘box’ sensor to understand? (Please draw a line to indicate your preference) Not Intuitive _____________________ Very Intuitive How intuitive was the ‘gauge’ sensor to understand? (Please draw a line to indicate your preference) Not Intuitive _____________________ Very Intuitive End of Questionnaire Download 3.41 Mb. Share with your friends:
|