REAL-TIME DATA VISUALIZATION IN ... - Semantic Scholar

This article was peer reviewed by two independent and anonymous reviewers

REAL-TIME DATA VISUALIZATION IN COLLABORATIVE VIRTUAL ENVIRONMENTS FOR EMERGENCY RESPONSE 1

1

1

Pan Wang , I.D. Bishop , C. Stock 1

CRC for Spatial Information and University of Melbourne, Parkville, Australia

[email protected]

ABSTRACT Collaborative Virtual Environments or CVEs are shared virtual environments used for collaboration and interaction of many participants that may be spread over large distances. CVEs are used in diverse areas, ranging from e-tourism and VR-shops to coal mining operations and interactive house design. Due to the benefits of CVEs, they have been introduced to the field of emergency management for education, training and assessment. However, there are some drawbacks in existing CVEs for emergency management: 1.Due to lack of real-time data, reality in most training scenarios is far from the real world, which cannot provide the trainee with vivid experience; 2. The connection between spatial information and Collaborative Virtual Environments still needs improving. Thus, a new method that can integrate real-time data into CVEs is in need. Our current research is based on the Spatial Information Exploration and Visualization Environment (SIEVE), which allows users to explore existing spatial data and hypothetical future scenarios in a real-time 3D environment. We are extending the function of SIEVE to be able to collect attribute data through ESRI ArcGIS Server using data mashup technology, which is a web-based technology that combines data or functionality from two or more external sources to create a new service. These data include weather conditions or real-time traffic, which enhance the reality of emergency training. Then data are transferred into SIEVE and visualized in real-time. Such data are very important for decision makers and data presented in a real-time 3D environment may provide a more realistic scenario and improve the effect of training.

Wang, P., Bishop, I.D. and C. Stock (2009). Real-time data visualization in Collaborative Virtual Environments for emergency response. In: Ostendorf B., Baldock, P., Bruce, D., Burdett, M. and P. Corcoran (eds.), Proceedings of the Surveying & Spatial Sciences Institute Biennial International Conference, Adelaide 2009, Surveying & Spatial Sciences Institute, pp. 435-441. ISBN: 978-0-9581366-8-6.

Pan Wang, I.D. Bishop, C. Stock

INTRODUCTION Disasters, be they natural or human-induced, can greatly endanger the security of human life and property. It is manifest that such disasters cannot be avoided by individuals. Nonetheless, we can apply new technologies in the field of emergency management, thereby diminishing the loss to its minimum level. One of them is Geo-information technology, which mainly deals with spatial data collection, storage, analysis, and display to provide useful information to decision makers such as local government, emergency responders, and medical rescuers (Radke et al, 2000). Another technology involves the development of Virtual Reality and visual simulation. Due to their particular advantages in practice, they have been widely used in emergency training like urgent health care and fire fighting, which offer the trainees immersive and verisimilar training environments. Without having experienced the hazards in a real situation, trainees, however, are able to improve their ability of making decisions and response to the emergency hazards simply by controlling the avatars and reacting to the various circumstances in the virtual environment. (Burdea & Coiffet, 2003) Admittedly, the presentation of real-time data plays a pivotal role in emergency training. Such data changes all time and the variation of it will influence the decision making of the users. Hence it would be desirable if a system can combine the advantages of Geo-information technology, virtual training as well as real-time data in one package. A collaborative virtual environment can be used for training people to deal with emergency situations. Clearly, the response of people in an emergency depends on the specific circumstances at the time. The change of environmental factors such as weather, visibility, traffic levels and also individual incidents, can affect the response behaviour. In a training environment it is ideal to provide a set of circumstances which is known to be possible. This can be done by recording variations in key parameters in real-time and playing them back into the simulated training exercise. Even more significant may be situations in which in-field operators are communicating with head-quarters staff who are directing the emergency response. A CVE can then be used to give those at HQ a close approximation of on-ground conditions. However, this will be of very limited use unless the virtual environment is reflecting the true conditions. That can only be achieved through streaming of real-time data into the CVE.

REAL-TIME DATA INTEGRATION IN VIRTUAL ENVIRONMENT Generally speaking, virtual environments are composed of static components such as terrains, buildings, roads, etc; and dynamic components such as weather condition, sea tide, and vehicle flow. Real-time data has the advantage in representing variational situation in a dynamic scene, and it is especially valuable and essential when it comes to emergency training and response in areas such as fire fighting response, anti-terrorism, or first-aid service. However, unlike creating static objects in CVE, visualizing dynamic real-time data is more complex because these data are constantly changing and it is necessary to update every short period, thus a new method is under research for visualizing real-time data. Inspired by mashup technology, which is a newly emerging web application that integrates data from different sources into one single tool, we design a new framework for data mashup in a 3D environment. Some examples of mashup are Travature, a travel portal that has integrated airfare meta search engines, wiki travel guides, hotel reviews which benefit the travellers; Weatherbonk , an web 2.0 application that provides highly contextualized relevant weather information for people from around the world; WikiCrimes,a wiki-style Web site where users can report crimes by placing pins on a GoogleMaps based map, etc (Mashup (web application hybrid), 2009) . The architecture of a mashup application (Figure 1) is comprised of three participants: API/content providers, the mashup site, and the client's Web browser (Merrill, 2006). •

The API/content providers. They provide the content which can be mashed up by the users. In the Weatherbonk mashup example, the providers are Google, Weather Underground, Weather Bug, etc. Providers often publish their content through Web-protocols to facilitate data retrieval.

•

The mashup site. This is where the mashup is hosted and where the mashup logic resides, not necessarily where it is executed. Mashups can be implemented similarly to traditional Web

436


applications using server-side dynamic content generation technologies like Java servlets, CGI, PHP or ASP. •

The client's Web browser. This is where the result is displayed and where user interaction takes place.

Figure 1 Mashup Application

Currently mashup applications are invariably embedded in web-based applications, and usually based on text content or in 2D environments, such as maps or graphics, which are inadequate for the virtual training purpose. In this research we intend to extend this technology into the 3D environment (Figure 2). In this research, we selected SIEVE (Spatial Information Exploration and Visualization Environment) as the 3D visualization platform. It is developed by CRCSI in 2005, enabling users to explore existing spatial data and hypothetical future scenarios in a real-time 3D environment. The framework and working procedures are as follows: Firstly data are transferred from different data sources, such as weather information websites to a realtime database engine. These data have dissimilar types but are all in standard data format such as ATOM or RSS feed based on XML, in order to facilitate data processing. Secondly, real-time data are stored in geo-database for reuse or later analysis. The stored data are necessary if we want to use historical data in certain locations. Next, real-time integration engine initiates and collects real-time data from different data sources and transfer them into a visualization modelling engine. Finally, the visualization modelling engine plays a key role in transferring abstract data to visual representation. As a result, the virtual environment is supplemented with real-time data on a CVE server. And users are enabled to connect to the server from different clients to share information with each other.

437


Figure 2 Framework of Real-Time Data Integration

MODELING OF REAL-TIME DATA After collecting data from various sources, we then transfer them into the virtual environment for visualization. There is a projection between the original data and visualized data. Take weather data as an example, information in the weather report are usually provided in the form of an XML file, including temperature, moisture, and precipitation. While the weather simulation in SIEVE is scripted, key parameters include cloudSpeed, fogVolume, windVelocity, etc. Thus a converter from the source data format to SIEVE supported data format is necessary. Figure 3 shows some of the conversion from source data to SIEVE supported data. Weather condition: information in the weather report are usually provided in the form of an XML file, including locations (longitude and latitude), time (date/hour/minute), wind (wind direction and speed), astronomy (sunrise, sunset), precipitation (type and quantity), Condition code (integers indicate weather conditions such as sunny, cloudy, rainstorm, blizzard, etc). Then these data are converted to visualization data key parameters and displayed in collaborative virtual environment. Real-time traffic flow: Real-time traffic data are often represented by three parameters: road, start and end point, traffic jam level. And the visualization result will demonstrate 3D traffic data flow in real-time, which enable emergency managers or trainees to take different measures in a complex environment. Incident Status: parameters in status of incident are similar to real-time traffic condition, including Incident ID, location, time, type and status. In the virtual environment, these real-time data are represented as incident scenarios that change with time.

438


Category Real-time data Key Parameter

Visualization data Key Parameter

Weather Location Condition Time

Coordinate in CVEs Synchronized Time

Wind

,

Astronomy

Precipitation

Condition Code (0,1,…)

Datablock based Weather Condition

Traffic Location Condition Start-End Point

,

Level , Incident ID Status Location

ID Coordinate in CVEs

Time

Synchronized

Type

Type Model

Status

Figure 3 Visualization Data Conversion This converter is initially designed for converting weather conditions; later it will also support multi-type data. Those data should be in accordance with the following rules: 1. such data can potentially be visualized in SIEVE Viewer; 2. data is published in some way by content providers; 3. data is useful for emergency response or training. At present, millions of online content providers publish their data through web feed. Everyday there are new content and providers. In this research we aim to build a common tool for real-data collection and integration for SIEVE. Currently there are limited types of data complying with the rules mentioned above, but with the increasing amount of data content online, more available data will be integrated into SIEVE, or similar tools, to make the virtual environment get closer to the vision of real world.

DEVELOPMENT ON MULTI-USER COLLABORATIVE VIRTUAL TRAINING To achieve a better understanding of real-time data integration and visualization technology, an incident scenario at Sydney harbour has been selected as a case study. This project is based on a hypothetical terrorism incident scenario. More precisely, a terrorism action attacking Captain Cook dock in Sydney Harbour is assumed. In this study, we will build a collaborative virtual environment for military training in maritime security applications. Following, we describe the main approach of this study currently in progress to verify the real-time data integration and visualization technology. (Figure 4) 1. Building the scene of area of interest. A mirror world of Sydney Harbour with full featured urban environment will be created, including terrain, water, vegetation, buildings, roads, vehicles, etc. 2. Simulation of dynamic situation. The SIEVE-VTS real-time data processor will connect to online data sources from different content providers as selected by users. These data will largely vary with time and will be updated frequently. SIEVE Viewer will display the corresponding result. 3. Collaborative training in SIEVE-VTS based on hypothetical Sydney harbour incident scenario. The variation of incidents will be updated by sending live data stream to the training system.

439


Sydney Harbor Scenario

Rainy condition

Real-time Traffic flow

Incident location and status

Figure 4 Sydney Harbour Incident Scenarios Real-time data are divided into two parts: one is public data including weather information, real-time traffic and incident locations and status; another is secure data from our project collaborators including some warning information and real vehicle movement information from GPS. Real-time weather in CVEs may provide users with real experience and lead them to respond to different weather condition; traffic information may remind users to select alternative routes when facing heavy traffic jam in an emergency; Change of incident locations and status gives users additional information to take measures to deal with the emergency. While systems development and proof of concept will focus on the training example, the processed learned will later have potential application in real emergency response situations.

CONCLUSION AND FURURE WORK Real-time data plays a significant part in emergency training, management and response. The development of a component for integrating real-time data and CVEs into an overall framework for a virtual training system will improve the immersion experience and reality of the virtual training environment, as well as strengthen the ability of decision making and collaboration for the trainees, thereby facilitating better decision making processes and offering additional insights to users in a complex environment. The application can extend to any situation in which monitoring data can be quickly collected and transferred into the virtual environment. In addition to the terrorism example, this might include fires, floods, shipping emergencies, gas leaks and a range of other hazards and threats. In the longer term, such a system may save lives and property by improving emergency response decision making. Future research includes specific real-time data modelling in certain areas, advanced real-time data integration technology as well as data management such as data storage and query. Meanwhile there are also some issues of implementation to be solved. One is data accuracy, e.g. when real-time weather data

440


or traffic information is not accurate enough for modelling. To solve this problem, we need to design new methods to collect high-accuracy real-time data meeting the user’s requirement; another issue is data availability, i.e., if real-time data are not available for a place of interest. Collecting historical data in a different format or interpolation from existing data in a nearby district could be a potential solution.

REFERENCES Boulos, M. N. K. & D. Burden (2007). "Web GIS in practice V: 3-D interactive and real-time mapping in Second Life." International Journal of Health Geographics 6(1): 51. Burdea, G. & P. Coiffet (2003). "Virtual reality technology" Presence: Teleoperators & Virtual Environments 12(6): 663-664. Chen, T., Stock, C., Bishop, I.D. and O’Connor, A. (2006) Prototyping an in-field collaborative environment for landscape decision support by linking GIS with a game engine. Proceedings of SPIE Volume: 6418 Geoinformatics 2006: GNSS and Integrated Geospatial Applications, Editor(s): Deren Li, Linyuan Xia, Wuhan October 2006. Churchill, E. F., D. N. Snowdon, et al. (2002). "Collaborative virtual environments: digital places and spaces for interaction." Educational Technology & Society 5(4). Merrill, D. (2006). Mashups: The new breed of Web app. Retrieved March 15, 2009 form http://www.ibm.com/developerworks/xml/library/xmashups.html?S_TACT=105AGY82&S_CMP=GENSITE J Radke, T Cova, MF Sheridan, A Troy, L Mu, Russ Johnson(2000). Challenges for GIS in Emergency Preparedness and Response, ESRI Library, Redlands, California, USA Mashup (web application hybrid). (2009, March 26). In Wikipedia, The Free Encyclopedia. Retrieved 03:55, March 27, 2009, from http://en.wikipedia.org/w/index.php?title=Mashup_(web_application_hybrid)&oldid=279825274 O'Connor, A. N. (2007). Automatic virtual environments from spatial information and models, Thesis (Ph.D.), University of Melbourne, Dept. of Geomatics

BRIEF BIOGRAPHY OF PRESENTER Pan Wang, PhD student in Department of Geomatics and CRC for Spatial Information, University of Melbourne, interested in Geo-Visualization.

441