Visual Analytics on Mobile Devices for Emergency Response

Transcription

1 Visual Analytics on Mobile Devices for Emergency Response SungYe Kim, Yun Jang, Angela Mellema, David S. Ebert Timothy Collins Purdue University Regional Visualization and Analytics Center (PURVAC) Purdue Homeland Security Institute (PHSI)

2 Demo

3 Introduction (1/2) Mobile Visual Analytics Visual Analytics using mobile devices (PDA, Cell-phone) Effectiveness and interactivity of analysis on-site/in-field Mobility Rapidly growing capabilities Wireless network infrastructure

4 Introduction (2/2) Goal Mobile Visual Analytic System for emergency response Prototype for general purpose tool of mobile visual analytics Interactive visualizations that help first responders (police, fire fighter, etc.) in answering general emergency response (or training) related questions Challenges H/W limitations Small screen (various resolutions) Limited processing power and memory Large data generated for desktop system Visual analytics for emergency response and training

5 Related work (1/2) Visualization of sensor data <For Mobile Devices> Pattath A., Bue B., Jang Y., Ebert D., Zhong X., Ault A., Coyle E., Interactive visualization and analysis of network and sensor data on mobile devices, VAST 06. Gunnarsson A.-S., Rauhala M., Henrysson A., Ynnerman A., Visualization of Sensor Data Using Mobile Phone Augmented Reality, ISMAR 06. <For Desktop systems> Fan F., Biagioni E. S., An approach to data visualization and interpretation for sensor networks, HICSS 04.

6 Related work (2/2) Visualization on Mobile devices Zhou H., Qu H., Wu Y., Chan M., Volume Visualization on Mobile Devices, PG 06. Chittaro L., Visualizing information on mobile devices, IEEE Computer, Duguet F., Drettakis G., Flexible Point-Based Rendering on Mobile Devices, IEEE CG&A, Masoodian M. and Budd D., Visualization of travel itinerary information on PDAs, AUIC 04. Hekmatzada D., Meseth J., Klein R., Non-photorealistic rendering of complex 3d models on mobile devices, IAMG 02. Visual analytics on Mobile devices Herrero Á., Corchado E., Sáiz J. M., MOVICAB-IDS: Visual Analysis of Network Traffic Data Streams for Intrusion Detection, IDEAL 06. Sanfilippo A., May R., Danielson G., Baddeley R., et. al., InfoStar: An adaptive visual analytics platform for mobile devices, SC 05.

7 System overview Local Visualization

8 Input Data Moving entities (Personnel) (simulated) ID, position, time stamp, [value, etc.] Emergency situation (simulated) Type, position, time stamp, [value, etc.] e.g., fire, dispersion of toxic gas, water contamination 2D Map / 3D model of an environment Image of an environment by orthographic view Definition of specific spots, e.g., emergency exits

9 Visualization Issues Dynamic (D) vs. Static (St) Moving entities Static scene Temporal (T) vs. Spatial (Sp) Information changed at certain time Information at different places at same time Aggregation (A) vs. Non-Aggregation (na) Population Specific individual Vector Graphics for flexible screen resolutions 3D Viewing

10 Visual Analytics (1/2) Object Personnel Scene 1. Who is he/she? 2. Where is he/she? (Where are they?) 3. What about his/her/their movement pattern for evacuation? 4. How is his/her/their health condition? 5. How does their health condition change? 6. Did he/she success in evacuating? 7. How many entities succeed in evacuating? 1. What is the environmental condition? (distribution of temperature, toxic gas, etc.) 2. How does the environmental condition change? 3. What is the environmental structure? (map, emergency exits) Questions for Analysis Type of Vis. (na) (A+nA) (A+nA)(D)(T) (A+nA) (T) (na) (A)(T) (Sp) (T) (St) cf. (D):Dynamic, (St):Static, (T):Temporal, (Sp):Spatial, (A):Aggregate, (na):non-aggregate

11 Visual Analytics (2/2) Object Questions for Analysis Functions Personnel 1. Who is he/she? 2. Where is he/she? (Where are they?) 3. What about his/her/their movement pattern for evacuation? 4. How is his/her/their health condition? 5. How does their health condition change? Visualize personnel info. Visualize movement path Visualize health change (Result (3/4)) (Result (1/4)) (Result (1/4)) (Result (3/4)) Scene 6. Does he/she success to evacuate? 7. How many entities success to evacuate? 1. How is environmental condition? (distribution of temperature, toxic gas, etc.) 2. How does environmental condition change? 3. How is environmental structure? (map, emergency exits) Visualize personnel s status Visualize emergency status Visualize environmental condition information Visualize change of environmental condition Visualize 2D map/3d model (Result (3/4)) (Result (3/4)) (Result (4/4)) (Result (4/4)) (Result (2/4))

12 Test Scenario The Station nightclub fire of West Warwick, Rhode Island, Feb. 20, 2003 Computer simulation for desktop system Simulated fire (300 sec.) Simulated evacuation of 419 people [Image Source: NIST]

13 Results (1/4) Moving entities and their movement path in 2D environment Each entity s condition changed at each time Alive (Healthy) Unconscious

14 Results (2/4) Moving 2D entities in 3D environment < moving entities in 3D perspective views (solid shading/wireframe modes) >

15 Results (3/4) Personnel-related information # of evacuated personnel through each exit Info. of selected personnel ( ID / effect by gas / status ) Evacuation graph

16 Results (3/4) Personnel-related information # of evacuated personnel through each exit Info. of selected personnel ( ID / effect by gas / status ) Evacuation graph

17 Results (3/4) Personnel-related information (40 ~ 60 sec.) # of evacuated personnel through each exit Info. of selected personnel ( ID / effect by gas / status ) Evacuation graph

18 Results (4/4) Emergency situation-related information Environmental condition change (Time = 76 sec.) (Time = 113 sec.) (Time = 76 sec.) (Time = 113 sec.) Temperature change Toxic gas (CO) change (source of fire)

19 Results (4/4) Emergency situation-related information Environmental condition change (Time = 76 sec.) (Time = 113 sec.) (Time = 76 sec.) (Time = 113 sec.) (kitchen area) Temperature change Toxic gas (CO) change

20 Implementation Development environments Microsoft Visual Studio 2005 Hybrid Graphics Rasteroid 3 OpenVG, OpenGL ES, EGL Open 3DS library Test environments Dell Axim TM X51v PDA Intel 2700G graphics processor Sprint PCS Vision SM smart device PPC MHz Intel processor Win32 emulation

21 Running on PDA and Smartphone Dell Axim x51v PDA Sprint PCS Vision smart phone

22 Conclusions Our system can be used as a general purpose Prototype of Mobile Visual Analytics for Emergency Response run on any devices using PocketPC at interactive rates used for mobile devices with various screen resolutions without deteriorating the quality of visualization Informal feedback Purdue University fire department, Purdue Homeland Security Institute Useful for training such as pre-planning scenarios and site inspection Need real-time tracking capability for real emergency situation

23 Future work As visual analytic tool, include more analytic functions to enhance emergency situational awareness e.g., 3D optimal path finding for virtual rescue To be used in real emergency situation, add real-time tracking visualize video information for training and in-field deployment support To be more useful, integrate social network data, e.g., family, friends, police, fire station, hospital and government department, etc.

24 Acknowledgements We thank the reviewers for their suggestions and helpful comments. We also thank the Purdue Homeland Security Institute (PHSI) for supplying the simulation data. This work has been funded by the U.S. National Science Foundation (NSF) under grants and , and by the U.S. Department of Homeland Security Regional Visualization and Analytics Center (RVAC) Center of Excellence.