Big Data Analytics System for City Emergency Alerting

Big Data Analytics System for City Emergency Alerting Advisor: Prof. Jerry Gao Koushik Ram Sindhuja Narra Saikrishnan Baskaran Jeyanthh Venkatachari Ravikumar Introduction • • Our project is a smart city emergency project to conduct a comprehensive study about the existing city emergency alerting systems and infrastructures using the data-driven approach. To overcome the shortcomings of legacy disaster alerting systems, this project will provide a big data based analysis system for studying, evaluating alerting, and analyzing different types of city emergency alerting mechanisms with wider coverage to accurately deploy and distribute the available resources. Existing Alert Systems Alert. SCC: Alert. SCC can send text or voice messages to all mobile devices, cellphones, landline phones, laptops, IPADs, desktops computers, and to the telecommunication devices used by the hearing and speech impaired. Alert. SCC can send alerts and notifications only in English. Project Objective #1: To find out the data-driven emergency alerting coverages for four different nature disaster scenarios, such as earthquakes, floods, fire accidents, and so on. Objective #2: To find the system performance, limitations, and research ability problems in underlying emergency system infrastructures. Motivation • Mass Warning System available for use by the city of San Jose, (population 1, 015, 000 residents) should be capable of warning 90% of the population within ten minutes of an occurrence • Need to gauge the effectiveness of alerting systems in San Jose for high probable disaster scenarios like Anderson dam failure, Oil train derailment, Earthquake, Biological hazards • Need for an application that provides a proof-ofconcept for analysis of alerting systems in San Jose and their ability to alert their target demographic in a timely manner. Implementation Data sets Biological Flood Earthquake • The different types of data like demographic and alerting system coverage and statistics were cleaned and stored into the My. SQL database. • The analytics engine runs on Node JS Server as an independent web service and exposing REST APIs for the clients to communicate with. • The results of the computation are summarized and wrapped in a http response and sent back to the web client. IPAWS: IPAWS receives and authenticates messages transmitted by using alerting authorities and routes them to IPAWScompliant public alerting general public via radio, television, cell phones, social media. Oil Train Derailment Logistical Formulae Flood: Oil Train Derailment: Qp = 1. 268(Hw + 0. 3)2. 5 Ri : ∑[ Qi*CEIi ] Where, Qp – Outflow Discharge Hw- Outflow Discharge Where, Ri – Radius of Impact(Oil Train Fire) Qi – Quantity of Oil per container (in gallons) CEI – Chemical Explosion Index Earthquake: M = log 10(A/T) + Q(D, k) Where, M- Magnitude A - Amplitude of ground motion (in microns) T - Time period (in seconds); Q(D, h)- correction factor D - Degrees between epicenter and station and focal depth k - Kilometer for earthquake Demographics sample Screenshots of dashboards showing the impact and statistics for each scenario: Methodology Conclusions • The project is a prototype for disaster analytics systems by studying, analyzing and interpreting the data from multiple alert system and mapping them with demographic data. • It provides a solution for city planners and emergency operations center operators a way to gauge the impact of a disaster and the effectiveness of the alert systems available. Biological hazard: I 2 = I 1 x D 12 / D 22 Where, I 1 -Intensity 1 at D 1 I 2 - Intensity 2 at D 2 D 1 -Distance 1 from source D 2 -Distance 2 from source Results • The web client is the dashboard running on the user’s browser. The Dashboard values are populated dynamically using REST calls made to the Node. js web services running in the back end. • Node JS and its HTTP libraries as backend server for handling requests • AJAX calls for these asynchronous communications between for the Web client and the analytics engine Acknowledgement We are deeply indebted to Prof. Jerry Gao for his invaluable comments and assistance in the preparation of this study, The San Jose City Disaster Management Team and Prof. Dan Harkey for their special interest in our work and encouraging words that kept us going. www. postersession. com