INSIGHT / Big Data

Delivered a comprehensive GIS-based data analytics solution to ensure effective real-time communication between decision-makers, support teams, and commoners.

Big Data

About client:
Client is the State Disaster Management Authority (SDMA) of a south-eastern state of India.

Scenario:
The state has the second largest coast-line among the 29 states of India, which made it inherently risk-prone to natural disasters. That besides, rapid urbanization had led to populous inhabitations and a wider adoption of information technology, especially the internet and social media networks, increased the susceptibility of civilian life from cybercrime and crime in general. That meant there existed an uncertainty over when, where, or how a weather-related or societal calamity might occur and cripple civilian life.

Consequently, the SDMA wanted to equip itself with critical information in real-time to help it in better responding to and managing both, man-made as well as natural disasters. It aimed to use the information to not only mitigate the risks associated with disasters but also to proactively monitor and spot any potential calamities well in advance to be able to take pre-emptive (safety) measures and minimize the impact of such events.

Scope:
The scope of the project was to integrate all the disaster information into a single platform from various sources and analyzing the same to arrive at patterns, spot outliers or indicators and passing on the information to help decision-makers / stake-holders. Disaster types included cyclones, tsunami, drought, epidemic, floods, heatwaves, landslides, thunderstorms, man-made disasters, air quality monitoring, river water, reservoir levels, rainfall, lightening, etc.

The solution was also required to integrate all administration-related activities such as employee on-boarding, asset management, events (trainings, mock drills, workshops and other programs) and decision-support activities. Generating PDF reports for daily activities to monitor disaster information and sending automatic notifications/alerts to personnel concerned in the event of a disaster. Critical of all was to facilitate two-way communication between disaster team(s) and commoners to eliminate or minimize the impact of a disaster / calamity right from the pre-disaster stage to during a disaster situation and afterwards as well.

Solution:
MSRCosmos implemented a solution incorporating a typical MVC framework to facilitate easy collection, aggregation, and analysis of the required disaster information from various sources. Multiple components were integrated to streamline the process to achieve the project objectives – primarily enabling faster response times during any disaster.

The following components/modules were incorporated for data integration from various disaster sources:

The Data Integration module fetches all disaster related information from various sources either through direct integration or through specific API calls at specified time intervals.

The Data Ingestion module involves data verification and saving the same into a database at specified intervals of time. A viable information ingestion process starts by organizing information sources, approving individual documents and steering information to the right destination.

The Data Preprocess module enables saving only the data that is required to mitigate disaster risks, while cutting out the unnecessary data.
The Reporting module is aimed at a component-rich and easy-to-use web interface for accessing historical reports that can help in making informed disaster forecasts.

The GIS-based Decision Support module collects information from raw data and integrates a Geo-spatial based user interface to show all locations (districts, blocks and village layers) to find infrastructure (roads, stations, tanks, river lines, major canals and such) and assets (hospitals, shelters, habitants, industries and such) for disaster preparedness and response. Also integrated spatial statistics and network analytics to assist the affected population with driving, cycling and walking during or after a disaster.

An integrated mobile app was also incorporated into the solution to ensure better and faster communication between various teams and commoners in the affected areas. It helps to reduce the communication gap between various disaster teams, people and victims. An integrated dashboard shows the statistics of requests and responses by teams, people, and locations.

Automatic notification alerts to stake-holders, response teams, and people (commoners & victims alike) were configured for specific time intervals as well as that are triggered when threshold levels of disaster impacts were recorded. The notification alert module is built to monitor threshold levels on each of the disaster types thus helping faster decision making and reducing response times to minimize impact of disasters.

Dashboards were built to visualize various indicators (rainfall, air pollution, reservoir, capacity building calendar, forecast & current weather information, forecast images, minutes of meeting images and document repository) on top of the various types of information related to disasters so that monitoring all levels (highest-high-moderate-low) could be facilitated. Dashboards display easily understandable graphical and drill-down representations of disaster related information.

Impact:

  • Reduced times and, faster and effective communication during disaster response.
  • Enabled live (data) tracking of disaster-related activities / situations.
  • Quicker risk identification, report generation, alerting & decision-making to take timely action.
  • Automatic alerts system is now helping various teams in responding faster to distress calls or any civic emergency.

  • Improved the process of disaster administrative activities and preparedness of the decision support systems which are so critical for disaster management and recovery.
  • Reduced communication gap between disaster team, people and victims.
  • Helping to reduce loss of life and property in disaster affected / prone areas.
SHARE: