Skip to content

Latest commit

 

History

History
41 lines (36 loc) · 4.57 KB

README.md

File metadata and controls

41 lines (36 loc) · 4.57 KB
Raw

Submission of Idea for Code.fun.do++

MOTIVATION

We see dozens of large scale natural disasters happening all over the world causing mass destruction of human life and property, it is our duty as a technologically advancing society to try to predict, attempt to prevent and manage support and help the society in the case of such diasters using the up and coming fields of historical data analytics, prediction tools and modern day daily-use technologies.

AIM

We are making an app/web-app interface that aims to cater to the populations specific to regions suffering from any natural disaster be it an earthquake, flood, tsunami or tornados. This app would have a bi-ended approach using a rescuer and refugee side app and unique features adhereing to both these ends. It is our hope that using this app we could optimise manpower and resource allocation to a disaster affected societies and structure and stabilise help and rescue support to cater to those in its urgent need.

ASSUMPTIONS

We have to look at a set of assumptions taking into consideration our resources, availability of data and privacy of our users along with the technological limitations:

  1. We need to assume that the rescuers(Government or army services, WHO, hospital services, refugee campers,etc.) have our app downloaded and are acquainted with the interface.
  2. We also need to assume that our user end population suffering from the disaster has atleast SMS availability and has offline Google map on his device to figure out his/her location.
  3. We need to assume that we have a number on which the disaster-ridden people will SMS their location such that rescuers can find them.
  4. We are assuming that shortest and safest paths can be trusted based on the live user and rescuer inputs based on moral hazards and asymmetric information transfer.

KEY FEATURES

  • We build a database based on the location data we receive from the sms' sent out by the people or which we receive from the user end of the app.
  • We cluster the population density using k-means.
  • We optimally decide safe zones/Hotspots/Camps based on past disaster data, clustered population zones and the safety of a zone based on it's Altitude data, water level data and other quantitative parameters taken into consideration on the rescuer side of the application.
  • We decide red/danger zones based on data updated by both the users or rescuers if they see blockage or destruction of surroundings or routes.
  • We make a SOS/urgent rescue feature that caters to a proiority accidents or someone stuck in an emergency and it will be directly sent to the rescuers.
  • We would build integration with a map services app which would help us marking unsafe zones, relief camps and safe routes for both the users as well as the rescuers.
  • The app would give directions, safe and short routes to the users as well as send an urgent evacuation notification to users stuck in highly unsafe locations.

IMPLEMENTATION

  • On the basis of the location data received by the people we would create a cluster of people in a diameter of about 5km. The stranded people would receive the location of nearest cluster center so that they can gather at a common point so that it makes easier for resecuers to reach them with a smaller manforce.
  • The common point of gathering would be decided by the rescuers with help from the prediction based on current population clusters, altitude and water level data and hotspots(hospitals,helipad, resources) and the nearest gathering point will be conveyed to the person through SMS which would be integrated with Bing Maps.
  • The SOS feature is for the people who are trapped and cannot reach the safe locations without help from the rescuers. There location would be directly sent to the rescue operations team for immediate action.

dataflow

Tech Stack

  • Azure SQL Database
  • Azure App Service
  • Bing Maps API
  • Django/ Flask FrameWork in Python for Web-App
  • Scikit Learn/ PyTorch

CHALLENGES

  • Physical restraints to a rescuer based on resources and availability of manpower.
  • Time constraints on hotspots, routes and danger zone availabilities based on real-time problems like repetitive disasters
  • Route limitations based on the APIs used or algorithms!
  • Danger zones and blockage on routes precision as well as precision of hotspots based on other clustering algorithms used based on the location data.
  • SOS can be sent out by people in distress or by those who are desperate but we need to no moral hazard to cater to emergencies first.