Human Factors

The emergence and spread of infectious diseases, like COVID-19, are on the rise. Can you identify patterns between population density and COVID-19 cases and identify factors that could help predict hotspots of disease spread?

Obelus

Summary

Obelus project has the objective of building software that helps health agencies acting in vulnerable areas (like favelas and slums) by getting to know their demographic profile better, therefore improving their ability to take contingency actions against COVID-19.

How We Addressed This Challenge

As of now, low income areas like the favelas in Brazil are not only suffering from underreporting, but also from a lack of support from their health and governmental institutions based on the inaccurate epidemiological models that are used to analyze the virus progression in those regions.

In order to consider the unique lifestyle of those areas, the Obelus project thinks of those places as independent regions. Thus it is possible that our epidemiological models include most peculiarities that would not be available without it, leading to a more precise result.

Correlating satellite images, population distribution data and anonymous mobile phone call and text detail records (CDR's), Obelus is able to operate as a DaaS (Data As A Service) platform that offers complementary data to those who need it.

How We Developed This Project

Obelus gets it's necessary information from 3 resources:

1. NASA WorldWind - Interactive UI to visualize data;

2. WorldPop - Datasets containing mobile phone call and text detail records from various regions;

3.  WOPR Vision/API - Population estimates for specific locations and demographic groups.


The project proposes a method that follows the following steps:

1. Determine the lines that indicate the internal migration from the datasets from WorldPop and the Flowminder Foundation;

2. Overlays the geographical locations and frontiers (also from the Flowerminder Foundation and WorldPop datasets), locating invalid places for our considerations (like oceans and any off-land regions);

3. Perform the division of the area based on the geometry of the lines presented, creating smaller local sub-regions; 

4. Using linear algebra, defines the perimeter, area and coordinates that defines the sub-regions;

5. With WOPR data, retrieves the estimate population residing in that sub-region and some of its details;

6. Analyzing the internal migration that crosses the sub-region, estimates the average number of wanderers per day;

7. With the local population and daily migrations in mind, calculates the probability, and therefore the number of people  an individual gets in contact in a daily basis;

8. Using NASA WorldWind, displays this info in a user-friendly way to all its users.


For the initial demonstration we decided to exemplify this method with data and estimates from Nigeria. We selected this country due to great data availability and easier manipulation.

Project Demo

https://drive.google.com/file/d/1U47KGFavoa5WrBU-Ir-c0KwzfjS9RD9_/view?usp=sharing

Data & Resources

https://worldwind.arc.nasa.gov/

https://apps.worldpop.org/woprVision/

https://www.worldpop.org/geodata/summary?id=1283

https://www.worldpop.org/geodata/summary?id=1281

https://www.worldpop.org/region/ebola

https://web.flowminder.org/


Tags
#equality #epidemiology #data #forecast #population #prevention #vulnerability
Global Judging
This project was submitted for consideration during the Space Apps Global Judging process.