Reshape

https://youtu.be/WbgbJB3DPqw

https://reshape.duckdns.org

Worldwide, disparity in neighborhood quality is a multifaceted, complex challenge. We thought that the first step to solving any complex challenge is having a rich source of data to form and test hypotheses against. With that in mind, we envisioned a platform that can serve as a unified source for data that incentivizes change towards more equal, accessible urban planning.

To achieve this vision we began by developing an MVP (minimum viable product) as a proof of concept. To that end, we began by collecting the necessary data which we can use to calculate a preliminary set of indices that showcase our vision. 

To begin, we extracted data from satellite imagery using the EC-GHSL Population dataset (provided by NASA’s Sentinel-2* mission). In combination with local-specific data that is publicly available, we processed the data to extract intelligence about each district in Dammam. Some of the indices include: hospital availability, population density, and Internet coverage.

The collected and processed data is publicly available for researchers and urban planners. This allows others to build upon and use our data for further improvement. Additionally, we provide a publicly hosted web application to showcase a use case of the data.

The web application features Dammam districts, and visualizes the overall neighborhood quality index per district. This gives a high-level overview of the quality of life in different neighbourhoods. Clicking on a specific district further expands the information provided with specific indices used to calculate the overall neighborhood quality, shown on the right hand side of the web app.

Front-end source code:

https://github.com/yazeed44/reform-front-end

API source code:

https://github.com/yazeed44/reform-api

Next steps

We want to scale the current curated dataset as well as the indices covered. To do that, we will need expertise in urban planning and human development to come up with research-derived indices that are shown to factor into neighborhood quality, accessibility, resilience, and sustainability. We also plan to tap into the rich tapestry of data that drones can provide through processing drone footage using computer vision. Which would provide more accurate data for our indices calculations and open the possibility for new indices to be calculated, for example: road quality, traffic data, etc... 

Tools

• Pandas, Google Colab jupyter notebook, and GeoPandas 

For data processing and preparation

• QGIS 

To convert satellite imagery into a vector, and from that to geoJSON where we loaded it into GeoPandas

• FastAPI 

For publicly providing the data through our API

• Requests library 

For scraping local-specific data

• Google Cloud Platform

To manage Linux VM instances

• nginx 

To host the front-end website

• Duck DNS

Provide a DNS record pointing an easy to remember URL to our static IP

• European Commission's GHSL - Global Human Settlement Layer (Population)

In combination with publicly available data, we’ve used GIS data that captures population density to calculate services-per-capita. Doing so, we hope to draw attention to districts that need more urgent urban development. With our solution, these districts will benefit from increased inclusiveness, resilience, and sustainability.

The data for the population density inspired a heuristic for measuring the ‘popularity’ of a district based on the number of people that live there.

When the team first learned about the space apps challenge, it was clear that this would be a great opportunity to improve urban planning and derive social change through the use of data. As some of the team members live in districts that are in need of more attention, it was imperative that a solution be developed and the team was enthusiastic to answer that call.

The team itself, composed of four members, possess a wide-range of knowledge and technical expertise, making it a well-balanced one. 

To conduct the team meetings, and collaborate remotely, the team used several platforms for communication and development. For communication, the team used Telegram, Discord, and Jitsi in order to share files, screen share, and chat. For development, the team mainly worked on a Google Colab Jupyter Notebook which handled the grunt of the data processing work. Hosting, backend API, and frontend development were handled separately and integrated together later on Google Cloud’s platform.

A brief timeline of the project development is described below:

1 . Node.js

Static proof-of-concept website

This was used to gauge the UX/UI and envision how the MVP would look like

2. FastAPI

Fixed test data

This was used to test the functionality of the API in order to share the data with researchers

3. Data Collection and Processing

Data was collected from public sources and NASA supported satellite imagery

This data was processed and used to calculate different indices which were then used to calculate an overall quality-of-life index for each district

4. Integration

The different parts above were integrated together into a single, functional, MVP

Data

• Schiavina, Marcello; Freire, Sergio; MacManus, Kytt (2019): GHS population grid multitemporal (1975, 1990, 2000, 2015) R2019A. European Commission, Joint Research Centre (JRC) DOI: 10.2905/42E8BE89-54FF-464E-BE7B-BF9E64DA5218 PID: http://data.europa.eu/89h/0c6b9751-a71f-4062-830b-43c9f432370f
• Internet Coverage Data

Scraped from telecommunication service providers coverage maps

• Pubic services (mosques, hospitals, schools, etc.)

Scraped from Google Maps data

Tools

• Pandas, Google Colab jupyter notebook, and GeoPandas 

For data processing and preparation

• QGIS 

To convert satellite imagery into a vector, and from that to geoJSON where we loaded it into GeoPandas

• FastAPI 

For publicly providing the data through our API

• Requests library 

For scraping local-specific data

• Google Cloud Platform

To manage Linux VM instances

• nginx 

To host the front-end website

• Duck DNS

Provide a DNS record pointing an easy to remember URL to our static IP

This project has been submitted for consideration during the Judging process.