We developed a Multi-Mission Explorer which is a generic tool to explore and visually fuse data from multiple missions, satellites and sources. Our agenda for making this explorer tool is to visualize all the related data of EO Dashboard together. 

We believe that NASA, ESA and JAXA have created their own set of independent tools, and generated data of specific regions for analysis when the pandemic hit. But these individual tools don't provide a unified way to view and analyze all the data from different agencies/sources at one place. That's where we believe, the reason EO Dashboard was created which is based on the RACE tool developed by ESA providing data from only European continent. So there is a need to build a true collaborative tool to visualize and fuse all the data of EO dashboard together.

Multi-Mission Explorer offers features like:

• Fusion of multiple dataset onto single reference view
• Viewing time-series data with map with timeline comparison
• Compare multiple layers/Indicators together to form a unified view
• Plot the machine learning detections onto the map.

We chose this challenge because it involves interacting with different data-sets. Rather than focussing on a single use-case our solution to this challenge can be generic enough to generate insights which can complement other challenges and visual fusion of data is most exciting and challenging task.

Our approach was to develop a generic UI based tool which help to visualize the data coming from NASA, ESA and JAXA's Satellites. We created a lightweight application using React.js and Node.js. We used HTML, Javascript to write the code and we used several open source javascript plugins to ease our development, we used Mapbox api to render our maps. We integrated this application with eodash by adding a new button on eodash home page and providing a re-direction to our node based application. For collecting, analyzing and understanding the data, we used python in Jupyter notebook where we mostly interacted with SentinalHub apis. We hosted both the applications onto google cloud platform so that it can be easily accessed by everyone.

We wanted to render data in real-time in our application but we faced challenges with cross origin restrictions from our app.

We also faced a few challenges as the data which we queried from SentinalHub was not returned in geotiff data map. Another major challenged which we faced was with respect to Recovery Proxy Map. For some reason, the Recovery proxy maps on the EO dashboard did not have the comparison time/window. But, we also got great help from the SMEs to solve our issues. 

We used data from EDC's SentinalHub and Nasa's Open-source Covid API hosted on AWS instance in our solution. We majorly used SentinalHub's APIs for researching and studying the data.

We analyzed that all the data in EO Dashboard's Indicators is coming from Nasa's Covid API, hence we used all those APIs in our application and consolidated the data representation based on Areas/Regions so that all the related data to a particular region can be analyzed together.

Additionally, we also tried to create and use data/tools from other open-sources and tried to run detect the rooftops from city of San Franscisco.

Please find the application demo files, links to source code & documentation on this google drive - https://drive.google.com/drive/folders/1IZhdHL19wWk5W5QzlkS0oyAmBA24WfGg?usp=sharing

We have also hosted our application on google cloud platform which can be used to view the live running apps - 

eodash: http://35.239.104.198:8080/

multi-mission explorer: http://35.239.104.198:8081/

We already tried to integrate our solution with Earth Observing Dashboard, though hosted as a separate application. The details are mentioned in the integration document available on google drive.

The source code of customized eodash application (forked repository) and multi-mission app is available at https://github.com/GoldenEye-EODHack

The links to the Integration documentation would be available in the google drive - https://drive.google.com/drive/folders/1IZhdHL19wWk5W5QzlkS0oyAmBA24WfGg?usp=sharing

We have used following data and resources:

1. Nasa's covid_api - https://8ib71h0627.execute-api.us-east-1.amazonaws.com/docs#/
2. Map rendering library - https://www.mapbox.com/
3. For ML detection - https://github.com/dselivanov/robosat.pink
4. SentinalHub APIs - https://docs.sentinel-hub.com/api
5. SentinalHub request builder - https://apps.sentinel-hub.com/requests-builder

#multi-mission, #explorer, #fusion, #visualization, #layer_comparison, #generic_tool

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