The project addresses the development of a pipeline for the treatment, analysis and prediction of air pollutants using deep learning techniques and enabling the indication of current and future air quality index (AQI) with the construction of a flexible dynamic API and interactive WEB interface.

As an early stage of the project, a mind map was developed that would allow us to visualize how the information provided by NASA could be correlated so as to deliver some value. It has been identified that much of the themes make it possible to deploy tools that contribute to government action (theme: "Smash your SDGs!") And that "From Curious Minds Come Helping Hands" is a centralizing theme for all tools. Thus, our ultimate goal would be to develop a tool that supports this idea, but to get to that centralized one needs to start with something specific.

So we chose the "Surface-to-Air (Quality) Mission" theme that will take an initial step towards graphically displaying poor air quality regions, initially using NASA-provided Los Angeles datasets; graphics for time series display using node.js; and tool for pollutant prediction and air quality index calculation using statistical and deep learning techniques (Multilayer perceptron feedforward back propagation) in python with the open source tensorflow library, maintained by Google. We then consider linking this data to global emissions inventories such as Geoschem and regional emissions such as CMAQ to fill in gaps and later also use burning and human health data that will generate global feedback on the problems facing the world, feeding an artificial intelligence that will add value to predictions and anticipation of government actions.