On Tuesday March 17, the French government decided to lockdown the country in response to the COVID-19 pandemic. While the lockdown has had a very negative impact on the country's economy, it has also had a positive impact on air quality, especially in large cities such as Paris. That's why we chose to focus on the first challenge (Quiet Planet) and more specifically in the air quality aspect of the challenge.
As described in the challenge, several recent studies have shown that NO2 emissions from satellite data serve as an effective proxy for co-emitted CO2 emissions from cities. Our objective is to obtain daily data of the levels of NO2 in the city of Paris and use this data to encourage Parisians to use bicycles instead of cars as a means of transport. In this context, the project directly addresses the first challenge (Quiet Planet) and also addresses the goal of the French Government and the Town Hall of the city of Paris putting a 20 million euros government plan to encourage cycling during deconfinement.
We started the process by a brainstorm session to identify a specific problem around the first challenge (Quit Planet). Once we defined the problem statement: How might we motivate the citizens of Paris to continue to improve the quality of the air of the city after the COVID-19 lockdown?, we wanted to develop a solution using bicycles, especially after seeing the interest of the French government and the city of Paris in the use of bicycles after the pandemic.
The first step was to explore the variables affecting air quality in the world and specifically the impact of these variables during the lockdown in Paris. We started by exploring the NASA COVID-19 Pathfinder. We identified two interesting Datasets: 1. Find Environmental Impacts Data, and 2. Find Seasonality Data. The second one could be interesting in an evolution of the project (create a Machine Learning Model to predict air quality from Air Temperature, Humidity and Ultraviolet Radiation), but not for our current goal. So, we decided to only use the first Dataset (Find Environmental Impacts Data).
We identified four variables to do further exploration: 1. Aerosol Optical Depth/Thickness, 2. Nitrogen Dioxide, 3. Carbon Monoxide, and 4. Ozone. To analyze the Data, we used Python and the libraries Pandas and Plotly. Our strategy was to analyze the data only for the French lockdown period (January to May), we decided to compare what happened with the air quality variables in this period in 2019 (no COVID-19 lockdown) and 2020 (COVID-19 lockdown).
We created a Radar Plot that you can find here to illustrate the evolution in 2019 and 2020 (during the COVID-19 lockdown period) of the following variables:
Globally, we identified a difference only in 4 variables: 1. Nitrogen Dioxide (NO2), 2. PM2.5, and 3. PM10 and O3. Specifically, we decided to zoom each variable with a Line Plot to better understand the evolution of each one of these 4 variables. You can see the Plots here: Paris NO2 levels - Paris O3 levels - Paris PM10 - Paris PM25. Finally, we observed that the major changes and variations are observed in the NO2, PM2.5 and PM10 levels.
Exploring these variables that were impacted by the COVID-19 lockdown, we found that NO2 is one of the highly reactive nitrogen oxides (NOX) gases. Its major source in cities is the combustion of fossil fuels. It is generally produced in larger quantities by older vehicles with diesel engines. NO2 is also a main contributor to the formation of nitrates in the atmosphere and in the presence of ultraviolet light to the formation of ozone (O3). PM2.5 can be of primary or secondary origin. The primary fraction is directly emitted from the source (e.g. from cars). The secondary fraction consists of sulphate, nitrate, ammonium and organic carbonaceous materials formed through chemical reactions of gaseous precursor such as NOx, SO2, NH3 and VOC.
So, the main variable affecting Paris Quality of Air is NO2 (PM2.5 and PM10 are a consequence of NO2 among other factors). Note that regarding NO2 levels, we did not find the same trend in cities such as Lyon, Marseille or Berlin, where we only noted a difference in PM2.5 and PM10 levels. The difference in NO2 is more noticeable in cities like Paris or London.
We started to build a prototype of a mobile application that will retrieve the daily NO2 variable from NASA dataset and use it to encourage citizens to use their bicycles as a means of transport.
If we observe the EU reports of NOx, road transport is the main source of NOx urban pollution, especially in big European cities like Paris or London. A closer look on the road transport sector shows that NO2 in cities mainly originates from the emissions of diesel vehicles. Diesel fueled vehicles are responsible for the bulk of road transport NOx emissions across all EU countries, and this is especially true in Paris.
We used AdobeXD to create an interactive prototype of the solution, after the COVID-19 Space App Challenge, our goal is to make a User Research (UXR) to better understand how Parisians move around the city and whether the application could play a role in motivating them to use the bicycle as a means of transport. Our objective is also to do 5 user tests of the prototype to better understand the user experience of the application. With more time, the application is easy to build using a NASA (or a NASA partner) API (to recover NO2 daily data), React Native or Flutter for the UI of the application and Django (Python) for the Backend. We also explored the possibility to use a no code solution because we are friends of the people behind this no code solution project in France: https://oneproteus.com/.
The proposition is very simple to understand. We propose a mobile application that allow users to visualize daily the Quality of the Air retrieving the NO2 measure from NASA. Each time the user takes his bike, the application records the time and the kilometers he has travelled. For each ride, the user wins a bike to collect inside the application. The application has different models of bikes that the user discovers according to the kilometers covered with his real bike. Finally, we have some surprise bikes that represent different cities in Europe.
30 seconds video of the solution prototype: https://youtu.be/d88txORVxWU
The solution in 3 slides: https://drive.google.com/file/d/1Oazlkh-b7n2wq42DyaRj00YGt1y6ESYC/view?usp=sharing
Our solution uses the NASA Environmental Impacts Data from the COVID-19 Data Pathfinder. Specifically, our solution uses data concerning the emissions of Nitrogen Dioxide (NO2) that are at the same time responsible for PM2.5 and PM10 particles in the air. So, our solution consists in a simple call to retrieve this data to show it daily to the user, encouraging him to use his bicycle to replace the use of the car or even public transportation.
We used Python (Pandas and Plotly) to analyze the data, you can change the city (Paris in our case) to see if other cities around France (or Europe) follow the same pattern (the evolution of NO2 due to the COVID-19 lockdown).
The design principles (collect items [ex. bikes]) are research based in psychology and they are usually used to motivate and influence user behavior (ex. going to work each day using a bicycle instead of a car).
Finally, we want to build the final version of the app with a participatory design approach. Our goal after the Hackathon is to integrate users in the User Experience Design and Research process to better iterate and scale. So we plan to integrate different User Experience Research (UXR) techniques and probably use the no code solution mentioned above. If it is not possible, we will use React Native or Flutter for the UI and Django for the Backend.
