A case study on using AI to estimate air pollution in Bogota, Colombia. You'll work with sensor data from the city, compare baseline models, and build a simple interface to measure impact. Coding is optional: run provided notebooks or tweak parameters.
Now is where things get really interesting. Up until now, you've been learning what AI for good is about. You've been seeing examples of some projects and you've been practicing with a framework that you could take to approach any kind of AI product. Now you'll get some hands on experience in a project with a very specific goal. We are going to be looking at estimated air pollution in Bogota, Colombia for citizens and public health professionals. As you work through this project, you can keep in mind that while we are looking at Bogota, Colombia, this is a problem that people all around the world are working on. And you could take a lot of the techniques and methods that you will learn in this lesson and apply it to air pollution and related pollution use cases in your particular area. And so we'll start with an overview of the problem of air pollution, the dangers that it presents to society. And then next you'll see how various groups around the world are doing their best to measure air pollution, because some signals are easier than others to measure. And this in turn results in more or less data that's available for you to build your models on. After that, you'll dive into the case study that you'll be working on for the rest of this course, estimating pollution levels throughout the city of Bogota. You'll be walking through the case study using the framework that you saw in the last lesson of this course. So you'll start by exploring the project. We'll define a problem statement, consider the various stakeholders and look at the data to determine whether AI could even add value to this project in the first place. Next, you'll enter the design phase where you'll analyze the data in detail, run a comparison of some basic models, and then design an end to end user experience that you aim to create. After that, you'll create a simple user interface using the models that you've developed, and then you'll evaluate the success and impact of your project. Throughout the project, you'll be working directly with air pollution sensor data published by the city of Bogota. For this project, you'll be using Python code to explore and analyze the data. But don't worry if you don't have a background in programming or with Python. The code that you'll be running is already written and your job is just to run the code and explore the results. Occasionally, you'll get the chance to change a parameter in the code and see what happens. For example, you might run a different version of the model, but again, to change these parameters, you don't necessarily need to have a background in Python or programming. If you do have a background in Python, then feel free to modify the code and experiment a little bit. You can also check out some of the utility functions and other code that we've abstracted out of the notebooks, and then think about how you might use some of these methods as a starting point in your own projects.
