Users create a bus route following a pattern. They then train a model, which learns a pattern based on the type of store and the location of stores. With this information it predicts its own route, taking the shortest path between stops, and is able to incorporate new stops into the route.

The user places stores for the bus to stop at. Finding its own pattern, based on location and store type, the model predicts a route for the bus. 

After completing both simulations, the user is able to learn the differences between Supervised and Unsupervised Learning. There is more information on the website as well.