Now you’ve seen two examples of car motion models. In the first example we assumed our car was moving at a constant velocity. But in the second example we said that the car was slowing down with a constant acceleration. And after three seconds had elapsed we ended up with different estimates for our car’s state depending on which motion model we relied on. In the case where the car was slowing down we had to include an acceleration value in our calculations in order to predict where it would move. And so we also added this value to our car’s state. In fact, how many variables the state of our car contains depends on what motion model we use. For a constant velocity model, position and velocity will suffice. But for constant acceleration you’ll need to include an acceleration value as well, and these are all just models. For any state you should always choose the smallest set of values that you’ll need to work with your chosen motion model.