Let’s talk a bit more about why uncertainty is so important in the field of robotics and self-driving cars. We know that measurements like the speed, the direction, and the location of a car are challenging to measure and we can’t measure them perfectly. There’s some uncertainty in each of these measurements. We also know that many of these measurements affect one another. For example, if we are uncertain about the location of a car, we can reduce that uncertainty by collecting data about the car’s surroundings and its movement. Self-driving cars measure all of these things from a car’s speed, to the scenery and objects that surround it, with sensors. And though these sensor measurements aren’t perfect, when the information they provide is combined using conditional probability and something called Bayes rule, they can form a reliable representation of a car’s position, its movement and its environment. Let’s take a look at Bayes Rule.