In the previous administration, navigators offered various suggestions on what to do next. The instructions they gave were about changing lanes,
After the lane, turn, etc. But there really aren't as many types of advice you'd expect to hear from a navigator. In this lesson, we will teach a behavior planning method
Behavior planning problems are solved using something called a finite state machine. A finite state machine makes decisions based on a finite set of discrete states .
There are five states in this example. When initialized, the finite state machine starts out in some starting state, let's call it zero.
Any pair of states within a finite state machine can be connected by one or more transitions. Sometimes there is a transition back to the same state.
This is called self-transformation. Not all conversions are necessary.
For example, S4 will not transition to any other state.. In the language of finite state machines, this would be called the accepting state.
For non-accepting states, there may often be multiple potential inherited states. To decide to transition to the next state,
A finite state machine needs to process some kind of input and then use a state transition function to decide the next state .
We'll talk more about the transition function and the states associated with a self-driving car in a minute.
But first, we want to formalize what a finite state machine is, with the help of a simple example.