One method is to use sensors on the outside of the stator which will react to the magnets on the rotor accordingly and give a signal that can be used to determine the position. Typically, three Hall-effect sensors will give a 60 degrees difference between two transitions.
The advantage of using sensors is that by knowing the state of each hall-effect sensor, the position of the rotor is known even at zero speed, on system power-up.
The main downside of using sensors is that they add cost and complexity to the motor, while also needing auxiliary power.