The Arduino is set up with a microphone module that detects the noise from the smartphone and moves the RVR just after or during the speech to provide more realism. There is also a touch sensor for tactile input that makes the RVR purr, and an ultrasonic distance sensor. The distance sensor will follow your hand forward and backward on the spot when in desktop mode. But if you press the touch sensor during the initial setup, the RVR will launch in autonomous mode and explore the environment.
This is a small demo of what the RVR companion can do: