There is a far field microphone array built into the collar. Speech is processed by a speech recognition module for Arduino that manages the suit lights and relays. It is serial connected to a Raspberry Pi that performs speech synthesis via Cepstral.
