Just a simple lip-sync test that I made using drivers and shape keys along with a python script I wrote to import phonemes from a program called sapi lipsync that I found that matches phonemes from a text to where they are in a sound file.
I realize the robot’s facial expressions are rather simple but the same method will hopefully work at least as well on a nicer mesh with some better detail.
http://video.google.com/videoplay?docid=-4804939258981284297