LiveLeak: The Brown University Robotics Group has been able to teach a PackBot to respond to verbal commands, gestural commands, and is capable of following a human around. The robot has a depth imaging camera which is able to interpret both shapes and distances, enabling the robot to recognize gestures and the human giving the gesture by analyzing how the silhouette of the human changes. It can then correlate a particular silhouette with a particular human, and follow that human. The PackBot is very good at keeping a respectful distance, and also knows enough to back up if necessary.