The Uncanny valley of mimic design

Was at a presentation on robotics by the clever folks at iRobot, makers of the iRoomba, in which they discussed the uncanny valley as well as an improved interface for a military search robot.

The original interface was put together in a weekend for a critical mission in Afghanistan, while an amazing job, it was engineering centric. The new interface had a 3d model of the robot indicating position, bearing, pan/tilt etc. It was a fully rendered 3d image, very nice, but a little hard to make out some details as the rendering wasn't 100%, it had some polygon artifacts.

Thats when I made the connection between the uncanny valley and my own research on displaying complex information through mimics. If you reduce the detail in a mimic you reduce the cognitive load, in addition, with the mimic 'cleaned up' you can introduce additional data-points