Not as far as I know. You could ask the researchers at the Bristol Robotics Lab or the University of Liverpool. They've all got active groups looking into it.
The autonomous system would be the AI control unit of a vehicle. With this in mind, and using Nevada's roads as an example of a test bed (both real and simulated), it would at a minimum follow all traffic laws (this alone would probably make it not pass a Turing test..:). It would then need to pass decision tests based on certain input scenarios (urban, weather, road conditions, etc). I believe these would mainly be detections of pedestrians/cars, road conditions, its own state (vehicle speed, orientation, heading), etc. Not a general test, and would require quite possibly a huge set of assorted scenarios. I've checked out Jubatus and Bristol Robotics Lab, and was wondering if there might be others.