On Thu, Dec 9, 2010 at 2:53 PM, wrote: >  A question about LAGR. In the TR paper referenced in the ROS docs, you say in the intro that the controller accounts for the non-zero turning radius of the robot. However in the control section you mention that it samples trajectories with pure theta-dot as well, but this should not be possible with non-zero turning radius? Ah, that's because the LAGR robot isn't actually car-like; it has two driven wheels in the front and two passive casters in the back. So you can command pure theta-dot, and it will rotate in place, about the center of the front axle. brian.