Just a couple thoughts: the sonic dispersion of the sensor is a cone, (30 degrees) which means having the transducers mounted horizontally leads to a distortion effect. As in, things nearer the output side are received faster than things approaching from the receiver side. Turning the ultra sonic range finder 90 degrees relative to the rotational axis eliminates this distortion. I wonder if they used polar coordinates in the mapping. As range would be theta 0-180, and Radius = distance. might make an Arduino version. save a couple steps in coordinate calculations.