An RBF neural network approach to geometric error compensation with displacement measurements only
This report examines the results of the ball plate round robin administered by NIST. The round robin was part of an effort to assess the current state of industry practices for measurements made using coordinate measuring machines. Measurements of a two-dimensional ball plate (240 mm by 240 mm) on 41 coordinate measuring machines were collected and analyzed. Typically, the deviations of the reported X and Y coordinates from the calibrated values were within ± 5 μm, with some coordinate deviations exceeding 20.0 μm. One of the most significant observations from these data was that over 75 % of the participants failed to correctly estimate their measurement error on one or more of the ball plate spheres.