Testing Methodology
Testing was broken into two categories: scale error and drift error. In this study, scale error was
determined first in an effort to isolate and measure the impacts of drift alone. Each component
was assessed to quantify the precision and accuracy of the model viewer within TCH. Testing
was performed using two versions of TCH v3.2 with WLT (released October 2020) and v3.1
(released July 2020) without WLT.
Figure 1: Precision vs. Accuracy
Before assessing the ways in which precision and accuracy can be determined, it is important to
first define the difference between these two terms. Accuracy represents how well a
measurement compares to its true value (ie: hitting the bullseye of a target) whereas Precision
refers to the repeatability of such measurements (ie: tight groupings of measurements). Figure 1
above illustrates these differences.
Scale Error - Methodology
To assess the scale error within TCH, testing was focused on
observed dimensions and the ability to precisely measure
various components of a model. The accuracy of the model
viewer in TCH was determined by comparing model
dimensions to measurements in the real-world. At a 1:1 scale
these measurements should be the same, within a
reasonable tolerance.
4
