Custom Application for Ford continued...

Over time, however, the design engineers found that they needed to quickly compare the results of individual samples in different tests. It was nearly impossible for the engineers to grasp the interrelationships between these tests simply by laying the plots out next to each other and shifting their eyes back and forth from one to another. For one thing, the axis scales on the different plots did not correspond to one another. The design engineers requested that all the test results for a particular sample be plotted on one multiple axis graph but were told that this is not possible with Excel.

They then asked that the axes of the individual plots be carefully scaled and aligned so that they could be fastened together and held up to a light in an effort to view detect interdependencies between the results of different tests. The need to scale and align the axes of the Excel graphs increased the difficulty of the graphing process. The engineers assigned to manipulating the data and generating the plots found that it took about 5 hours to create the graphs required for a single group of six samples. These graphs consisted of plots for each of the four main conditions - constant in, coast drive, converter stall and performance stall - for each of the three tests - capacity factor, torque ratio and efficiency - for each of the six samples. Design engineers made a sandwich consisting of all of the graphs for a single sample and held it up to the light in an effort to view interdependencies between different tests. The design engineers complained that the graphs were difficult to read like this and that a slight misalignment of the axes would give spurious results that they then had to spend time trying to track down.