How much Data is enough Data?

At the beginning of my career I worked at a large aircraft manufacturer in the Northwest as a Jig Builder; fabricating jigs that ranged from small tabletop size to jigs that were over a hundred feet long and four stories tall. I learned to measure with Vernier Micrometers, Calipers, Height Gages, “Suitcase” Micrometers, Ball Gages, Adjustable Parallels, Transits, Optical Levels, Alignment Scopes etc….Single dimension measurements….it took multiple setups and a great deal of time and skill to properly set tooling details to the tolerances we were working back then (+/- 0.010 inches) and occasionally working as close as +/- 0.005 inches.

Working with single dimension measurement devices often required the use of three different types of measurement devices, three different indexing setups to set a detail in all three directions (X, Y, Z) or in aircraft terms; Station, Buttline, Waterline. This is how Aircraft Tooling was typically built until the early 1990’s when Computer Aided Measurement Systems became user friendly enough for shop use. For decades prior to 1990 Tooling was a highly skilled profession requiring the “Toolmaker/Jig Builder” to have a detailed understanding of several types of measurement equipment and how to properly utilize them in order to fabricate the tooling to the tolerances required. The Toolmaker/Jig Builder would also be called upon to verify aircraft parts to tolerances of 0.030″, this might include taking measurements at specific locations on the “skin” of the aircraft by holding by hand optical scales that were read with two different measurement devices, a time consuming task.

In the late 80’s early 90’s Computer Aided Theodolites (CAT) became available and quickly became the “measurement device” of choice for building tooling. The CAT Systems allowed the Toolmaker/Jig Builder to measure all three dimensions from one setup for the first time ever. While these devices increased the time in which tooling could be fabricated, they still required a great deal of skill to obtain the precision and accuracy required for aircraft tooling. Verification of aircraft parts could now be accomplished by placing sticker targets on the aircraft and measuring them with at least two theodolites (they are a triangulation based system).

By the mid 90’s Laser Trackers made their appearance and as fast as the CAT System appeared it quickly became obsolete as the Laser Tracker could provide real time 3D Measurements, Laser Trackers were not as difficult to setup as the CAT system or earlier one dimensional measurement devices and were able to achieve a much higher degree of accuracy and quickly became the measurement device of choice all around the world for Tool Fabrication in not only aerospace but the automotive industry.

The arrival of the Laser Tracker with improved accuracy and the ability to measure 1000 points per second, also brought about big changes in tooling tolerances, suddenly, the +/- 0.010 inch tolerances of the past, were no longer acceptable, with many Engineers expecting that tooling could be built to the tolerance of the Laser Tracker (=/- 0.0020 inches)….Aircraft parts could now be measured by “scrubbing” the surface with the Laser Trackers spherical mounted reflector (SMR) collecting thousands of points very quickly and the ability to characterize surfaces faster and more accurately then ever before.

By the end of the 20th century Laser Scanners of all varieties became available and now millions of points could be collected very quickly without having to contact the surface of the part being measured.

So with all of the technology, the ability to measure faster, collect more data Tooling and Manufacturing have improved dramatically in the aircraft industry, however, the overall result may be even better if the Engineering requirements in some areas were a little more reasonable.

The ability to measure more accurately and precisely should not always equate to the “tightening up” of tolerances as has been seen worldwide, the reality is our measurements are more accurate then back in the 1D measurement device days. When tolerances are “tightened” in almost every situation it equates to a higher cost for the end product as more time is required to meet the new tolerance. The ability to measure millions of points on parts that were previously measured with hundreds of points does not always mean a better product as the data must be analyzed and deciphered, sometimes the “overload” of data can actually stop production as a decision is made on if the part is good or bad.

Having spent a lifetime working in measurement I have enjoyed being on the cutting edge of the new measurement technologies and would not trade it for another profession, but I do think back on the “old days” occasionally and wonder; how much data, is enough data? There are many airplanes and automobiles still operational today that were built using 1D measurement devices, while clearly not as accurate as today’s products, are we getting to the point of data overload? Are we expecting too much of our measurement systems given the limitations of manufacturing materials and environments?