Log in

No account? Create an account

June 30th, 2014

We have all suffered from changing requirements. We are almost done with implemention, maybe even most of the way done testing, and then a change in requirements invalidates all of our hard work. This can of course be extremely irritating.

It turns out that changing requirements are not specific to software, nor are they anything new. Here is an example from my father's area of expertise, which is custom-built large-scale food processing equipment.

My father's company was designing a new-to-them piece of equipment, namely a spinach chopper able to chop 15 tons of spinach per hour. Given that this was a new area, they built a small-scale prototype with which they ran a number of experiments on “small” batches of spinach (that is, less than a ton of spinach). The customer provided feedback on the results of these experiments, which fed into the design of the full-scale production model.

After the resulting spinach chopper had been in production for a short while, there was a routine follow-up inspection. The purpose of the inspection was to check for unexpected wear and other unanticipated problems with the design and fabrication. While the inspector was there, the chopper kicked a large rock into the air. It turned out that spinach from the field can contain rocks up to six inches (15cm) in diameter, and this requirement was not been communicated during development. This situation of course inspired a quick engineering change, installing a cage that captured the rocks.

Of course, it is only fair to point out that this requirement didn't actually change. After all, spinach from the field has always contained rocks up to six inches in diameter. There had been a failure to communicate an existing requirement, not a change in the actual requirement.

However, from the viewpoint of the engineers, the effect is the same. Regardless of whether there was an actual change in the requirements or the exposure of a previously hidden requirement, redesign and rework will very likely required.

One other point of interest to those who know me... The spinach chopper was re-inspected some five years after it started operation. Its blades had never been sharpened, and despite encountering the occasional rock, still did not need sharpening. So to those of you who occasionally accuse me of over-engineering, all I can say is that I come by it honestly! ;–)