Of tracers and traceability

I recently had the good fortune of chatting to an automotive engineer at a secret and highly unusual testing session. Such vagaries of course hardly serve to enlighten particularly but I mention it to show that I do get out and about occasionally.

Naturally, as two engineers, we ended up discussing various engineering type things. Specifically he told me about when he worked for a well known luxury car manufacturer who, to protect the innocent I shall call ”JB Cars.”

Apparently there was a rather sad and unfortunate incident where one of their American customers was killed when a drunk driver ploughed into the back of his “JB”. Such was the force of the impact that he was thrown from the car still strapped to his seat. America being America the family sued claiming that the seat rails had not been up to the job. As sympathetic as one may wish to be in such a situation this was a claim that, of course, had to be refuted.

Thanks to modern data handling systems the full build sheet for that particular car, along with the design analysis and test results for the seat rail, were returned to the lawyers within 24 hours - and as one would expect the case was dropped.

This is an extreme illustration of the everyday power that the computer revolution has brought to our desks. It is something that we now see as commonplace and unremarkable but which is in fact utterly astonishing. I am hardly an “old lag” but even in my time the change has been immense.

Although “tracers” were an extinct breed when I joined my first design office, they weren’t long gone as I was given a tracer’s board to work on (it was the only one available). This was in a top rank aerospace office as well, hardly an innovation back water. Today I suspect you are lucky to even see a single drawing board of any sort in use anywhere in industry.

We all know the advantages of this metamorphosis but it strikes me that we are reaching a tipping point.

No longer do we share time on slow mainframes battling with clunky wireframe models. Solid modelling with various analysis packages gives most of us the capabilities that whole departments used to provide. But this generates an awful lot of data that has to be stored and, as seen above, readily available if it is to be of any use.

As we press on into the future, data storage and control as an intrinsic part of the wider design control will inevitably become of paramount importance. With the data control system we have at Amalgamated Products Limited we were recently advised “not to delete anything.” Rather if there was a problem we were to blank the offending item from our minds because deleting it “could be tricky” and “after all memory is cheap.” Surely an unsustainable strategy from any point of view and one which has to change?