In the first of a new series of articles discussing the issue of component and even skills obsolescence, Nigel Wallis outlines some of the basic decisions engineers need to make when tackling the problem
It is a fact that obsolescence, and the issues surrounding it, have been around since time immemorial.
However, not only have the causes of obsolescence changed over time, it is no longer seen as “just an aerospace/defence problem”- it is now having far-reaching consequences in other markets.
In the first of a series of obsolescence-related articles, I thought it would be relevant to “go back to basics” and look at the fundamental issues surrounding obsolescence.
What is obsolescence? A basic statement could be: “Systems have a longer life-span than the components within them”.
With regard to the causes of obsolescence, perhaps the most common example would be diminishing demand for a component meaning the manufacturer no longer views it as a profitable item to manufacturer- the fact that perhaps quite a few people may still need it notwithstanding.
Or, devices being replaced by newer technology – which is fine as long as you don’t have systems in the field that rely on the old technology.
Then there’s obsolescence caused by unavailability of material, and indeed obsolescence caused by legislation.
The RoHS (restrictions on hazardous substances) legislation is a particularly good example of this, having effectively “wiped out” whole series of components more or less overnight when it was introduced in 2006.
Obsolescence is not just limited to electronic components. Software, test equipment and indeed engineering skills can also be affected by obsolescence.
So what are the solutions to obsolescence? Let’s use the most common form of issue: An obsolete component.
If you were notified in good time, you could conduct a last-time buy. But then you have to ensure the parts are stored correctly in the right environment: It is of no benefit to store a component for 10 years only to find out they have deteriorated over time and no longer function.
The quantity also needs to be considered: To few and you’re back where you started, too many and you could have a significant inventory cost on your hands.
Often, of course, there are alternatives or substitutes available, but these can incur engineering costs in ensuring the replacement is Form Fit and Function compatible.
Semiconductor devices can be remanufactured from die, but this takes time and can lead to additional expense.
Another course of action would be emulation or simulation- again, this can take significant resource to achieve, and the design authority may also need to be consulted. Excess stock can be sourced from the open or “grey” market.
But never has the phrase caveat emptor been more appropriate: It is vital to have a robust and trusted supply chain, and it is imperative the appropriate checks and balances, not to mention some significant measurement equipment, are in place to avoid receiving counterfeits.
Finally, there is the option of “cannibalisation”. This is the re-using of components from other units. This can be done with the appropriate procedures, but perhaps is not suitable for certain safety critical applications.
In summary then, it is not possible to stop obsolescence happening. There are several solutions, none of them perfect and all with negative elements. One thing however is clear: Obsolescence is an issue that must be taken seriously- being aware of a component’s lifecycle and acting in good time can make all the difference.
Author is Nigel Wallis, business development manager at CMCA(UK) and is the current chairman of the Component Obsolescence Group (COG).
www.cog.org.uk