🚀 go-pugleaf

On Thursday, January 25, 2018 at 3:54:14 PM UTC-5, Peter Flass wrote:

> > The modern mindset is to accept that the failure will occur despite 
> > preventive maintenance and instead work on making the failure 
> > convenient. You can prevent failure by throwing money at the problem 
> > (which is a diminishing returns situation) or you can detect an 
> > impending failure and just schedule a replacement at the next docking 
> > period, production line shutdown, holiday period etc.
> 
> Sounds like that IBM Watson ad where it schedules airplane maintenance.

Was this in the late 1940s with tab machines?  IBM did a lot of
advertising back then.

In the 1960s, an engineering firm developed railroad maintenance
software.  By tracking failures of every component, they were
able to identify problem components, more so than practical before.
On a big railroad, it ran on a 360-65.  On a small line, it ran on
an 1130.





> 
> > 
> > Most items, say a bearing, will degrade over time in a regular and 
> > predictable way* from tiny pitting in the bearing face,noise and 
> > deflection from the bearings hitting the 'potholes' that result in heat 
> > and vibration, through the complete collapse of the bearing. At various 
> > stages of the collapse the impending failure is detectable through 
> > various means such as vibration monitoring, temperature monitoring, 
> > acoustic monitoring etc. If you can detect it early enough then you can 
> > say with reasonable confidence how long it will be until failure.
> > 
> > *note that an individual bearing will show a wear-out curve but the 
> > point at which the wear out starts is random hence the graphs for 
> > populations of bearings show a constant failure rate flat graph.
> > 
> > sorry rambling on... 30 years at this and I get a bit geeky.
> 
> This sounds good for mechanical parts, but what about electronics? Things
> seem to fail catastrophically with lots of magic smoke.
> 
> -- 
> Pete