Failure Mode shows fewer upgrades is better


Thread Starter

Bronson, Robert

I recently sat in on an Aladom Reliability Centered Maintenance presentation. It was interesting to see how control system/safety system reliability (or lack thereof) is being presented to maintenance personnel. More specific was the discussion on failure mode. Six different failure modes were discussed. The bath tub curve - high initial followed by random and then high failure rates as products age - is not the only failure type and in fact nor is it the most common. Electronic failures are generally considered to be high infant mortality and then random.

The point of this is to reduce unnecessary maintenance. Not only will this save money but it is during maintenance that we are often vulnerable to other safety, production issues. This fits with my observation that: It's when our fingers are in the cabinets that we have problems.

Given this information, I believe we need to work harder at having fewer maintenance releases and fewer upgrades. I certainly haven't seen this from the vendors. The trend in industry seems to be more.

Darold Woodward

I would like to get some more information on this topic. I searched the web for Aladom and didn't find anything. Can you give us some more specific information on where this might be available?


Darold Woodward PE
SEL Inc.
[email protected]
I love this point . . . perhaps the most entertaining (and wasteful) approach I have seen is when Manufacturing Company A is purchased by Manufacturing Company B. I have now witnessed
this with at least 4 fortune 500 companies.

Company A uses brand X PLCs / Drives / Networks and Company B uses brand Y PLCs / Drives / Networks. Company B has a mandate to change company A over to all brand Y stuff.

Brand Y comes in full force and replaces brand X stuff (which is working in many cases better than brand Y) and after a few years of faltering performance and upgrades and retro_*fits* . . . the company has all brand Y . . . the vendors and management all celebrate the golden udder they have milked to achieve the STANDARD that looks nothing like it did when they started. Then .
. . . brand Y buys brand X, puts a new label on it and they then proceed to switch most of it back so it will finally run as well as it did in the first place!

In my mind, unless there is a compelling economic reason to retrofit / upgrade it shouldn't be done and any argument that points to implementation of a "standard" should take a hard look at whether or not a "standard" really exists in the first place.

Look at any PLC vendor in the market place and grade them on how many families of PLCs they have, how many software packages it takes to program them, how old the revision can be before it becomes useless, how many different network options they have, how many different network configuration packages you need to buy (the favorite price seems to be $1k / package), how
many gyrations do you have to go through to share a tagname database between the PLC, MMI and SCADA systems . . . I could go on and on and on . . . based on these criteria I have yet to see a
"standard" for any vendor of PLCs / drives / networks / stuff.

Whew . . . I feel better now : )

Ken Brown
Applied Motion Systems, Inc.