Worn out Green Spaghetti

Picture of Adam Bahret
Adam Bahret

Like all families, mine has a few holiday traditions. Some traditions are passed down, like our recital of the poem about the Italian Christmas Witch “Befana” who brings children a single present on Epiphany Eve. Other traditions just “come to be”, like in late January when I get to set fire to everything that is made of awesomely explosive dry pine in the patio fire pit: wreaths, entryway garland, and that big long centerpiece from the dining room table with the candles in it that makes everyone nervous anyway. I always dress appropriately, wearing my t-shirt with a “safety third” graphic and fire proof sandals.

However, there are also the traditions that are really just burdens that occur with such regular consistency that they mockingly get labeled as traditions. In this category we have me fixing broken Christmas tree lights while everyone waits to decorate the tree. Ok, before you get all in my face about how now they have LED Christmas lights and other incandescent types that don’t take out the string when they fail and blah blah blah… Look!!! I know all of that! But, just like you, I am an engineer. I’m not replacing something I can fix… and is as an excuse to buy a $50 tool to solve a $6 problem.

So, we have a fake Christmas tree for several reasons, I’m sure one of them is to restrict the amount of fuel I have for my post-Christmas backyard explosive blaze. It’s actually a great tree that looks fantastic despite being 12 years old. At the time, it was top of the line (as I was told by the guy who sold it) and has lights woven into its branches. A very appealing feature where your target market is clearly lazy people. The problem is that after so many years there are sections of the lights that have completely gone out. Thus, the tradition of “Watch Dad figure out which bulb is the bugger that’s causing everyone to wait around.” I have, however, submitted to failure a few times and just added a small string of lights into those dark patches.

So, this year when we took the tree down, I thought I would be proactive and find & fix all the burnt bulbs, every last one, before putting it away. Thus mitigating the tradition of angry elves glaring at a man arguing with a plastic tree. Genius! After all, I’ve only been prescribing preventative maintenance programs to the world for like 20 plus years.

I started my task of “preventative maintenance” after we stripped the tree. I pulled out the new tool that I had just “coincidentally” bought, which can detect an open line connectivity EMF by just being in proximity to a wire, and got to work. Unfortunately, five minutes into this task I realized that my plan really wasn’t all that great. I’m sure when we set the tree up next year and flip the switch they will all be lit. But I am also sure that in the first few weeks after setting it up we will have one or two new burnouts.
So, how is this method preventative?

Yes, I am doing maintenance to improve performance while in the customers hands—the aforementioned decorating elves. And yes, when I am done it will be ready when called into service. Yet, I have done nothing to actually prevent failure. I don’t think maybe there will be failures, I guarantee there will be new failures. It’s guaranteed because we are clearly deep into the wear out curve for these light strings. They are all in the third stage of the bathtub curve, wearout. So a guaranteed disappointed customer is my objective? Angry elves don’t make nice treats or give presents.

I realized that I am in a never-ending cycle of repairing a failure mode that is not even a failure at this point. These “failures” are how the typical life of any great christmas tree light bulb ends. They were probably designed for a five-year life and held on for an amazing 12 years. That’s not failing, that’s just impressive.

I’m a monster! These bulbs don’t deserve to be cursed at as they are being hunted in fake pine branches. They should receive a service and burial that includes kind words from colleagues such as the kids’ elementary school ornaments with glued on glitter.
“There’s no sparkle with just glitter. You need light to make the magic happen,” we so often forget that when we compliment ornaments.”

I knew what I had to do; all the lights had to go. We needed to reset the clock and put the tree’s illumination system back in the early part of the bathtub curve. All of it, the whole thing was in phase three (wear-out) of the life curve, and there is no phase four.
So, switching tools, I put away my cool new proximity conductivity checker and took out an old pair of wire cutters. I started the surgical task of cutting out all the carefully interwoven lights. In no time I had a pile of green spaghetti on the floor.

But wait, don’t go anywhere. We didn’t solve the problem yet. Did we really do anything preventative? Sort of… I mean, I did reset the entire string back to the beginning of the bathtub curve, i.e. new components. Ok, so what happens next as far as Christmas lights go? If the new set works perfectly then in about six years or so, I’m going to be reaching into that tree again trying to figure out which bulb blew. I won’t be grumbling so much now that I am more conscious of the true struggles of Christmas lights. I’ll just say “you did a great job little buddy” as I extract and replace each one.

But, in the end, I’m still reacting to failure. Prevention means you “prevented” something, not just postponed what is inevitable. Let’s back up.

The objective of preventative maintenance is to have the user not experience a predictable failure mode, typically this is a wear-out failure mode. Making faster tow trucks is not preventative maintenance. In my case, preventative maintenance would be changing strings or bulbs in advance of a failure based on a known performance profile or measurable parameter. So simply replace the entire string at year four or at 1,000 on/off cycles. It could be that I measure resistance in the line or current draw and have a degradation curve to convert that into remaining life. In my case, I’m not trying to get every last second out of each bulb. I’m going to assume the light manufacturing people have done their homework, so I’ll just change it at the design lifetime they advertise on the box. Every five years, I’ll throw out the old lights and put on new ones. PM done, happy elves.

So this is what preventive maintenance is. A prescriptive action that when carried out will reduce field failure rate. It can’t be done to mitigate quality or random life failures. Preventative maintenance will only work if we have a degradation relationship between use and performance. If we study it we can just about eliminate that failure mode from occurring.

Here are some tips for getting started.

Investigate the percentage of wear-out failures your customer’s experience with your products? Putting a number to this can be a great asset in justifying investment.
Create preventative maintenance studies outside of a product development program when possible. When a study is linked to a product development program it is subject to all the forces within that program, schedule changes, budget cuts, facilities reallocation.
Look to industry to see what studies have been done. There are probably 1,000 PhD thesis out there on bearing grease degradation. The only reason you should be doing a bearing grease study is if you have extremely unusual use or environmental conditions that can’t be found in “the tables.”

So start small. Grab a bench in the back corner of a lab and pick a wear-out failure mode you want to know more about. Full systems are not required. You often only need the parts associated with the actual failure. So older sub-assemblies with only the parts under study being replaced may be all that is required.
It’s interesting how funding for larger wear-out studies seem to magically appear when a little under the radar study provides valuable information.
-Adam

Share this post