machine superstition

[IO_Rack]

Five years ago I was sweating in a cleanroom "bunny suit", a few meters from the Sea of Japan, a hot and humid August day between Osaka and Kobe.

I've been in Osaka in August. It's brutal!

ゴジラ.

東京にいなくてラッキーでしたね。

 

[cardosocea]

I've been in Osaka in August. It's brutal!

Can confirm, it's horrendous.

 

[WhinnieThePooh]

Here in central Wisconsin it seems as though 90% of the plants we work in are builded upon sacred Indian burial lands. Gremlins are also common.

 

[I_Automation]

I worked in a shop where the boss logged all machine breakdowns - not my response time or repair time, just how often each machine broke down.

He called me in when he noticed an ongoing trend across the shop.

As soon as one machine or line was fixed the machine next to it broke down immediately. If the loader on a line was down as soon as it was running the unloader broke down.

The breakdowns were not related in any way nor the machines connected together even.

With a straight face I told him it was jealousy - the other machines were jealous another machine was getting all the attention.

Then I added that was the only possible explanation because there was no other explanation at all

 

[Ken Roach]

I worked on a homebrewed ASRS for stacks of particleboard early in my career (up in Sault Ste. Marie !). One of the complaints I got after we finished commissioning was that the storage cart was "afraid of the new forklifts".

Wait, what ?

"Sometimes when we drive by on one of the new forklifts, the cart just takes off in the other direction. Doesn't happen with the old Hysters."

The system had been built with a long-range retroreflective laser distance sensor to determine the position of the storage cart. For some reason they'd installed the reflectors it used on a wall on the far side of a cargo and personnel aisle.

I had tuned the sensor to ignore no-return conditions where the beam was blocked or absorbed by people or vehicles or spilled stacks of particleboard. I had not tuned it to reject a reflection from the high-viz sticker tape on the new fleet of forklift trucks.

 

[ASF]

That reminds me of a commissioning project where I had a graduate engineer helping me. Nothing really tricky, just roller conveyors moving tubs around in particular patterns.

Whenever I stood and watched, the conveyors behaved perfectly. As soon as my graduate took over, things would immediately go south.

Turns out that even though we'd tuned all the PE's down to only detect a crate 100mm away, his brand new hi vis vest was clean and shiny enough that they'd detect the reflective strips on it from over a metre away, and immediately try to discharge the rogue "tub" out of the way.

 

[JeffKiper]

20 years ago, a machine would just stop. The light curtain would flicker. I aligned, tested, algined. Great I had it and moved on. 2 to 4 days later same rinse lather and repeat. Replaced components with the same brand and type. I called the light curtain manufacturer told me to put a "GREEN" trash bag over the light curtain reciever and see if the problem went away. sure enough not a problem for 2 weeks. Removed the trash bag and the problem returned.

It was the strobe light causing interference with the existing light curtain. It was a good one to find. It only happened when the crane was loading 1 certain machine. The moon and starts all aligned just right

 

[joseph_e2]

...
As to your example, I have a similar one. I worked in offshore drilling in West Africa. Our winches had proximity switches to assess the position and also to stop it before hitting the derrick. We had two winches, so 6 of these P+F proximity switches installed. One Sunday I get a call that one failed, when I get there I realise it's actually two and by the time I finished the call three had failed. All in, four failed on our ship that Sunday afternoon in sequence.

I didn't have four spares, so installed whatever I had in the warehouse and kept the machines running... come to my office thinking to write to the nearest ship to us (2 years newer) to pinch their stock thinking that since they have redundant proxes, they'll have enough for me to get up and running. Lo and behold, I already had an email from them saying that 8 of their 12 switches failed and they needed my stock to get back up and running.

Worst? Went to the workshop, wired the proximity switches and all of them were working but clearly did not work when in the derrick. We sent all of them to P+F in Germany and got told that these things don't like tropical weather and it's ikely the cause of the failure. Bear in mind that the proximity switches were installed nearly 2 years apart too.

it was one of the most stressful times I had but also pretty cool as I had them replaced without going into downtime.

Similar issue with Festo reed cylinder switches. Work fine for a bit then fail. Worked great on the bench afterward. Maintenance finally figured out (with the OEM in touch with Festo) that they weren't rated for the temperature/humidity inside the ultrasonic washers. We ended up cleaning out all of Festo's North America stock of the more expensive version that could handle the environment.

 

[I_Automation]

I worked on a homebrewed ASRS for stacks of particleboard early in my career (up in Sault Ste. Marie !). One of the complaints I got after we finished commissioning was that the storage cart was "afraid of the new forklifts".

Wait, what ?

"Sometimes when we drive by on one of the new forklifts, the cart just takes off in the other direction. Doesn't happen with the old Hysters."

The system had been built with a long-range retroreflective laser distance sensor to determine the position of the storage cart. For some reason they'd installed the reflectors it used on a wall on the far side of a cargo and personnel aisle.

I had tuned the sensor to ignore no-return conditions where the beam was blocked or absorbed by people or vehicles or spilled stacks of particleboard. I had not tuned it to reject a reflection from the high-viz sticker tape on the new fleet of forklift trucks.

Two photoeye problems I had to deal with:

The first was a conveyor with a photoeye looking up between rollers to sense product, then unload it. Ran for a couple years fine then suddenly one day it's trying to unload non-stop. New photoeyes did the same.
Turns out there was a high-bay light directly over the photoeye that didn't matter until the shop had all the lights replaced with LED lights.
Had to have a canopy made to block the light.


I was called to a different unloader on a spraycoating line with a similar setup that today wouldn't pick up panels at all.
As soon as I got there I immediately saw what was happening.
I asked the engineer standing there if he really went to college, and told the manager standing there that I thought he was better than that.
Turned out that batch of product was being sprayed a flat (very-flat) black that no photoeye would have ever gotten a reflection off of..