3 ms·
I used to do plant floor support in an automotive assembly plant. Think desktop support, but super extra. Think multiple serial single points of failure, on a o
by csours 7y ago
I used to do plant floor support in an automotive assembly plant. Think desktop support, but super extra. Think multiple serial single points of failure, on a one minute metronome. Here are some things I've screwed up.
---
A simple, early one: We used VNC for remote desktop support of line-side production computers. One of my team leads was walking me through what was on the screen and what was going on. I was used to right clicking on these screens to see more of what was going on, but this one happened to be running a script that was interrupted by this. When I right clicked, my team lead freaked out, and the operator on the floor freaked out, and started moving the mouse themselves and clicking everywhere. After a while they started just doing their job again, but shortly after we got a call from a supervisor.
---
I got a call saying tracking was off on a production conveyor. This means that operators were getting incorrect instructions and work was being recorded on incorrect units. I adjusted tracking to match what I was being told. All good.
I shortly got a call from the same conveyor saying tracking was off. I told them "Yes, I just fixed it". "Well it is wrong now, it was fine a minute ago". So I adjusted it to match what I was being told now. Who knows what that other guy was smoking.
Right as I finished re-adjusting tracking I got another frantic, high energy, expletive filled call saying the tracking was off.
Dear reader you may have guessed what was wrong.
Since I got multiple sets of contradictory information, I decide to go out to the floor. This is what I see (simplified):
Footprint 1,2,3,4, etc
FP1 FP2 FP3 FP4 FP5 FP6 FP7 FP8
008 007 006 ___ 005 004 003 002
You see, the first person was on the first half of the conveyor, and the second person was on the other half. They were both correct, but neither had the full story. There was an empty carrier in the middle of the conveyor.
---
Last one. One of our weekly ops tasks was to verify that the 3 (three!!) scheduling services agreed with one another and also the production schedule. Unfortunately sometimes we got late breaking schedule changes, like running extra time or extending or moving lunch. On a Friday night/Saturday Morning, I got one such change. We were going to run an hour extra to make up for earlier lost units.
I made the requested change and went back to "compiling code" etc. (Perks of night shift)
Some time later... I get a call on one of my radios (Nextel) saying the lights were off in the back of the shop. I say, "hmm, that's odd" and go to the screen to turn the lights on in that area. I get a call on my other radio, saying the lights were off in the middle of the shop. Oh sh*t. For context, the lights were now off for over 200 pissed off people who just wanted to finish their overnight shift and go home. I continue to press buttons to turn lights on, hindered by the fact that the lighting controllers were on a very very slow daisy chained serial bus. My radios continue to go off with people urgently and excitedly informing me that the [expletive deleted] lights were off. I also got a visit from the Plant Shift Lead (2 or 3 steps down from the plant manager). I was pretty surprised to see her, as I was kind of wedged in a corner with a bookcase blocking half of the entryway to my cubicle.
Anyway, I eventually got the lights turned back on. Looking at the schedule changelog, I had successfully extended the shift, but for the wrong day. I had done it for Saturday, as the clock was past midnight when I edited the schedule. Oops.
---
These were all relatively early in my career, but I think they're pretty colorful.