Life in a vigh volume manufacturing environment

Have been assigned the care of this barcode scanning system that worked externally from the TEL Tracks. This system has been designed by this electronics prodigy a few years back, and recently moved on to another company.

Now manufacturing have been highlighting the case of dysfunctional barcode scanning for their daily chemical use... Since June, I had to learn the stuff, the wiring, the functionals, and the what is's and what isn't's of the system. No schematic, and the last person who had a hand at this system proly said good riddence to it when I was assigned to this. Anyways, I have spent days and days figuring up this system... and ways to make it better. Because my boss ain't hearing this = my description of this system is "a few compenents soldered with shit!".

Yeah, it is frustrating when every month your monthly report says "I.O.U" because you simply have no time to look at all these things. Between the three of us track owners, we are up against over 40 machines that may require our attention 24/7, and there are times when we work 24 hours straight on. We also have projects under our care, on top of other daily responsibilities such as quality feedback from KLA group (DO issues), championing of certain systems and key performance indices (I take on MTBI), cost reduction activities (second sourcing repairing is hot in the industry now) and various other logistics command and controls implementation. There simply just isn't time to do a lot of monkey work around.

Then there is the human factor to come into play. High turnover industries such as semiconductors have a predominantly high worker turnover rate, depending on the economic cycle. When people leave a company, they take along with them the skills and experience that have been nurtured thus far. If a major movement of people occur in a specific division, then there is bound to be a vacuum of skills, and those experienced ones who are left, have to take on extra burden Replacements that come will take time learning and assimilating to a new system and culture. It is a fact that training takes a long time in semiconductor manufacturing, and the effect of this void will immediately hit the uptime (equipment) and utilisation (manufacturing). Even process engineering group will suffer from wrong disposition and human errors.

Anytime anyone out there thinks your job is tough, think again. Think when everyday when you plan to sit down to do something in your cubicle, there are issues and your presence is needed in the production floor (fab). Don't think you are engineer you are excused from doing the dirty jobs. I change valves, motors, drives, I sniff carcinogenic chemicals, crawl over and under multimillion dollar machines. I analyse problems ranging from electrical connection failure, to encoder problems, and intense troubleshooting. Our set of machine manuals, when arranged side by side, are around 3 meters thick! These do not test my knowledge of the Fourier Series, or, even my B and D matrices in composites. These test how much I can troubleshoot a problem, starting from gathering of information (usually very sketchy), assess the problem, choose and discuss the best possible solution, assess the potential risks to both the machines and the product that will run after, and make it happen. All this have to happen in the fastest time possible, or, you will definitely get it in tomorrow's morning meeting. A lot of times there will involve external parties such as the KLA (Defect Density people a.k.a. "the fab police"), process engineering, engineers from other modules (Etch, Diffussion, CMP etc.), manufacturing engineers etc. Within, you will have to control your pool of Associate Engineers (AE's) so that they remain vigilant throughout the intense 12 hour shift.

101 things are happening everywhere at the same time, you bet at least 5 have your name in it at any given time.

"Where do I put my attention then?" Then I was thought to ask myself further:

"By not doing which one now, will I be fucked for tomorrow?" and then the answer immediately becomes clear. Same goes with clearing your mail. Always ask "By not reading which one will I be screwed by someone next?" Coz we have heaps of mail to read every morning.... from various departments/sections..

My boss hardly sleeps these days. Because being the most highly ramped fab in the company, we are busy. And as if that is not enough, all fabs are designed to have the lithography (masking) to be bottlenecks. What does it mean? It means people working in lithography are fucked twice over. Every second missed, the entire fab misses 1 second. Shipment will be delayed and this will warrant the attention of the Fab VP, and in worse cases, our CEO. But who can guarantee 300-400 machines, 900 people, not to go wrong or make mistakes? Most machines therein are designed to last for 5 years, and they are already running 10! Financially, they have been fully paid up years before this (talking a billion dollar worth of capital assets) and on finance papers, they have $1 value, but they are capable, in the example of the TEL track and ASML stepper machine cluster, in producing a throughput of 1000 wafers per day.

So, my boss asked the line AE shift leader to give him an update every 3 hours 24/7. Still, he is expected to come to work on time at 8 a.m. till 8 p.m. most of the time! Saturdays he comes back for Operations meetings. In fact, the Fab VP, a Malaysian Indian, cancelled his Deepavali leave yesterday to come back to meet up with his 40 over directors, managers and section managers, and no one is surprised..... Coz I bet you he asked himself this question: "By not doing which one now, will I be fucked for tomorrow?" while thinking of his boss, the Senior VP of Operations'look.

Such is the life working in the semiconductor business. They pay relatively ok, but they throw you in a pool of aligators. You either swim like you never did before, or you die. Then we have this freakking barcode scanning system that I hate to further troubleshoot... because it is plagued with bad design for serviceability. At one glance, it does resemble some ITE (of TAFE) student's final year project that was poorly documented. Last night would be my umpteenth time overtiming without pay to troubleshoot this system. I left the company at 5 a.m. (after working for 21 hours) and that was how I spent my Deepavali wee hours - cursing on a poorly designed, fragile, yet critical system that I have to carry on however I hate it.

I think I have better things to do.