A few months ago, I was reviewing a bill of materials for a mid-sized packaging line upgrade. We had specified a Siemens S7-1500 controller, a pair of ET200SP remote I/O stations, and a new TIA Portal engineering station. It looked fine on paper. But as I dug into the component list, something felt off.
I'm a quality compliance manager at an automation integrator. I review roughly 250 unique deliverables a year—BOMs, wiring diagrams, software archives, test protocols. In Q1 2024 alone, I rejected about 22% of first-round submissions. Not because teams are sloppy, but because details get missed when people are under deadline pressure.
This particular job had a tight schedule—eight weeks from kickoff to factory acceptance test. The project manager was pushing to order hardware. The engineering team was confident. But my job isn't to be confident. My job is to verify.
So I started tracing the CPU selection back to the spec. The customer had asked for a CPU that could handle 40% spare capacity for future expansion. The team picked an S7-1511-1 PN. It's a capable CPU, but its maximum code block size is 150 KB. When I checked the estimated memory utilization from their earlier tests, they were already at 110 KB after the basic logic. With HMI communication and data logging, they'd hit the ceiling within a year.
(Note to self: never assume the default choice is adequate.)
"The spec said 'spare capacity,' but the budget said 'keep it simple.' The team satisfied one requirement at the expense of the other."
I flagged it. The engineer was frustrated—thought I was being pedantic. But I called a quick meeting, pulled up the sizing guidelines from Siemens' support portal (Siemens Industry Online Support, document ID 109478140, April 2024 revision), and walked through the math. The team relented. We upgraded to an S7-1515-2 PN. The cost delta was about $420. On a $78,000 hardware package, that's 0.5%. The alternative: a $22,000 field upgrade later, plus production downtime.
Here's where it gets interesting. While reviewing the same BOM, I spotted a line item for contactors. It listed a generic "3-pole contactor, 25A." No part number. No manufacturer. Just a description.
Now, in the world of industrial control, types of contactor matter a lot more than most people realize. A standard AC-3 rated contactor designed for motor starting behaves differently from an AC-1 rated contactor for resistive loads. Using the wrong category can lead to premature welding of contacts, arcing failures, and—worst case—a downstream machine fault that shuts down a line.
I flagged this too. The purchasing agent pushed back: "It's just a generic part. We'll grab whatever is in stock."
I didn't back down. I asked the engineering lead to specify the exact application: three-phase induction motor, 5.5 kW, across-the-line starting. That meant we needed a contactor rated for AC-3 utilization at 25A minimum, with a mechanical life of at least 10 million operations. We ended up specifying a Siemens 3RT2024-1BB44. It was a 30-minute change on paper but prevented what could have been a catastrophic field failure.
(The surprise wasn't the pushback. It was how many people in the room didn't know the difference between AC-1 and AC-3 ratings. I'm not a power engineer, so I can't speak to every nuance of contactor design, but from a quality perspective: if the spec doesn't say what you're switching, you haven't specified the part.)
The client also asked about general maintenance items. Someone on the thread mentioned a "toro 20339 spark plug replacement." It was a total non-sequitur—someone on the client side was mixing up their equipment logs. I'm not a small engine mechanic, so I can't speak to Toro lawn mower maintenance, but I can tell you this: the moment a conversation jumps from PLC memory sizing to spark plugs, you're losing focus.
I recommended they handle that request separately with a local small engine service. It felt a bit dismissive in the moment, but in my experience, scope creep isn't always price creep—it's attention creep. An informed customer asks better questions and stays on topic. This one was just a bit disorganized. We gently steered them back to the automation scope, and they appreciated the clarity.
Later that same week, the client asked "where to buy a battery charger" for their instrument technician's kit. I had to laugh (internally). We're designing a $250,000 control system, and suddenly we're sourcing personal tools.
I get it—project managers wear too many hats. But here's a real lesson: don't let the tail wag the dog. If you're buying a battery charger for your own toolbox, check Amazon or your local tool distributor. But if you're specifying a UPS or a battery-backed power supply for a Siemens PLC cabinet, that's a different conversation entirely—and one that belongs in the electrical design review.
My team helped them find a 24VDC UPS (Siemens SITOP series, 10A) for the PLC rack, and separately, I pointed them to a tool distributor for their personal charger. Two different problems, two different solutions. The person asking didn't know the difference; I did. That's the value of customer education.
Looking back, here's my takeaway from that project:
That project delivered on time and passed factory acceptance with zero non-conformances. The upgraded CPU was praised during commissioning—the customer added two more I/O modules after go-live and didn't hit capacity. The contactor spec held up through the first 100,000 cycles. And I learned that sometimes the most valuable thing you can do is ask one more question before signing off.
Take it from someone who rejected 22% of first-round deliverables last year: the pain of a verified spec is far less than the pain of a field failure.