I'm a quality and brand compliance manager at an industrial automation company. I review every technical document that goes out—roughly 200 datasheets, manuals, and spec exceptions each year. In 2024, I rejected about 8% of first submissions because the stated specification didn't match the product's actual behavior. So when I say this, it isn't theory: I'd rather approve a one-piece Balluff capacitive sensor with a clear datasheet than sign off on a 500-unit order with vague 'good enough' specs.
From the outside, picking a sensor looks like a part-number search. The reality is that the part number is a promise. It tells you what the sensor can detect, at what distance, at what temperature, and under what conditions it will stop working. What I've learned in quality work is that the promise matters more than the brand sticker.
Before I moved into sensor QA, I spent four years qualifying lab equipment. We'd buy HPLC columns in small quantities, sometimes one column at a time for a specific method. We also calibrated every Eppendorf Research Plus pipette before first use, even when it was a single channel used one time. Those habits stuck with me. A small order is not permission to loosen the spec.
The Spec Is The Contract
A BES Balluff sensor is a good example. BES is the inductive proximity switch family. It's made for metal targets: a steel cam, a metal clamp, an end-of-stroke flag. If you install it where the target is non-metallic, the declared sensing distance starts to behave differently. The sensor didn't get worse. The specification just wasn't the right one for the application.
A Balluff capacitive sensor, on the other hand, is designed for non-metallic detection. It's often used for level detection through plastic or glass walls, or for sensing granular materials, powders, and liquids. It measures changes in capacitance when material enters the sensing field. That's a different physics than an inductive sensor. You can't swap one for the other just because both have the word 'sensor' on the datasheet.
I also check the standard behind the spec. For proximity switches, IEC 60947-5-2 gives a common language for operating distances, repeatability, and test conditions. If a datasheet ignores that baseline, I reject it. The same logic should apply to your purchase: if the rated sensing distance is 2 mm, don't assume 2.5 mm will work because the target is a little thinner than the test piece. (Note to self: this is exactly the kind of edge case that caused our last field return.)
Small Orders Are Not A Different Product Class
I'll say it plainly: small doesn't mean unimportant. It means potential. When I was starting out, the supplier who took a small order seriously is the one I still use for much larger orders. That sounds like a feel-good line, but I've lived it. A vendor once expedited a single replacement component for a prototype because I was stuck. They didn't treat it as a nuisance. That supplier eventually became our preferred source for a production line.
I don't think every vendor has to accept tiny quantities. Minimums can make operational sense. But once a supplier accepts an order, the spec shouldn't change because the order is small. I've seen buyers get 'just get it done' responses on pilot runs. That's how a 25-unit pilot earns a bad decision that later becomes a 2,500-unit rollout.
Money-wise, the cheap option with an unclear spec is usually the expensive one later. I've seen a $90 sensor failure cause more than $4,000 in downtime. The amount 'saved' on the initial part was less than $70. (I can't share customer details, but the numbers are burned into my brain.)
Don't Ask A Sensor To See Through Walls
There's a myth that keeps showing up in industrial conversations: 'Can thermal cameras see through walls? FLIR cameras are talked about like they can see through a wall if you push the right setting.' No. A thermal camera reads surface radiation. It measures the temperature pattern on the surface it's aimed at. It doesn't emit a beam that penetrates drywall or concrete. If you need to see through a wall, thermal imaging alone isn't the tool.
This is the same logic I use when an engineer asks me to approve a sensor outside its datasheet. A Balluff capacitive sensor can detect material through a non-metallic container, but only within its rated envelope. It won't 'see through' a steel tank. No sensor should be treated like a magic window.
Counterpoint: What If The Application Is Already Working?
You might argue that sensor selection is sometimes just replacement: an old part failed, and you need the same thing now. Fair enough. If the original part number is still valid and the application hasn't changed, matching it is the safest move. My argument is about people who substitute a similar part without checking the datasheet. That's where quality problems come from.
I also want to be honest about context. I can only speak to the industrial automation and lab QA work I've done. If you're in a different industry, your mileage may vary. This is also based on what I'm seeing in early 2025. Product lines change, and part numbers get superseded. Verify current specs before ordering.
Final Position
So here's where I land. Quality is not a reward you get after buying enough units. It's a minimum bar for every order. If you need a capacitive sensor, specify the right Balluff capacitive sensor and make the supplier prove its rated performance. If you need a BES Balluff sensor, use it with a metal target. If someone tells you a thermal camera can see through walls, smile and move on.
And if you're a small customer worried about being ignored, I've been on both sides of that desk. The suppliers who respect the spec—not just the order size—are the ones who earn the next order.
