What we are for

We sell industrial robots to small and midsized manufacturers. This page is what we believe about that, including the part of it most companies in our position would rather not put in writing.

We are a dealer, and we say so on every page

Steel Collar Robotics is an independent, multi brand robotics dealer. We earn a margin when you buy equipment through us, and we have referral relationships with the certified integrators we introduce. That is the whole model. You will find it in the footer of every page on this site, because a customer who has to work out how we get paid is a customer who will eventually wonder what else we did not mention.

We carry multiple brands deliberately. A dealer who represents one manufacturer will always find that one manufacturer's arm to be the answer. Because we carry many, our answer can be the one that actually fits your job, including the answer that none of them do.

The honest read is the product

Anyone can build a website that says yes. The reason ours exists is that most of the jobs people ask us about should not be automated yet, and almost nobody in this industry will tell them so.

So we built the opposite. Our concierge is instructed to say when a robot is the wrong answer: when the task runs under about two hours a day, when the real bottleneck is upstream of the station you asked about, when a three thousand dollar conveyor or a fixture or a process change would solve it for a fraction of the cost. Our process library has pages that conclude a dedicated denester beats a robot, that a print and apply labeler beats a robot, that a wide belt sander beats a robot. Those pages cost us sales. They are the point.

We do this because a robot is bought with real money by people who do not get many chances to be wrong. The dealer who talks you into a bad first project does not get a second one, and deserves not to.

We are never the integrator of record

Cell design, guarding, risk assessment, commissioning and safety validation are performed by independent certified integrators, in accordance with ANSI and A3 R15.06 and ISO 10218. Not by us. That is a contractual boundary, not a style choice, and it means we will never tell you a configuration is safe. Only an assessment on your actual floor can determine that.

It also means we will not tell you a cell needs a fence. Over guarding is a real and expensive failure, and guessing high is not the safe answer, it is just a different wrong answer with a bigger invoice.

Now the part about people

Every automation company eventually has to answer whether it is putting people out of work. Most answer with a slogan. Here is what we actually see, and where we think the interesting opportunity is.

Nearly every shop that comes to us is not trying to shed labor. They cannot find it. Read our own casebook, which is sourced and linked rather than asserted: a Washington machine shop won a round the clock medical order and could not hire a third shift. A Tacoma aerospace supplier was looking for as many as a hundred people who simply were not available in its labor market. An Ohio fabricator could not hire enough welders to keep up with demand. A small fabricator that grew out of a garage was turning down work because it could not train welders fast enough.

That is the actual shape of the problem. The jobs going unfilled are the ones nobody wants: the repetitive lift, the dusty grind, the second shift on a machine that just waits. When those jobs go unfilled, the plant does not modernize, it simply turns down orders. And when it turns down orders, everyone there has a smaller future, not just the person who would have taken the open req.

The person at that station is your best robot technician

Here is the thesis we care most about, and the one we think the industry has almost entirely missed.

When a station gets automated, the natural instinct is to look outside for someone technical to run the cell. We think that is backwards. The person who has been doing that job knows things no new hire will learn for a year. They know which cartons arrive damaged and how often. They know the sound the machine makes before it faults. They know that the label has to face out for one particular customer, that the pallet leans if you build the third course wrong, that the material from one supplier runs differently from another. That knowledge is the exact input a working cell needs, and it does not exist on paper anywhere in your building.

They also have something harder to hire for than technical skill: they already show up, you already trust them, and they already know your plant. Robot operation is learnable. Judgment about your process is not, at least not quickly.

Our own casebook has this pattern in it, in the accounts that publish enough detail to see it. A Virginia fabricator that put a grinding robot in moved its grinders into robot operating roles rather than letting them go. An Illinois machine builder states the cell shifted labor rather than cutting jobs and that headcount was not reduced. A Quebec manufacturer that automated a press brake unloading station states no layoffs resulted. None of these are our claims. They are published statements from the companies, linked from the entries, and you can read them yourself.

We want to be careful here, because this is where a manifesto usually starts lying. Automation does not guarantee this outcome. It removes tasks, and what happens to the people is a decision the owner makes, not a property of the technology. What we are saying is narrower and, we think, more useful: when owners do choose redeployment, the evidence suggests it works, and the person already standing at that station is usually the strongest candidate rather than the obvious casualty.

What that looks like in practice

The roles are real and they are not exotic. A cell needs someone to load and change over, to clear faults, to judge whether a finish or a weld is acceptable, to handle the exceptions the robot cannot, and eventually to teach it a new part. Those are day one, week one and month six versions of the same job, and they are a ladder rather than a single step.

They also pay better than the station they replaced, which is the part that matters to the person doing it. A robot operator is worth more to a plant than a manual grinder because their judgment now multiplies across every part the cell touches instead of every part their own hands touch.

The training is more accessible than most owners assume. Robot makers run operator courses. Integrators train on the specific cell as part of commissioning, and that training is negotiable, so ask for more of it than the quote includes. Community colleges and manufacturing extension partnerships run programs, often subsidized. And the modern arms are genuinely easier: hand guiding, palletizing wizards and graphical teaching have moved a lot of what used to require a programmer into what a capable operator can do in an afternoon.

Our advice to owners is unglamorous. Decide who will own the cell before you buy it, not after. Name the person. Send them to the training. Have them in the room during commissioning, asking questions, because that week is the cheapest education they will ever get and most shops waste it.

Where this goes

The wider version of this argument is that the people who have spent years doing physical work in manufacturing are among the better prepared, not the worse prepared, for a workplace where machines do more of the doing. They already know how a process actually behaves as opposed to how it is documented. They already know that the exception is where the cost lives. They already have the instinct that something is about to go wrong before the data says so. Those are the skills that matter when your job shifts from performing a task to supervising and improving a system that performs it.

We are not going to pretend that transition is automatic or painless, or that every plant will handle it well. But we sell into an industry where the constraint is people who are not there, not people with nothing to do. If the machines take the jobs nobody would take, and the people who took them anyway move up into running the machines, that is a good trade. It is also, on the evidence we have gathered, what is actually happening more often than not.

We would rather be the dealer that says this out loud and gets held to it.

What we will not do

We will not tell you a configuration is safe. We will not give you a fixed price or a firm payback before an engineer has walked your floor, because every number produced before that is arithmetic rather than a commitment. We will not advise you on layoffs, employment law, financing or taxes, all of which need people qualified in those things. We will not take investment from a robot manufacturer's venture arm, because neutrality between brands is the entire product. And we will not sell you a machine we cannot obtain, which is why the humanoids we track are marked as not orderable rather than dangled as an option.

If any of that changes, this page changes with it, and you should hold us to what it says.