SteelCollar Robotics
Home / Notes from Joe / Why we will not tell you a cell is safe

Why we will not tell you a cell is safe

We get asked to confirm that a proposed setup is safe, usually by someone who just wants to move forward. We will not do it, and the reason is worth understanding rather than working around.

Safety in an industrial robot cell is not a property of the robot. It is a property of the whole installation: the machine, the tooling, the part, the speed, the layout, the traffic pattern around it, the people who walk past, and what happens in every failure mode. Two plants can install identical arms doing identical work and reach different conclusions, correctly, because the surrounding conditions differ.

That determination has a name and an owner. It is a risk assessment, it is performed by a certified integrator on your actual floor, and it is governed by standards including ANSI and A3 R15.06 and ISO 10218. It is not paperwork you obtain afterward to satisfy an insurer. It is the process that decides what the cell needs in order to be installed at all: guarding, light curtains, speed and separation monitoring, or in some cases a conclusion that the layout you had in mind does not work.

We are a dealer. We sell equipment, we introduce integrators, and we never take the integrator of record role, which means we are never the party performing that assessment. If we told you a configuration was safe, we would be issuing an opinion we are not qualified to issue and not accountable for, and you would be relying on it. That is a bad trade for you even though it feels helpful in the moment.

This cuts both ways, and the second direction is the one people find more surprising. We will not tell you a cell needs a fence either. Over guarding is a real and expensive failure, and we have seen shops spend money on fencing that a proper assessment would have shown was unnecessary. Guessing high is not the safe answer. It is just a different wrong answer with a bigger invoice.

The place this comes up most is collaborative robots, because the word collaborative is widely misunderstood to mean fenceless. A force limiting arm can often work near people, and frequently that is exactly why it is chosen, but whether it can in your cell depends on speed, payload, tooling and what the arm might pinch a person against. A cobot holding a sharp tool or a heavy case is a different assessment from the same cobot handling small light parts.

So when you ask us, the honest answer is that we can tell you what class of solution typically fits, what usually gets guarded and what usually does not, and what questions to put to your integrator. What we cannot do is sign off on it. When someone in this industry does sign off casually, the right response is to wonder what else they are casual about.

Safety is determined by a certified integrator's risk assessment on your floor, not by a dealer, and not by the robot's category. Over guarding is a wrong answer too.

Put it to the test on your own job

Tell Joe what you are running and he will apply this to your numbers, including telling you when a robot is the wrong answer.

More notes