Collaborative does not mean fenceless

The most expensive misunderstanding in this industry is that buying a collaborative robot means you have bought your way out of a risk assessment. You have not.

A collaborative robot is an arm with force and power limiting built into it, designed so that contact with a person is detected and the arm stops or yields. That is a genuine and valuable engineering capability, and it is the reason cobots have opened automation to shops that could never have accommodated a caged cell. It is also routinely oversold.

What the technology provides is the possibility of operating near people. Whether your particular cell may do so is determined by a risk assessment that looks at the whole application, and the arm is only one input. Speed matters, because force scales with it. Payload matters. Tooling matters enormously: a force limited arm holding a sharp deburring tool or a forty pound case is a different proposition from the same arm handling a small plastic part, and the assessment will say so.

Geometry matters too, in ways that are easy to miss. A great deal of collaborative risk is about pinch points, meaning what the arm could trap a person against. An arm operating in open space beside an aisle is one situation. The same arm working close to a machine frame, a conveyor rail or a wall is another, because a person's hand can be caught between the arm and the fixed object regardless of how gently the arm is pushing.

The commercial consequence is that a cobot cell can end up needing safety hardware anyway: light curtains, area scanners, speed and separation monitoring that slows the arm when someone approaches. That is a perfectly good outcome and often still cheaper and more compact than full fencing. It is only a problem if the budget assumed none of it.

There is a second consequence that catches people out on throughput. Running collaboratively usually means running slower than the datasheet, sometimes considerably. If the rate you quoted your customer assumed full speed and the assessment requires reduced speed near people, that gap has to come from somewhere. This is why we insist on sizing cells with derates rather than headline numbers.

None of this is an argument against collaborative robots. We recommend them constantly, and for a shop with no floor space and no automation history they are frequently the right answer. It is an argument against treating the word collaborative as a permit. It is a capability, and the assessment still decides.