Collaborative robots: safety before productivity
What a manufacturing manager needs to understand before cobots start working alongside people on the line.
A few years ago, talking about robots in manufacturing meant heavy industrial robots behind fencing - a human-free zone, high speed, high payload. That works well for automotive assembly lines with large volumes and standardised operations.
But most manufacturing businesses are not automotive plants. For flexible production with frequent changeovers, short runs, and operations that are hard to fully automate, an isolated industrial robot does not fit. This is exactly the situation that collaborative robots - cobots - were designed for.
Cobots are built to work alongside people - without a cage, without permanent separation. That expands the range of applications. But it also creates risks that need to be understood before a cobot arrives on the line.
How a cobot differs from an industrial robot
The main technical difference is in force and speed limiting systems. Cobots have sensors that detect contact with a person and stop or slow the movement. They run slower and with lower payloads than industrial robots.
That does not mean cobots are safe by themselves. It means that safe collaboration with people is possible - when the application is properly designed. "When properly designed" is the key condition.
Where risks arise
Risk assessment for cobots is more complex than for isolated robots. In an isolated zone, people do not work - the risk of contact is minimal. In a shared zone, people do work, and every operation must be analysed for possible contact.
Typical sources of risk:
Tooling and end-effectors. A cobot itself may be designed safely, but if it is holding a sharp tool or a clamp, contact is dangerous regardless of how gently the robot stopped.
Speed and payload for specific tasks. Standard safety parameters are designed for typical operations. If a cobot is reconfigured for a non-standard task with higher payload or speed, the assessment must be revisited.
Changes to the workflow. If something begins happening near a cobot that was not part of the original scenario, new risks arise. People adapt and improvise; automation does not.
What to do before launch
Risk assessment for the specific operation - not a general assessment of the cobot as a product, but of what exactly it will do, with which tool, alongside which people. This requires participation from an occupational safety specialist.
Staff training. People who will work alongside a cobot need to understand its capabilities and limits: that it stops on contact but not instantaneously; that certain actions near a cobot are unsafe; and how to stop a cobot when needed.
A change process. When an operation or a tool changes, a repeat risk assessment is mandatory. This should be part of the change management process, not something remembered afterwards.
A practical question
Before deciding to deploy cobots, it is worth asking: do we have the competence to properly assess the risks of collaborative work - or are we assuming that the product's safety certification means everything is fine?
A product's safety certification and the safety of a specific application are two different things. Responsibility for the second belongs to the company.