Cobot safety: the four types of collaborative operation

Universal Robots A/Svisit website


Collaborative robots have become the buzzword in industrial automation. But how do you technically define collaborative operation?

Cobot safety: the four types of collaborative operationLast year, Universal Robots welcomed the new ISO/TS 15066 guidelines on safety in collaborative robots systems that lists four types of collaborative operation:

  • Safety-rated monitored stop
  • Hand guiding
  • Speed and separation monitoring
  • Power and force limiting

When it comes to the safety-rated monitored stop, the robot system stops before the human operator can access or be exposed to any hazard in the collaborative workspace. Only when there is no human operator can the robot move as a non-collaborative robot. In other words, either the robot system or the human operator moves, not both at the same time. This method cannot utilise the advantages of collaborative operation and it requires safeguarding of a traditional industrial robot. The benefit that this offers is the ease and speed to resume automatic operation.

With respect to the hand-guiding operation, the human operator uses a hand-operated device and the robot system moves based on motion commands of the operator. It is a kind of manually controlled operation in that the operator is in direct control of the robot system's operation. This is considered automatic operation, not manual operation.

In reality, only the third and fourth options are used for beneficial collaborative applications in industry.

Currently, most collaborative robots including Universal Robots are inherently designed to limit power and force. If the robot detects a certain level of power or force, the robot controller stops the robot system to protect the human operator. Due to the safety functions for motion, speed, force and power monitoring, the human and robot system can move at the same time in the same workspace. As long as the risk assessment is conducted properly, traditional guards and protective devices are not needed.

The power and force limiting method assumes that the human can contact the moving robot system. Hence, we should consider the impact to human body during the risk assessment process. To prevent pain or injury, the application restricts payload and speed. As a result, the robot speed will likely be too low to be useful for high-risk applications.

In order to use the speed and separation monitoring method, external safety devices such as a safety scanner have been used to lower speed as a person approaches the collaborative workspace.

These days, the research on additional technologies for sensing humans is being conducted. For example, a capacitive sensor could be used to measure the distance between the human and robot system. These new technologies have yet to be commercialised for safety functions. Regardless of technology, safety sensing needs to meet the required functional safety and the application be verified and validated.

If the sensing technology to detect humans is successfully developed and certified as a safety-rated product, it is possible that the safety device would be integrated with the robot as a common feature.

In the future, the speed and separation monitoring operation can be deployed in the real industry upon the development of advanced safety sensors. The robot could slow the system's speed if the human operator comes closer to the robot. If the human gets too close to the robot system, the protective stop would be triggered. The robot could move faster and the payload could be heavier because the robot system will not contact the human operator.

For more efficient operation, a dynamic safety system configurator can be used. The safety setting changes corresponding to the distance between the human and robot. Furthermore, the robot can change the path of motion planning based on human movement. Then, the robot system does not have to stop even if the human operator enters the robot moving area.

Go to for more information about collaborative robots.

© Copyright 2006-14 Damte Ltd.