There was a time when adding a robot to a manufacturing line meant losing floorspace to the robot enclosure and changing the workflow. Today, a new trend is underway – collaborative robotics. Collaborative robotic workcells are designed to permit direct interaction of human beings and robots. In this new paradigm, robots no longer lay claim to large amounts of square footage. The guards frequently come down, leading to significantly smaller installations that are easier to integrate into the production line without disruption. Courtesy of inherent design, functional safety, and a methodical approach to risk management and test, collaborative robotic workcells are easier to add to the factory floor than ever.
Collaborative robots 101
The first thing to understand is that there is no such thing as an out-of-the-box collaborative robot. We can’t talk about a collaborative robot in a vacuum. Instead, we need to think in terms of a collaborative workspace, which is the operating space within which the robot system and human can perform tasks concurrently during production operation. It is the implementation that dictates whether a robot is collaborative, not the robot itself. A manufacturer that advertises a “collaborative” robot has simply given a user/integrator the tools to accomplish this task. It is up to the user/integrator to ensure a safe cell is built.
There are four classes of collaborative operation:
- Safety-rated monitored stop: The robot operates autonomously within collaborative workspace when no people are present. When a person enters the workspace, the robot stops, resuming only when the individual leaves the enclosure.
- Hand guiding: Operator controls robot by hand guiding the robot arm.
- Speed and separation monitoring: System maintains a safe separation between operator and robot at all times.
- Power and force limiting: The system limits potential for injury by constraining power and force of the robot.
The latter class is the one most commonly associated with collaborative robotic workcells by the uninitiated.
Functional safety and collaborative robotics
Even in the best designed system, contact between robot and operator can occur unintentionally, as well as intentionally. Power and force limiting can help protect operators in both cases. Power and force of a robotic system can be limited with two techniques: inherent design and/or use of safety-rated drives and controllers. Inherent design focuses on developing low-inertia robots by applying suitable geometry and materials. Structural members are made lighter. Motors and gearboxes can be selected to operate below some maximum torque level.
Design alone may not be sufficient to support collaborative operation, however. More to the point, design can’t necessarily address changes in applications, use cases, or the regulatory environment. Many automation vendors offer controllers and drives with various safety-rated functions that enable OEMs and integrators to develop safer workcells.
Safety-rated controllers use functional safety to manage operation. Specialty functions include limiting speed, torque, and force. This can be adjusted on an application-by-application basis or even a recipe-by-recipe basis. A safety-rated communications protocol such as PROFIsafe should be used to ensure that the system falls to a safe state if any errors are detected in communications.
Risk assessment and testing
With proper implementation, a force- and power-limited robot can be used for collaborative operation, but that is by no means guaranteed. A force- and power-limited robot with suction cups on the end effector may be collaborative. Swap the suction cups for a knife and the robot is no longer safe. Safety does not rest in the constraints of the axes. To ensure a safe system, risk assessment and testing needs to be performed on each cell.
The processes are covered in detail in EN ISO 12100 (risk assessment) and ISO/TS 15066 & RIA/TR 15.606 (testing). It is worth mentioning a few key points, however. Risk assessment and testing should cover as many contact situations as apply, including the collaborative workspace, access, clearance, ergonomics, use limits, and transitions. It is important to consider whether any stationary structures are involved and other aspects like the end-effector and workpiece. As the knife example shows, not all workpieces or end effectors will be suitable for use in a collaborative environment.
The assessment and test processes need to be strategic. Within the bounds of comprehensive assessment and test, contact scenarios should be consolidated to be fully representative while minimizing the number of test cases. Similar cells may be able to have a limited set of validation testing if the risk assessment shows that there are few or no changes that affect safety.
When the collaborative workcell is validated is just as important as how. It is vital to consider the requirements discussed above during the design of the workcell and not after it is built and installed. Failure to plan ahead can lead to unpleasant surprises that cause delays and other issues. Each new collaborative cell should be verified before operation. Changes to the cell such as workpiece, operator station, or programming of the robot will likely require retesting.
Developing a collaborative workcell begins with understanding the technology and how it can improve the process. Although manufacturers commonly want “guard-less” robotics when they are considering a collaborative workcell, that is not always possible. In more than a few cases, the manufacturer doesn’t particularly want or need a collaborative workcell, they just seek to have a guard-less system. In these cases, it’s better to identify the pain points and try to find a more effective way to address them. For the right application and circumstances, however, a collaborative workcell can increase productivity, decrease worker injury, and streamline the production floor for improved overall operations.Have an Inquiry for Siemens about this article? Click Here >>