Power & Force Limited or Speed & Separation Monitoring?

By Alberto Moel (Vice President Strategy and Partnerships)

Welcome back, dear reader, to one more installment of the Veo Robotics blog. Over the last two years (has it been that long? Yikes!), we have written about how flexibility in manufacturing is increasing in importance, how humans and machines working collaboratively increases flexibility, and how this flexibility has value. We have backed up our assertions with numerous case studies showing the economic value of human-robot collaboration

In recent postings we have been discussing the particulars of human-robot collaboration as defined by the ISO safety standards for robots in industrial environments, particularly Speed and Separation Monitoring (SSM). We studied how guarding and fencing can (paradoxically) improve human-robot collaboration and reduce workcell area; how robot reach, payload, and speed affect how “collaborative” the robot can be; and how robot dynamic and controller characteristics enhance or limit its collaborative capabilities.  

As our readers well know, some applications are quite a good fit for human-robot collaboration, such as palletizing, machine tending, and in-cycle assembly. But the workcell geometry, what kind of robot used, the workpiece, and the process steps can all vary from workcell to workcell, which will determine how the human and robot will collaborate. It could be that the application requires the simplest form of collaboration, where the robot and human are in proximity to each other but do not really interact with each other during robot operation. In today’s post, we provide some thoughts on the best collaboration mode, using either a PFL robot or SSM with Veo FreeMove® while keeping within the ISO 10218 and ISO/TS 15066 standards.

We begin by asking whether a Power and Force Limited (PFL) robot is adequate for your application. If not, the key questions that would determine if SSM with Veo FreeMove works for you depend on how far the human can be from the robot while the robot is moving, whether and how fast the robot needs to move while the human is near, and how big of a safeguarded area is required. We will numerically show that there is a broad space, the “Protective Separation Distance (PSD) Envelope”, bounded by minimum and maximum workcell areas as a function of robot speed and FreeMove sensor coverage limits) within which SSM human-robot collaboration is feasible.

If you are in a rush, Figure 1 shows the high-level decisions that need to be taken, which we believe are (relatively) self-explanatory. If you have a few minutes and want to learn about the details, please read on!

Figure 1. A Decision Tree when considering Speed and Separation Monitoring collaborative applications with Veo FreeMove®

Figure 1. A Decision Tree when considering Speed and Separation Monitoring collaborative applications with Veo FreeMove®

Power and Force Limited or Speed and Separation Monitoring?

The first decision point is whether your collaborative application can be supported using existing Power and Force Limited (PFL) robots. But because we need to consider the entire application, including a risk assessment, we have to bear in mind constraints on performance and safety. Very simply, and as shown in Box 1 of Figure 1, your application can be straightforwardly implemented using PFL robots if you can answer “yes” to all the following questions:

  • Does the robot payload (including End of Arm Tooling) weigh less than about 15kg?  Most PFL robots are limited to small payloads if they are to remain safe upon human contact.

  • Is the maximum Tool Center Point (TCP) speed required for your application slower than about 1.5m/sec? Because PFL robots must be safe on human contact, they are speed limited. 

  • Is the required robot reach 1m or below? To remain safe, PFL robot moments are limited by keeping reach (and inertial forces at the TCP) to a narrower range.

  • Is human-robot contact while the robot is moving OK? In other words, will the robot making contact with a human not cause injury? This contact can be accidental or part of the process, where the human operator stops the robot on purpose by touching it. 

  • Will the application pass a risk assessment as outlined in ISO 10218? This is crucial to make sure the application as designed is safe for human-robot collaboration. 

If you can answer these questions in the affirmative, you are good to go, your application is suitable for PFL human-robot collaboration. However, if your application does not meet these criteria, it is time to consider SSM approaches, ideally using Veo FreeMove®.

Speed and Separation Monitoring with Veo FreeMove®

The key question whether your application is amenable to SSM is determining how far the human can be from the robot while it is moving, the Minimum Distance to Hazard (MDH).[1] In this case, the hazard is the robot, and a human must be far enough away while the robot is operating so that if the human were to approach the robot, it could stop safely before the human and the robot came into contact. This approach can be part of the application (where for example, the human performs a step on a workpiece carried by the robot) or unexpected (where the human and robot are near each other but do not interact during normal operation). Hence, there is a minimum distance (the MDH) at which the human must be while the robot is operating.

The MDH depends on robot speed, payload, pose, robot controller latency, and, of course, the dynamics and mechanics of the specific robot. Our previous work indicates that this MDH is at least 1-1.5m, for a wide range of robots, even at zero speed, as the MDH needs to include robot controller latency.[2] If the application incorporates such a distance, and the human can be at this MDH while the robot operates, we can proceed (Box 2, Figure 1). 

If the MDH condition cannot be met, perhaps some creative rethinking of the application is needed (Box 3, Figure 1). Maybe having the robot move at high speed or with the heavy payload away from the human or reconfiguring the workflow will allow enough distance between the robot and human. However, it may still be the case that the robot and the human have to be closer than the MDH while the robot is operating (perhaps from space constraints). In that case, it is bad news; the application cannot realistically be made collaborative, and physical fencing or guarding to keep the human and robot apart is likely necessary.

We now come to the last decision point (Box 4, Figure 1), which involves what we call the Protective Separation Distance (PSD) Envelope, which is the “space” of SSM operation “bounded” by the robot speed (or payload or extension) and workcell coverage (which for Veo FreeMove is roughly 64m2). The concept is quite simple: as the robot speed (or payload or extension) increases, the MDH will also increase, and hence the workcell area and required safeguarded space needs to increase. As the workcell area increases, at some point it may be bigger than the FreeMove workcell coverage area.

To keep it descriptively simple, we model the MDH as a function of speed (and with realistic parameters) as outlined in our previous work. A visual representation is in Figure 2. The bottom upward-sloping line is the minimum workcell area as robot TCP speed increases.[3] For sufficiently fast robot speeds, the required workcell area is larger than 64m2 and hence, if the application requires such speeds, it is outside the PSD Envelope (indicated by the shaded green area).[4] However, if the proposed application meets the PSD Envelope requirement it is a great candidate to use SSM with Veo FreeMove.

If the application does not fall within the PSD Envelope, could the application be reconfigured to meet the PSD Envelope? In the case of Figure 2, we were quite “lenient” by safeguarding an unfenced robot, requiring an MDH all around the full robot envelope, and a correspondingly large minimum workcell area. What would the PSD Envelope look like if we protected two of the potential entry points with fences, reducing the required safeguarded area? An example is in Figure 3, for which the application falls within the PSD Envelope for a wide range of practical speeds.

Even if you could not fit the application within the PSD Envelope, it may be possible to still use Veo FreeMove as a superior replacement for traditional active safeguarding methods such as (limited and hard to configure) 2D scanners or light curtains. Imagine, for example, a situation where the robot and the human do not interact but need to be close to each other without any guarding or fencing. It would be possible to set up 2D scanners or light curtains to protect the human, but this could be cumbersome for complicated geometries or workspaces with large occluded spaces. With FreeMove, you would get real-time, 3D, dynamic SSM safeguarding capabilities in an easy to install package, as in this machine tending case study.

Now that Veo FreeMove is safety certified to ISO 13849 PLd Cat 3, we look forward to working with you in 2021.Please drop us a line at sales@veobot.com and we can explore your collaborative workcell needs and applications with you.




[1] We are assuming away some limitations that are specific to the current FreeMove 1.0, which include support for single (not multiple) fixed robots (but not a 7th axis such as a rail or a turntable), and certain requirements on the composition of the other items in the workcell, such as whether they are moving or stationary, or materially changing shape. If you have gotten this far in your discussions with our capable sales team and applications engineers, that means those constraints are not binding for your application.

[2] Of course, for your specific robot and application we will estimate these numbers as accurately as we can.

[3] Specifically, the modeling here is for a fully unfenced workcell, where human approach towards the robot can be from any direction, necessitating an MDH that extends outside the full robot envelope.

[4] Let us note that a 5m/sec TCP speed is very fast, at the outer edge of what many robots are capable, so it would be unlikely your application would require such speeds.

 
Figure 2. A representation of the Protective Separation Distance envelope for an unfenced robot.

Figure 2. A representation of the Protective Separation Distance envelope for an unfenced robot.

 
Figure 3. A representation of the Protective Separation Distance Envelope for a robot workcell fenced on two sides

Figure 3. A representation of the Protective Separation Distance Envelope for a robot workcell fenced on two sides