Robot Cell Safeguarding

Not entirely PLC related, but here goes:


Question for you folks who are familiar with the RIA R15.06 standard: I don’t have access to the standard at the moment, but am safeguarding a robot cell used for R&D purposes and had a couple of general questions about guarding considerations.


1) Currently there are no guards installed, and the robot reach with tool is close enough to a wall to pin someone against it. I am installing a laser scanner (w/ Safety Relay) in the wall area, which would trip a protective stop covering the entire robot reach including safety distance, but is this enough? Or is there still a minimum clearance required from any hard guard, even when safety devices prevent the robot from moving?


2) I have seen a statement in the standard to the effect that a robot protective stop needs to be initiated when running an automatic sequence and a person enters the vicinity. But what does the standard say about teach and manual operation from the pendant? I seem to remember something like there were certain speed limits where motion was ok, but I can’t recall if this was true (talking standard robots, not cobots).


3) Very generic/general question, probably not so easy to answer: how far does the standard require you go for the safeguarding of equipment (this is a US installation, so OSHA)? For example, I can install dual laser scanners at about thigh height to cover the entire perimeter around the robot. However, it could be possible for someone to enter the robot perimeter, climb up on the conveyor out of sight of the scanners, which if the robot is set to auto-rearm, could turn out to be very dangerous. How far is “reasonable” when considering these types of dangers in a risk assessment?



Thanks for any input.

My two cents... I'm from Canada so I won't quote our rules; there may not be a direct link. Your last statement is where you should start: a risk assessment. Once this is determined, the degree to which you must provide safe design will be more clear. Robot programming can provide "soft stops". Interlocking a zone safety with a move into a dangerous area is a good start. I've seen both light curtains (with muting) and floor scanners (two zones of monitoring) used for these applications. What is the robot doing? If it is a "pick and place" application, your safety has to take into account the possibility of a part becoming airborn should a gripper fail.

I work at a plant where a woman got killed by a palletizing robot. Yes she pushed her way past the safety gate, but it seems to me if a robot is told to make a move and doesn't get there, it should go to a safe position. That would of saved her life. It basically went to pickup the same box she was getting and pinned her to the conveyor and she suffocated to death.


Should that robot not have given up and moved to a safe location?
or at least reversed the last move that failed?

Jieve said:

How far is “reasonable” when considering these types of dangers in a risk assessment?

Click to expand...

"Reasonable" will be decided in a court of law during any product defect litigation. Are you willing to testify in court that you did everything reasonable when someone gets hurt?
We use the everything "conceivable" standard. Think of every way conceivable someone may use it, misuse it, or how it might malfunction.

You need to contract someone who has experience in this matter.