Discuss How to Start Safe Mode while moving. Ensuring Human Life Safety, Avoiding Collisions, Repairing After Collisions, Electrical Failures on KUKA KRC2ed05 Robot.

    I have an old KRC2 Robot that is in the process of CNC machining configuration, I am learning the process of configuring the Robot's Safety mode to ensure that the Robot does not collide with fixed positions around it such as the external turntable, vise or other surrounding obstacles, ensuring operator safety, electrical safety such as starting up by itself, (most importantly ensuring human life safety). Then how to operate again after a slight mechanical collision such as a broken milling cutter.... Can anyone with experience share with me? Thank you very much.

    That's an extremely vague set of questions.

    If the robot is stopped b/c it vioaltes a safety zone, or because of a collision error, recovery required switching to Teach mode and jogging the robot clear. There is no automated recovery from this category of error, for safety reasons.

    "Human life safety" requires correctly fencing off the robot with physical barriers and safety rated components like light curtains, gate interlocks, or area scanners, and correctly wiring those safety signals to the correct safety inputs on the KRC. Every human fatality involving robots in the past few decades has involved either operator stupidity (hotwiring a safety gate, removing safety fence, closing a gate with a person inside the perimeter), or failure to meet safety design requirements (open fencing with no light curtains).

    Safety-rated components all work via dual-channel signals. Using non-safety-rated similar components (for example, normal relays) may allow the system to operate, but will not be safe.

    Quote from SkyeFire

    That's an extremely vague set of questions.

    If the robot is stopped b/c it vioaltes a safety zone, or because of a collision error, recovery required switching to Teach mode and jogging the robot clear. There is no automated recovery from this category of error, for safety reasons.

    "Human life safety" requires correctly fencing off the robot with physical barriers and safety rated components like light curtains, gate interlocks, or area scanners, and correctly wiring those safety signals to the correct safety inputs on the KRC. Every human fatality involving robots in the past few decades has involved either operator stupidity (hotwiring a safety gate, removing safety fence, closing a gate with a person inside the perimeter), or failure to meet safety design requirements (open fencing with no light curtains).

    Safety-rated components all work via dual-channel signals. Using non-safety-rated similar components (for example, normal relays) may allow the system to operate, but will not be safe.

    thank you, i dont know where to start with the keyword "Safety". what is the first job, what are the steps to implement safety measures? what are the safety methods, what are the necessary equipment to prepare? what are the steps to implement interlock for robots? can you help me answer these questions? i would appreciate it.

    Edited once, last by PhunxamLinda (December 20, 2024 at 10:26 PM).

    it all starts with training... first you need to become familiar with machine safeguarding in your country (this is general), then on top of that are special considerations... including robot safety, laser safety, press safety etc. (whatever hazards are present).

    standards tell you what you can and cannot do, what the restrictions are and how to calculate things to determine if your design meets the standard or it fails... for example if the fence around the robot is a wire mesh, size of the opening in the fence is important factor. the larger the opening the larger cell perimeter need to be,

    then comes cell evaluation - all hazards are to be identified and key things are measured (distances, velocities, stop types, stopping time...) ... after this suitable protection measures are determined.

    with that one can start working on design... part of this is mechanical (fences, posts, gates,...), part is electrical (safety circuit design, selecting suitable devices, determine their response time, calculate... replace what does not fit, do it again...).

    in general when it comes to industrial robots, safeguarding is a high standard, everything is redundant, everything is monitored, safety devices need to be very difficult to defeat (one must try really really hard to bypass some safety device).

    when interlocking different devices the same question always arises - can it hurt someone? if answer is yes, then interlocks will need to be of higher grade as above mentioned (must protect human life and limb).

    if answer is no, it is all about money... after all machines can be replaced. so you have a choice to make... do you go with simple standard I/O handshake or still go with some level of safety interlock. first one is cheaper and may be justified in places where possible damage is low or rare so it can be factored into production cost. but if potential damage is too significant or process is critical, you may have to spend more up front to prevent collisions, damage and downtime.

    1) read pinned topic: READ FIRST...

    2) if you have an issue with robot, post question in the correct forum section... do NOT contact me directly

    3) read 1 and 2

    thank you. I understand that safety is always a top priority, the first thing is to set up physical limits (such as protective fences, proximity sensors) or soft limits of the robot, but what I want to ask here is how to implement it? where to start (I mean software and electrical implementation)? what needs to be prepared. How can the Robot recognize obstacles to avoid collisions in a narrow environment with many obstacles around.

    Quote from PhunxamLinda

    How can the Robot recognize obstacles to avoid collisions in a narrow environment with many obstacles around.

    It does not. That responsibility is entirely on the programmer. The robot has no "awareness" of anything around it, it only repeats a set of recorded motions endlessly. Programs can branch on IF/THEN conditions, but that just plays a different set of recorded motions.

    Likewise, there is nothing that prevents the robot from simply punching straight through a solid fence if programmed incorrectly. There's also little that prevents a robot from striking an obstacle with lethal force and not even noticing.

    Preventing the robot from striking something inside the safety perimeter is the programmer's job. Preventing the robot from punching out through the perimeter is the job of whomever is integrating the robot. For a KRC2, this could take the form of bolt-on hard stops, or it could involve the SafeOperation option if the KRC has that option installed. For SafeOperation, it is possible to define the fence line, and a "globe" around the tool, that will E-Stop the robot if the fence and globe ever touch. This does not prevent someone from programming the robot to violate the safety perimeter, it merely monitors the robot in realtime and issues an E-Stop command. However, this still depends on the programmer to set up correctly, and must take stopping distance into account -- for example, at full speed, a KR210 might take 400-500mm to stop after an E-Stop occurs.

Advertising from our partners