Let's be careful out there!

My robots moving 15# castings could smash you like a bug before deciding, "wow, I think I hit something", can't imagine one that picks up car bodies.
 
I work with robots regularly, and frequently I have to work inside the cell while setting up. This involves having a buddy outside the cell to activate the safety circuits after the door is closed. While I am inside my safety is totally in my own hands. It is important not to allow oneself be distracted by cellphones and the likes which can break your concentration and perhaps cause you to make a mistake. As with all things, familiarity breeds contempt, and unfortunately even a tiny 10KG robot can kill. It would never be the machines fault though, so as in the topic header, lets be extra careful.
 
even a tiny 10KG robot can kill. It would never be the machines fault though

I am very sorry to hi-jack here, please except my sincerest, apologies.
But I just could not pass up this opportunity to point out the level headed, straight minded, reasonable thinking that went into producing this comment!

Bravo to you sir, [Tip Hat]

BCS
 
There was a lady killed at one of the plants I do work at (not programming robots) it basically went to pick up a box and she got in the way and it smothered her to death by keeping her pinned to the conveyor.

My question is why didn't the robot go to a safe position and shutdown the millisecond it didn't make it to the loading position? Was that bad programming on the robot companies part?
 
My question is why didn't the robot go to a safe position and shutdown the millisecond it didn't make it to the loading position? Was that bad programming on the robot companies part?
I've never seen a robot that functions this way. An overtorque condition on any of the joints will generally trigger an immediate fault and the robot will not move until it is reset by an operator. Usually the cause of the fault is a jammed part or loose tooling, and it is often necessary to examine the cause of the fault before moving the robot. I see no advantage to an automatic move back to a "safe" position--this could cause more harm than it prevents. What if the robot hits a person on the way back?
 
I see no advantage to an automatic move back to a "safe" position--this could cause more harm than it prevents.
And, who decides where the 'safe' position will be? If there's just one robot in a cell I guess it wouldn't be too difficult. OTOH, what about a cell with several interacting robots? That would be a tougher proposition.
 
I work with robots regularly, and frequently I have to work inside the cell while setting up. This involves having a buddy outside the cell to activate the safety circuits after the door is closed. While I am inside my safety is totally in my own hands.
For such a situation, shouldnt there be a safe slow movement speed of the robot ?
I have also seen a kind of tethered "dead-mans button". The guy inside the cell has to press the button constantly, allowing the robot to move at the special slow speed.
I am no robot expert though.
 
For such a situation, shouldnt there be a safe slow movement speed of the robot ?
The old ABB robot I worked with would only run in slow - 10% of max. programmed IIRC - when the programming pendant was removed from its cradle.
 
That takes care of Asimov's Rule #1 - how soon before the rest?
https://en.wikipedia.org/wiki/Three_Laws_of_Robotics

I don't think we have been smart enough to build robots in 2015 using a mere 3 little rules invented in 1942. The problem is that Asimov's Laws of Robotics are not built into robots, but instead crude efforts are made to add them externally. The external safety devices are better than nothing, but not fail-safe as shown by the fatal accidents. The robot itself is going to have to become "safe" just as Asimov predicted.
 
Last edited:

Similar Topics

Hello everyone! So basically I have this PF 525 that no matter what I do won’t let me speed down. It always shows 66.66 Hz and the parameters to...
Replies
11
Views
1,179
35 years ago - 167 Souls were lost on Pipe Alpha (Oil Platform in the North Sea) all due to a failed management system See more information here...
Replies
3
Views
801
I have begun using 480VAC power supplies in my panels to keep heat from a transformer out of the panel. This works great if the panel is mounted...
Replies
17
Views
4,127
I am deciding on components for a bayed modular enclosure containing multiple VFDs (480V/350A/321Hp each) for a dynamometer. I am concerned given...
Replies
7
Views
1,700
Hi, I am having trouble with the comms within FT view SE v11 (premanent license). I created an HMI program a while back which was working just...
Replies
0
Views
926
Back
Top Bottom