The A.I. revolt has begun! Robot kills Man!

[ControlD]

Haven't programmed one in many years though. Have set settings on them recently and modified general code but no movements. I usually program the cell PLC controllers and let another guy program the robot (he's much more on top of them than I am currently). I do design the cell electrical systems including robots and safety circuits though.

Sounds like we have very similar jobs. I will program a robot when I have to, but we have a crew of robot experts that are much better at it than I am. I am more of a PLC / NC project leader these days and I prefer to stay in that area as much as possible.

We do a lot of assembly cells that can have up to 10 robots inside. I would have to ask for permission before posting a video though. You really have to know what you are doing when you step inside the guarding of such a cell and I never take anything for granted. A robot can make some really funky moves and you would never know you are in trouble until it is way too late.

 

[Jeff7]

Sounds like we have very similar jobs. I will program a robot when I have to, but we have a crew of robot experts that are much better at it than I am. I am more of a PLC / NC project leader these days and I prefer to stay in that area as much as possible.

We do a lot of assembly cells that can have up to 10 robots inside. I would have to ask for permission before posting a video though. You really have to know what you are doing when you step inside the guarding of such a cell and I never take anything for granted. A robot can make some really funky moves and you would never know you are in trouble until it is way too late.

God that'd make me nervous.
Some of those things could kill you before your muscles even had time to contract.

 

[Sabrewings]

Will the robot be put down?

 

[Engineer]

God that'd make me nervous.
Some of those things could kill you before your muscles even had time to contract.

That's the reason that the deadman switch is a 3 position switch. If you squeeze too hard or let off of it, the robot stops. Still....very valid point.

I would have to ask for permission before posting a video though.

The video that I linked was posted on Youtube by the customer although they wouldn't care if we posted it. The small company I work for posts quite a few.

 

[yottabit]

Feel like I need to post this clip here
https://www.youtube.com/watch?v=bxbjZiKAZP4

 

[Engineer]

Feel like I need to post this clip here
https://www.youtube.com/watch?v=bxbjZiKAZP4

LOL....screw that. I don't know anyone who programs these things that would ride them....including me!

 

[pcgeek11]

Yes the guy should have locked out. But I know from experience that it doesn't always matter because I got bit once despite locking out an electrical panel.

If so then You Failed to Lockout / Tagout the machine properly.

 

[pcgeek11]

It is done all the time. Especially with robot cells.

It could be that lockout/tagout was not used here and that would be a problem. When working on a robot cell however, it generally isn't the main disconnect that is locked out. To do something like touching up robot points you do the following:

(1) Open the safety door next to the robot controller
(2) Place your lock in the safety door lockout point
(3) Grab the robot hand pendent
(4) Enter the cell
(5) Enable the live man switch giving control over the robot

Once #5 is active, safety is up to the technician. you can still kill yourself if you don't know what you are doing.

All of the robots that I have worked on the pendant is outside of the safety area guarding. You can teach points etc fine without going into the area if it has been properly programmed to begin with.

 

[ControlD]

All of the robots that I have worked on the pendant is outside of the safety area guarding. You can teach points etc fine without going into the area if it has been properly programmed to begin with.

Depends on the mechanical layout of the cell. Some can be taught outside some can't. It has less to do with proper programming and more to do with mechanical layout. Which explains why some machines have procedures for entering the guarding with the pendent.

 

[Engineer]

All of the robots that I have worked on the pendant is outside of the safety area guarding. You can teach points etc fine without going into the area if it has been properly programmed to begin with.

LOL. For every cell? Not a chance.....period. Maybe the stuff you worked on but based on that response, it must be very limited.

 

[futurefields]

Video of the guy being smashed by the robot?

It must exist, there are cameras in these facilities.

 

[BarkingGhostar]

 

[pcgeek11]

LOL. For every cell? Not a chance.....period. Maybe the stuff you worked on but based on that response, it must be very limited.

Nope just properly designed enclosures and taking the time to move the robot into the teach position then disabling the movement when it needs a closer visual inspection before proceeding. Takes more time but prevents injury and death.

We usually teach a series of fixed points that are visible from outside the safety enclosed area and the remainder of the picks and places are computed from those fixed references.

Like I said proper programming helps. Older Adept Arms, Staubli and Bosch / Rexroth Scara Robot Arms. Most are in a high precision application.

It takes longer, but I have never been hit or had any close calls. Just because you live dangerously doesn't mean everyone must.

 

[Kaido]

A good article on responsibility here:

http://gizmodo.com/robots-can-t-kill-you-and-claiming-they-can-is-dangerou-1715926940

 

[balloonshark]

If so then You Failed to Lockout / Tagout the machine properly.

No I didn't. I was checking a panel for bonded neutrals for the first time ever. Our supervisor decided to make a ground hot to fix a problem instead of tearing into a wall to fix it properly. I guess the only thing I did wrong was taking for granted that puling the main wires would remove all power from the panel and not checking every ground with a tester.

 

[ControlD]

A good article on responsibility here:

http://gizmodo.com/robots-can-t-kill-you-and-claiming-they-can-is-dangerou-1715926940

I think that article highlights some common misunderstanding due to the terminology being used. When non-industry people hear the work "robot" they tend to think of a completely different type of mechanism than what this accident involved. A "robot" in industrial terms is a mechanism like I linked to above (the yellow arm). There really isn't much in the way of artificial intelligence going on there, it is simply a tool. Like any power tool, improper use can result in injury or death. As programmable safety becomes more and more sophisticated, it gets harder and harder to cheat the system but you can still get around safety if you really work at it. I will be interested to see what the investigation shows if we ever get access to it.

 

[Rakehellion]

 

[pcgeek11]

No I didn't. I was checking a panel for bonded neutrals for the first time ever. Our supervisor decided to make a ground hot to fix a problem instead of tearing into a wall to fix it properly. I guess the only thing I did wrong was taking for granted that puling the main wires would remove all power from the panel and not checking every ground with a tester.

Then your supervisor should be fired. Why would you ever make a ground hot? That is just asking to kill someone.

 

[MongGrel]

I simply don't know how you're going to build some of these cells unless you're right up inside the guts of the cells. Now you don't necessarily need to have the robot set to 100% teach pendant speeds (not all robots support this anyway). You may have it set to max 20% speed just for rough-in purposes. Still can be scary when you're right up inside the cell (beside the place or pick points) trying to run at 20%.


We try to video our stuff now (phones everywhere) so we can analyze what's going on with picking / placing / general movements. Movements at a slow speed may not be the same path at high speed (as the robot controller interpolates points somewhat different at higher speeds unless you have it go to a ZERO point (i.e. no radius moves, etc).

Have seen several programmers hit with robots. Have nearly been hit myself a few times.

Haven't programmed one in many years though. Have set settings on them recently and modified general code but no movements. I usually program the cell PLC controllers and let another guy program the robot (he's much more on top of them than I am currently). I do design the cell electrical systems including robots and safety circuits though.

By the way, I'm not advocating anyone other than the guy holding the teach pendant with dead-man circuit go anywhere near the equipment like this. Without some sort of dead man safety handheld device, you're just asking to get killed, especially if running anywhere near full speeds.

Here's one of the last 'big boy' robots that I and the other guy worked on. Uses lasers and a vision system to find the brake drums (up to 150 pounds each) and unload them from the pallets. Then stacks the plywood between layers and finally removes and stacks the pallets. Starts over by conveying a new pallet in once old one gone. The openings are guarded with light curtains with the ones on the conveyor set up to allow the pallets to enter via a certain sensor sequence without stopping the robot.

https://www.youtube.com/watch?v=SNEloS7rU4E

You do have a point there.

You would have to admit something still did not sound right there.

You can always have the rapid moves on the controller turned down so it does not roll at full speed when trouble shooting things unless the programming somehow overrode the machine parameters also.

I've almost been hit myself once or twice myself in the last 30 years, once they are set to do something they will just do it and do not care if a person is in the way or not in some circumstances,

 

[MongGrel]

Even an old video these days, but are the type of thing you can do with a 5 axis on a daily basis non production.

The others are set up for full on production in cells, but even then you don't want to waste the time doing setup o these for a run of one thing.

https://www.youtube.com/watch?v=uc5P6Ss3LRE

 

[MongGrel]

Pretty nice one of the probe touching off on the part at the start there.

https://www.youtube.com/watch?v=a2sPUrtu3Us

 

[NoCreativity]

So many naysayers in this thread and yet the revolt continues.

 

[Eug]

https://twitter.com/sarahoconnor_/status/616282747200479232

 

[Jeff7]

Even an old video these days, but are the type of thing you can do with a 5 axis on a daily basis non production.

The others are set up for full on production in cells, but even then you don't want to waste the time doing setup o these for a run of one thing.

https://www.youtube.com/watch?v=uc5P6Ss3LRE

As someone who only has access to a fairly sluggish CNC router, those things are amazing to watch. Large swaths of metal just vanish so quickly.

 

[Engineer]

Nope just properly designed enclosures and taking the time to move the robot into the teach position then disabling the movement when it needs a closer visual inspection before proceeding. Takes more time but prevents injury and death.

We usually teach a series of fixed points that are visible from outside the safety enclosed area and the remainder of the picks and places are computed from those fixed references.

Like I said proper programming helps. Older Adept Arms, Staubli and Bosch / Rexroth Scara Robot Arms. Most are in a high precision application.

It takes longer, but I have never been hit or had any close calls. Just because you live dangerously doesn't mean everyone must.

I'll say that you may do that in a production world but you will not do that in a development / machine building world, and I still say that your situation is unique. If it's possible to do it that way, sure....the more safety the better. In the real world, most of the time somebody is going to have to grab a pendent and go inside lots of cells.

So you are into the machine and automation cell building business and have never had anyone go inside a cell with a teach pendent and teach anything at all? Was all taught outside of the cell area, especially high precision pickup and place points? Have you done anything like that for automotive parts? From experience, it's damn near impossible considering the shapes of the machines and the parts produced. I've done it on a few machines but they are few and far between (using reference points for the other points).

https://www.youtube.com/watch?v=f9vnwhBvbIc