Automation

KBW · Dec 28, 2007

Any one with some form of automation / robot attendant to machine experience, you would like to share?

We have a good candidate for robot pick, load - lathe job that could lend it's self to automation, with two lathe running one end the majority of machine work, feeding one machine handling a quick bore, threading op.

A year or so ago there was a show on Discovery or TLC that showed automation in the Beratta factory. Anyone remember this or maybe a link to where I can download all or some of it.

Thanks.

matt_isserstedt · Dec 28, 2007

I work in a shop with ~500 industrial* robots. 6 axis, fairly big models that can handle 250lbs of end effector.

The robot runs a program that you 'teach' it, then works like other automation, you have to develop digital interlocks with the other tooling that it interfaces with**, usually a good idea to have a PLC in there so you can troubleshoot online. HMI and safety stuff is going to be a significant part of the control system. You need an end effector, usually some sort of gripper for a material handling robot that can be shop-built or based on bought pieces from DeStaco (air-operated gruppo) or Norgren Automotive.

* as opposed to residential robots

** IOW don't call the robot program to enter a turning center unless the door-open limit switch is made and spindle speed is at zero

precisionworks · Dec 28, 2007

There are more books on the subject than you can read in one lifetime ... abebooks.com & alibris.com both have a good selection. I just looked at both sites & books are available starting at $1.00.

As Matt said, a PLC makes life easier ... as long as you have a strong background in PLC function/programming ... lots of books on that as well. One of the best basic books IMO is Introduction to Programmable Logic Controllers (ISBN 9780827378667), although it isn't cheap ... my copy cost almost $10

Another information source is the manufacturer's rep. Most are more than happy to come to your factory & provide ideas, estimates, quotes, CAD drawings, etc.

Tdkkart · Dec 28, 2007

As you start to study this, take into consideration not only the time savings available from the new automated process, but also the lost time that will happen when the shiny new system takes a crap.
If the new process saves 30 seconds/cycle, consider how many cycles you'll lose each time the machine screws up and comes to a screeching halt until repaired.
30 seconds is an eternity in an automated cycle, but a very short amount of time once the machine stops and the repairman has to be called.
Automatic toolchangers, pallet changers, bar feeders, robots, etc are all fine and good until they crash, and then they can cause some of the worst dog-knots imagineable.

G-Auto · Dec 28, 2007

Cost?

Just in general how much does a robot cost if it would be working with parts 2lbs or so? Just a rough figure?

LockNut · Dec 28, 2007

KBW,
We use robots here at work to make bone screws. We have Denso, Fanuc and Motoman in our plant. Could be one of the best decisions that were made here since time in machine has dropped to almost zero. The Denso's that I take care of do not go down. In 6 years we have not had a single breakdown of one of these robots where it was down for more than an hour or so. It does help that we have a decent maintenance group here. But, these thing run and run and run. The turning centers these are on have broken down way more than the robots have. Please don't take to heart what Tdkkart has to say. He has a good point I guess, but my experience has it that you will get way more out of it than you might think. Automation, for some jobs, just makes a lot of sense these days. You have to eek every ounce of efficiency from your processes these days if you want to make money. Just my 2 cents.

Paul

precisionworks · Dec 28, 2007

Automation, for some jobs, just makes a lot of sense these days.

Where I work, we fill bottles. Our manual bottle feeder (Mrs. Smith) works at a dead run to keep 100+ bottles/minute feeding down the flat chain conveyor. Then the bottles are autofilled, autocapped, autosealed, and auto labeled. At that point Mrs. Black & Mrs. White make cartons, place the bottles in the cartons, and push the carton through a taper.

Even with our manual bottlenecks (pun intended), the automation has more than tripled total product output. A couple of pick & place robots would more than double that.

clutch · Dec 28, 2007

Robots are fantastic devices if your process will accommodate them. We have a few machines with loaders, Mazak Multiplex's, Index G200, a CL25 cell with an overhead loader.

A number of years ago I used fanuc robots to perform waterjet profiling of headliner shells, put glue patterns on shelf panels, or used a in-house built X or XY motion system to perform pour patterns in S-RIM applications.

As far as the comment on taking a crap. When an operator blows an offset on a lathe, you come to a screeching halt. When the operator is taking a crap, the cell is still running. I'll go with the loader or robot.

If your support staff isn't up to correcting the occasional problem then likely you are bottom feeding pay wise when it comes to maintenance personnel. The guys and gals on the off shifts have to step in the for the engineers when the dayshift engineers go home and turn off their cells and beepers. Not that that ever happens.

Clutch

KBW · Dec 28, 2007

Thanks for all the replys, pro's and cautionary. We have a long run job that would lend itself to robotic usage in transfering from one machine to another.
This isn't going to be a by the seat of our pants install, we'll go pro all the way to interface with the Mazak Multi axis lathes it will be serving.

Any one else want to wade in...

boslab · Dec 28, 2007

misibushi movemaster are about the right size, the dcu can interface with a washing machine if you feel like it [so long as you know the rs232/whatever/424 the lathe is outputting via serial, stop bits parity and such]
have a look on thier site, the do some neat plc,s also
mark

matt_isserstedt · Dec 29, 2007

Along the lines of what has already been said, if you aren't already running robots you'll almost have to get someone to fill a position along the lines of "controls engineer".

Whether they are degreed or not is moot, whether they can get the job done setting up the automation and troubleshooting it is everything.

KBW · Dec 29, 2007

This particular job will be from stainless steel, how have the one's of you with robot tending, how have you over come the "stringers" and "bird nest" that happen with with stainless. How do you handle the blow off or cleaning the collet before loading the next part.

Ox · Dec 29, 2007

KBW:

Just want to make sure that you did in fact receive my PM. That's all.

The PM notification doesn't exactly jump out at you so I am not sure if you are familier with them.

..and you didn't list an e-mail either...

JoeFin · Dec 29, 2007

I’ve worked with the Yamaha robotics. It was handling small pieces of very expensive glass. Don’t know about maintenance intervals or down time as this was strictly turn-key design build stuff. How ever the teach-mode and logic programming was a breaze for anyone to pick up.

You will need a “Front-end” controller to play traffic cop between all your machines and the robots. Don’t let the bean counters catch wind the data logs can also track productivity. They’ll never let up on demanding the information once they get a taste for it

matt_isserstedt · Dec 29, 2007

KBW, you asked about how to clear the stringy chips and blow out the collet.

The robots I am familiar with hold spot weld guns or are material handlers with grippers or vacuum cups but I'd make a multi-function end effector, use something like a Cap'n Hook to clear the chips and give it a hard pipe that you aim at the collet and then set a bit to energize a solenoid valve and start the air blast, short time delay, then send the spool the other way by toggling the bit.

Think of how you'd do this by hand and then make the robot "do" the same thing.

KBW · Jan 1, 2008

The ones of you that implemented a robotic to tend machines, especially lathes do you see
528 hours real production time per month for 5 day work weeks and 624 hours for 6 day work weeks per month?

This is numbers simply pulled from multiplying by 24, but is it realistic?

And if it is why isn't every one going this way?

I'm just trying to see if this is a pipe dream or if numbers like this are do able.

Search

Largest Manufacturing Technology Community on the Web

Search

Automation

KBW

Cast Iron

matt_isserstedt

Diamond

precisionworks

Diamond

Tdkkart

Aluminum

G-Auto

Hot Rolled

LockNut

Stainless

precisionworks

Diamond

clutch

Hot Rolled

KBW

Cast Iron

boslab

Titanium

matt_isserstedt

Diamond

KBW

Cast Iron

Ox

Diamond

JoeFin

Stainless

matt_isserstedt

Diamond

KBW

Cast Iron

Similar threads