New Machine Day - Robot Invasion!

[Ox]

Shirley the robot programs are saved as well eh?
So the next time you shouldn't have much at all?


-------------------

Think Snow Eh!
Ox

 

[wrustle]

Shirley the robot programs are saved as well eh?
So the next time you shouldn't have much at all?


-------------------

Think Snow Eh!
Ox

I believe you can do that, but not 100% sure. I'll ask Andrew when he gets back from vacation.

In theory, as long as you're using the same table array and locating fixture plate for the array, and the same grippers, I can only imagine it could be done that way.

He's away for 10 days now so we're all left staring at very expensive piece of equipment sitting in the dark. You can certainly bet I will be cross training everyone on that machine once he gets back and gets into the swing of things again!


Later,
Russ

 

[Mcfish]

This is a great thread, thank you. We run all Haas and try to make it so any machinist can walk up to any mill and be off and running in minutes.

We do prototype work (for building robotic cells coincidently) so there are very few repeat programs, but the gist is the same as how you have your machines worked out.

Kudo's to you for looking to the future.

 

[wheelieking71]

At IMTS I had a chat with a vendor who was showing a robot loading pallets that then got swapped into the machine. And one of their staff opined that "yes, this is very powerful, but maybe too complicated for many shops to absorb, we suggest they start with just pallets". Of course they sell pallet pools for their machines so there may be some bias, but the point is well made.

(In other words I agree with dstryr.)

On the other hand, the workflow needs to support the investment - both the cost of the hardware, AND the learning curve. Russ bought something that is supposed to be reasonably well integrated and famaliar (they clearly know haas controls) - so maybe that's the investment that makes the most sense for their work flow.

There will be a prize for shops that can get automation to work for lowest total investment (money, time, floor space, headaches) to make it economic for relatively low value parts. Good luck.

Spot on! My system had about a 24 month ROI. Because it was for one part #. Once it was running? Just keep it fed, and cash the checks.
This is why I rolled the dice. As long as that part ran for two years, at worst case I was even with the iron on the floor.
To bad it only went about 6 months! LOL. I got very, very lucky when I was able to sell it in less than 2 weeks. Very lucky!
Now, had I brought the Flex cell in, intended to cover many different part #s like Russ's system? ROI would have been way, way out there.
The Yamzen Flex cell is a fantastic thing. But, for high mix low volume? Nope!
Not unless you are a Nachi control pendant whiz already, and love making end-effectors.

 

[wheelieking71]

I think if you would have gotten #1 a pallet load solution or #2 a robot load with intuitive easy software you wouldn't have sold that cell. You would have sold something else instead.
A pallet load system is more of an investment but its way more flexible .https://www.erowa.com/en/solutions

I had very, very specific parameters to cover when I set that system up. Simple A/B part. Part comes out finished. Run, reliably, un-manned for as long as possible.
So, the bot had a two gripper knuckle, there was a turn-over station in the bot enclosure, and two air-vises on the table.
Part probing and tool-detection. I could get about 300 blanks on the in-feed belt. That was about 7 hours.
I had the apps guy fully up to speed on the part. Even supplied the tooling and programs (even he was surprised at how dialed I had the process).
He and the automation guy did a fantastic job on the automation. Even I was impressed! LOL
But, it was a lot of work! To start over on another part, would not be a simple chang-over that can happen in one day tops.
And considering my shop dynamic? (its really just me) without that unlimited quantity job? It makes zero sense.
The Flex belongs in the super high volume production world. And, it fit my needs then cause that is where I was at.

 

[bryan_machine]

a key trend (i think!!!) has been for automation and flexibility to improve for ever shorter runs of ever more complicated parts - over say 75 years from hard tooling to numerical controlled tapes to conversational to ever better cam - with more or less parallel improvements in workholding.

no operation the size of wheelie or wrustle would have even considered such a scheme in say 1995....

come 2030 will it be practical to buy a machine without some kind of pallet/shuttle/auto loader?

 

[gustafson]

no operation the size of wheelie or wrustle would have even considered such a scheme in say 1995....

Think about it , in the mid 80's a Bridgeport CNC knee mill was 40k then.
In the early 90s everyone had a 2 axis prototrak
Then Haas was making VMCs as cheap as old CNC knee mills
Honestly there is a number that robots make sense, but it is pretty high
I think I would push parts on the table before that automation.
More parts on the table means a longer cycle, fewer tool changes per part and makes the operator interaction with the machine denser.
But then you battle space, and talent shortages, and , and and....
I think these robots are cool as hell, but for me, I think, they would have to come pretty integrated with the machine. I would want that 10 minute changeover, because with fast machines, 100, 200 parts is a couple hours and on to the next slightly different thing, or priorities change...that's the power we have over offshore, ability to change quickly.
I think a brother speed machine with an integrated robot programmed simultaneously would be a very, very attractive thing for a smaller shop.
I don't know how you do that...but....
Then there is cost
Automation is cheaper than people, but you can lay off people if need be, but boy they get funny when you don't make that machine payment.
So, yes, I see it, I see in the future lots of shops like Wrustles but he made very specific decisions for his very specific work and it has to get a little more flexible and universal before every 5 man shop has 3 men and 2 robots....

 

[bryan_machine]

@gustafson - well put - but don't fixate on robots. could be various kinds of pallet schemes instead. bar fed milling machines exist (very pricey and special) - maybe they will evolve. of course the trick might be some other weird thing - and end-of-arm tool with the abilities of a human hand (yeah sure), and AI software that watches a human do the task with say 3 video cameras, then spits out a robot program to do the same thing. then the human does touch up edits and away you go.... (say.....)

or an auto-fed bandsaw feeding the cutoffs to a pickup station inside the VMC, a schunk style actuator loads cut blanks into a vise.... sort of scheme that could do 1st op on any part from 1" x 1" x 1" to 4" x 6" x 10" (say) - 2nd op is exercise for somebody other than me...

I bet whatever comes to win, we'll be surprized...

 

[GiroDyno]

I saw a robot cell where the EoAT was actually the softjaws, and they were fitted with some type of quick change receiver and an auto-vise on the table. Swapping between parts didn't require changing out parts on the robot, just change the program and the fixturing and robot handling swaps out by itself (obviously not going to be as simple as it sounds). Kind of like the tricked out pallet pools on 5x cells.
Somehow it was kind of neat and also kind of a hack approach at the same time...

 

[wmpy]

I saw a robot cell where the EoAT was actually the softjaws, and they were fitted with some type of quick change receiver and an auto-vise on the table. Swapping between parts didn't require changing out parts on the robot, just change the program and the fixturing and robot handling swaps out by itself (obviously not going to be as simple as it sounds). Kind of like the tricked out pallet pools on 5x cells.
Somehow it was kind of neat and also kind of a hack approach at the same time...

There may be others, but these are the guys that I've seen doing that.

HOME - VersaBuilt Robotics

At VersaBuilt, we’re innovating CNC manufacturing through advanced CNC robotics. Our CNC automation systems are solving some of CNC’s manufacturing’s toughest problems, especially in high-mix, low-volume part applications.

www.versabuilt.com

 

[GiroDyno]

That's very similar but the one I'm thinking of used an off the shelf (but obviously high end) vise like a Schunk? and a yellow Fanuc robot in the demo. Seems like there are a few outfits tackling higher mix automation that way.

 

[Ox]

I don't think that I posted this here yet - as I was trying not to dink up Wruss's story, but if someone is looking for a system, Gosiger had a model at The Show last month. They had one in a few different booths around.

It is a Fanuc robot mounted on a large shelf unit that is totally portable. If you keep your acc/decel rates down, you should be able to use it w/o lagging it down. It is made to be picked up with a fork truck and moved to a new machine time and aggin. It has a built in [human] prox sensor that acts as your light curtain.

It was $140K.


-------------------

Think Snow Eh!
Ox

 

[wheelieking71]

Personally I think what wrustle is doing here is about the most practical application for automation in the small shop.
That being the automation first-OP everything from simple extruded bar. And, hand load any subsequent operations.
After my short stint dealing with "real" automation, the biggest take-away for me was: KISS is key to success for the little guy.
By nature of first operations, there is a margin for error in the extra stock on all sides. This helps greatly towards process reliability.
Of all the complications (or errors) I experienced, most were the vision system, turn-over station, or 2nd-OP loading.
Very few were related to the OP-1 vise. If a guy is just getting started in automation, and/or will be greatly reliant on his-self for all issues? KISS!
Especially those like myself who are not tech savants.

I also feel a greatly un-tapped weapon to dip ones toes in to automation is a Brother pallet-changer, combined with a spindle gripper.
Parts tray on one pallet, work on the other. This keeps the tray largely out of the machining environment (chips are the enemy!).
And the Brother pallet-change is so fast, it is of no consequence.
If I get to the point that I need un-manned operation again, this IS the direction I will go. A 40 tool R650 with a spindle-gripper and air-vise.

 

[Ox]

So you are saying that you would put an In and an OUT tray (?) on either side of one table side?

So you are machining on P1
When you are done, you would load a gripper of some sort into the spindle (BT30 mount)
Pull out the part - and likely blow the Shiite out of the vise now? (somehow?) (via air or coolant)
Index the pallet shuttle to P2
Set yuhr part down on one end.
Pick up a fresh blank on the other end.
Index the pallet shuttle back to P1
Blow the fixture aggin just fer Schidt's and Giggles (maybe)
Insert fresh blank, and run

Am I following you correctly?

And this gripper is actuated via TSC?
And actually may need two different grippers I spose...

So, that brings you to about the same part change time as Wruss's set-up, but with a faster / more $ machine, with less $ automation?

Am I hearin' you correctly???


---------------------

Think Snow Eh!
Ox

 

[bryan_machine]

Whether you are hearing Wheelie correctly, that is what I took him to mean, and think it makes sense, especially if the "working" pallet has some kind of 4th or 4th/5th access device on it.
It would also win (over just a big array of parts in a fixture) if the density of rack storage is much higher than the credible density of vises/fixtures - which would seem to be commonly true.
You could go one tier further, and have hand pallets (think Pierson or Erowa), one side for inbound, one side for outbound.

I note here that it seems Russ's system will be more versatile over time, at the cost of human and resource investment, and it *seems* like Russ's system can deal with bigger parts...

 

[Matt@RFR]

Watch this, Ox:

 

[wheelieking71]

So you are saying that you would put an In and an OUT tray (?) on either side of one table side?

So you are machining on P1
When you are done, you would load a gripper of some sort into the spindle (BT30 mount)
Pull out the part - and likely blow the Shiite out of the vise now? (somehow?) (via air or coolant)
Index the pallet shuttle to P2
Set yuhr part down on one end.
Pick up a fresh blank on the other end.
Index the pallet shuttle back to P1
Blow the fixture aggin just fer Schidt's and Giggles (maybe)
Insert fresh blank, and run

Am I following you correctly?

And this gripper is actuated via TSC?
And actually may need two different grippers I spose...

So, that brings you to about the same part change time as Wruss's set-up, but with a faster / more $ machine, with less $ automation?

Am I hearin' you correctly???


---------------------

Think Snow Eh!
Ox

For the most part, you git it. But, we don't need both in/out tray. Since the stock on the bottom is likely un-machined, just set the part back where it came from.
One real neat thing about grippers I learnt from Scott (awesome dude!) was, you can make one gripper do two things.
Shallow pocket to grab raw stock. Deeper pocket in same gripper to grab finished profile. That way one single gripper can load/un-load.
Effectively turning the spindle in to your robot. Now we are not reliant on a robots accuracy, or programming.
We are dealing with the comfort of the VMC's accuracy (dead nuts by comparison), and M-codes (which we already know).
The other benefit is single piece flow. In the nature of what Wruss is doing, this is huge come changover day.
And as someone who has way, way too much experience packing tables full of as many parts as possible? Single piece flow is huge period.

I'm not saying it is the best solution. But for a truly 1-2 man shop like myself or davidn? It could be game-changer type stuff.
If a guy can get say 50-100blanks on the tray? That is that many times a guy doesn't have to touch a vise.
Only investment in work-holding is an air-vise and set of gripper jaws. Yea, a guy is gonna have to make trays and grippers But, that is easy-peasy.
No, it isn't "true" automation. But it is damn sure better than an employee!

 

[Kingbob]

Watch this, Ox:

that's brilliant, looks like it's already at it's max Z to clear the partition though

 

[wheelieking71]

Watch this, Ox:

Hell yea! Exactly! Thanks Matt, LOL (how have I not seen that one?)

 

[wheelieking71]

Whether you are hearing Wheelie correctly, that is what I took him to mean, and think it makes sense, especially if the "working" pallet has some kind of 4th or 4th/5th access device on it.
It would also win (over just a big array of parts in a fixture) if the density of rack storage is much higher than the credible density of vises/fixtures - which would seem to be commonly true.
You could go one tier further, and have hand pallets (think Pierson or Erowa), one side for inbound, one side for outbound.

I note here that it seems Russ's system will be more versatile over time, at the cost of human and resource investment, and it *seems* like Russ's system can deal with bigger parts...

I'm sure most of you guys have seen me bitch on here about the family of parts I do where 4 of the 8 part#s take 7 operations to complete in a VMC (no way around it, trust me, LOL). and the other 4 part#s take 5 operations. If I had the set-up I elude to with a 5th on one side like bryan just suggested, I could tab all these parts off in one OP. The orders come in between 100-200pcs. The stock is all very similar. Instead of 500-700 vise handle cranks per 100pc order of each part#? I could literally put 100pcs on the tray, push the button, and come back when its done. If that ain't game-changer? I don't know what is. I've been doing these parts 8 years. Had I known everything I know now back then? This iron would already be on my floor! Just thinking through the numbers to type this post (and remembering how many of these parts Tommy scrapped!) is making my head hurt.