What's new
What's new

What’s the longest you’ve ever been “slow”?

Some customers have been telling me they are slow and others are flat out. Electronics guys are still trying to catch up with the backlog due to component availability.
 
Yes that is the question, which is better:

1. A horizontal with a high density work-holding (+maybe a cell system) and a guy changing parts all day?
or
2. A Brother/Robodrill fed by a robot making parts complete 24/7 and the dude just harvests and boxes parts up all day?

Which say you oh wise Internet Gods of CNC?
 
Yes that is the question, which is better:

1. A horizontal with a high density work-holding (+maybe a cell system) and a guy changing parts all day?
or
2. A Brother/Robodrill fed by a robot making parts complete 24/7 and the dude just harvests and boxes parts up all day?

Which say you oh wise Internet Gods of CNC?
We have four horizontals - a 2-pallet, a 6-pallet, and two 21-pallets. It takes between 30 and 90 minutes to completely load a machine to run unattended for the entire day (18-24 hours). Pallet-handling automation on HMCs efficiently handles high mix, daily production with low relative labor cost.

But work handling automation on machines running single piece flow have an even lower labor requirement. I'd estimate around 25% or less. The problem is the R&D/setup time, which traditionally takes weeks to dial in. Makes no sense to do things this way when your PO only requires a few days worth of production. I think there's a way around this though, using a combination of spindle grippers and cobots on 5-axis machines, reducing the automation setup time on a new job to a single day.

Side note: pallet handling on 5-axis machines is horrendous, IME. You spend 15 seconds loading parts and 60 seconds waiting for the pallet handler to grab the next pallet. An absolute flow-killer.
 
I think there's a way around this though, using a combination of spindle grippers and cobots on 5-axis machines, reducing the automation setup time on a new job to a single day.

Side note: pallet handling on 5-axis machines is horrendous, IME. You spend 15 seconds loading parts and 60 seconds waiting for the pallet handler to grab the next pallet. An absolute flow-killer.
Can you elaborate on this, with the context as you mention, IE you do manufacturing, I do a week or two of production then re-setup for something else.
I don't do onesie twosie job shop though either.

I debated on a 12-16 pallet horizontal, but we have started to get what others have mentioned, customers making multiple parts into one Franken part that has to be 5 axis.

I then was thinking we have to have a 5 axis.
robot /grippers load raw stock, or Erowa Compact80 loading small 6" puck pallets?

I thought the pucks pallets.
 
Can you elaborate on this, with the context as you mention, IE you do manufacturing, I do a week or two of production then re-setup for something else.
I don't do onesie twosie job shop though either.

I debated on a 12-16 pallet horizontal, but we have started to get what others have mentioned, customers making multiple parts into one Franken part that has to be 5 axis.

I then was thinking we have to have a 5 axis.
robot /grippers load raw stock, or Erowa Compact80 loading small 6" puck pallets?

I thought the pucks pallets.
Loading raw stock is ideal.
  1. Significant labor savings handling blanks in a tray vs. blanks in a fixture, about 75%
  2. More blanks in a given volume = longer unattended run times. If a compact pallet pool can hold 60 mini pallets with 1 part each, the same volume can hold 240 blanks
  3. You only need to setup and dial in one vise, vs 60 vises
So how do you do it?
  1. Pneumatic vise mounted in a 5-axis machine. I've shelved every project I was previously working on to develop our own.
  2. Spindle gripper(s) actuated by thru-spindle air or coolant
  3. Cobot that brings parts trays in and out of the machine for the spindle to grab
Robot programming is greatly simplified because it doesn't handle any parts directly. It only handles the trays. It can also tend multiple machines at the same time as long as they're all within reach. We just took delivery of a Fanuc CRX-25iA and it's the one I'd recommend with ~1800mm reach.

Spindle gripping is easy. It's just G-code. You can also have multiple spindle grippers in your ATC to handle different parts, the same parts at different stages of completion, or the same parts in different orientations.
 
It's fine if you make your own products in high quantity but most of us don't see the quantity and commitment from customers to justify/risk it on that. There's new machines for sale every week on contracts that "got canceled".
 
Loading raw stock is ideal.
  1. Significant labor savings handling blanks in a tray vs. blanks in a fixture, about 75%
  2. More blanks in a given volume = longer unattended run times. If a compact pallet pool can hold 60 mini pallets with 1 part each, the same volume can hold 240 blanks
  3. You only need to setup and dial in one vise, vs 60 vises
So how do you do it?
  1. Pneumatic vise mounted in a 5-axis machine. I've shelved every project I was previously working on to develop our own.
  2. Spindle gripper(s) actuated by thru-spindle air or coolant
  3. Cobot that brings parts trays in and out of the machine for the spindle to grab
Robot programming is greatly simplified because it doesn't handle any parts directly. It only handles the trays. It can also tend multiple machines at the same time as long as they're all within reach. We just took delivery of a Fanuc CRX-25iA and it's the one I'd recommend with ~1800mm reach.

Spindle gripping is easy. It's just G-code. You can also have multiple spindle grippers in your ATC to handle different parts, the same parts at different stages of completion, or the same parts in different orientations.
So are you doing an auto door on the 5 axis, and the robot holds a tray in for pickup?
And then I assume you have multiple trays?

Oh wait, I see you said that already, nice!
 
Loading raw stock is ideal.
  1. Significant labor savings handling blanks in a tray vs. blanks in a fixture, about 75%
  2. More blanks in a given volume = longer unattended run times. If a compact pallet pool can hold 60 mini pallets with 1 part each, the same volume can hold 240 blanks
  3. You only need to setup and dial in one vise, vs 60 vises
So how do you do it?
  1. Pneumatic vise mounted in a 5-axis machine. I've shelved every project I was previously working on to develop our own.
  2. Spindle gripper(s) actuated by thru-spindle air or coolant
  3. Cobot that brings parts trays in and out of the machine for the spindle to grab
Robot programming is greatly simplified because it doesn't handle any parts directly. It only handles the trays. It can also tend multiple machines at the same time as long as they're all within reach. We just took delivery of a Fanuc CRX-25iA and it's the one I'd recommend with ~1800mm reach.

Spindle gripping is easy. It's just G-code. You can also have multiple spindle grippers in your ATC to handle different parts, the same parts at different stages of completion, or the same parts in different orientations.
OMG that is straight up WIZARD! Robots slinging trays of parts instead of fixtures or individual parts is just genius.

I hope you don't mind if I use this for some inspiration for our next cell....
 
  1. Pneumatic vise mounted in a 5-axis machine. I've shelved every project I was previously working on to develop our own.
  2. Spindle gripper(s) actuated by thru-spindle air or coolant
  3. Cobot that brings parts trays in and out of the machine for the spindle to grab
I've been noodling kind of the same thing, but with a simple tray on a pivot (vertical axle) or linear rails, with a pneumatic ram that sends it in and out of the auto door instead of a robot.

I'm glad I'm on the right path. Thanks Eric.
 
It's fine if you make your own products in high quantity but most of us don't see the quantity and commitment from customers to justify/risk it on that. There's new machines for sale every week on contracts that "got canceled".

Light automation is possible and quick to implement with spindle grippers after you've done a little bit groundwork creating reusable templates. Even if you only automate 5 parts at a time, it can make a big difference.

In some cases you could forego the parts tray entirely by using multiple grippers, loaded with blanks and stored in the ATC. The grippers could be of a lower cost variety, while a better gripper is used for regripping and fine placement in the vise.
 
Light automation is possible and quick to implement with spindle grippers after you've done a little bit groundwork creating reusable templates. Even if you only automate 5 parts at a time, it can make a big difference.

In some cases you could forego the parts tray entirely by using multiple grippers, loaded with blanks and stored in the ATC. The grippers could be of a lower cost variety, while a better gripper is used for regripping and fine placement in the vise.
You did see this strange rig.
 
I've been noodling kind of the same thing, but with a simple tray on a pivot (vertical axle) or linear rails, with a pneumatic ram that sends it in and out of the auto door instead of a robot.

I'm glad I'm on the right path. Thanks Eric.

very simple to do and don’t even need any real electronics Could be all done with stops with adjusters to fine tune.

I like the idea of spindle grippers. My guess this would be overall faster then a Cobot and not needing anything beyond the cnc being programmed.
 
This is all really interesting and also really depressing. Anything requiring automation will be sent to China.

Edit
Should have said in Australia
 
Last edited:
very simple to do and don’t even need any real electronics Could be all done with stops with adjusters to fine tune.

I like the idea of spindle grippers. My guess this would be overall faster then a Cobot and not needing anything beyond the cnc being programmed.
Adding that you'd still need in-position switches that the machine waits for to eliminate crashing the server. But still much simpler than a robot.

Says the guy who's never been in the same room as a robot arm...
 
It's fine if you make your own products in high quantity but most of us don't see the quantity and commitment from customers to justify/risk it on that. There's new machines for sale every week on contracts that "got canceled".
Check out the "Incremental" podcast. Both of the hosts run spindle grippers, one of them in a job shop environment. The other guy is more what you'd expect, higher volume of very limited parts.
 








 
Back
Top