HAAS "Automation"

Been seein' this for some time...

Seems like most every artical aboot HAAS that has any amount of pics - shows a cell with a robot in it.

Let alone the name of the company is HAAS Automation!

They have done everything from putzy little lathes and mills to pretty large gantry machines, HMC's, and now even some big CNC "engine" lathes.

I have been watching to see when they develop an 8/9 axis full bore lathe segment or even if/when they will enter the Swiss market? But it has been mentioned that this likely is not done doo to power of the control. It seems that the HAAS control has maxed out at 5 axis. ???

Just wait 'till they git a new control!

So maybe this is one of the reasons that they have not entered into the robot market as of yet? (6 axis)

I can EASILLY see them selling a whole packaged lot to a LOT of smaller shops with a machine or two and a robot to tend them. IMO they would have a shoe-in on this whole trend that is soon to pass. 10 - even 5 yrs from now I think that you will see a lot diff "norm" than we have seen as of yet on the "job shop" level.

Thaughts?

-------

I am Ox and I approve this here post!

Hass has a few more problems to work out befor they start working on more axis, like (delivery) they wanted to be the lagest supplier in the world and got everybody using Haas machines but now cant keep up with delivery. Befor I left Continental Field Systems I ordered a EC-1600 Horizontal and it took 18 weeks for delivery. then i needed to order 2 HS-7's for an upcoming job and the sales guy said they only build 4 per year. when trying to keep all controls the same this doesnt help. even the 10" rotab i ordered took 2 weeks and it was in stock! Maybe Gene still has the lawsuit on his mind for tax evasion!!!!

I think Gene is more worried about getting buggered up the hoo hoo.

http://www.venturacountystar.com/news/2008/jan/15/haas-starts-prison-term/

Ox, I've wondered the same thing, just not with Haas, with Mazak, went to their little show/open house in December and they were running a couple of Fanuc robots. Mazak is all about their one and done thing, full automation, Blah, Blah, Blah, but they don't supply the full automation, wierd. They integrate it, but don't supply it.

Too big of a nut to chew since they are both(Haas and Mazak) at full capacity? Technology wise I doubt they are behind. Is it the controller, Mazak, buys Mits controllers and Haas controllers from what you just said are a little lacking? Does this make it unprofitable?


On the job shop front, me and my business partner have the goal to fully automate a job shop, horizontals with pallet pools, robots, robotic carts to and from the lathes. So far all we have is a little belt driven robotic pick and place arm with pneumatic fingers, $26 on E-bay, $130 for shipping.

How hard is it to build a robot, it doesn't have to be fancy, you don't need gigantic motors, PCs are cheap, and the hardware is cheap. You could even use a coolant proof button on the table to tell the external stuff when to do its thing. You have my mind rolling now, I might not get anything done tomorrow.

Probably one of the best articles I've run across in a trade magazine was a couple years ago about a company somewhere around Iowa that makes about all the hydraulic cylinder pistons, seal glands, and similar components for most of John Deere's domestically made cylinders.

These were all turned parts with no milling involved. They were running 6 Okuma lathes in a row, each with a gantry loader/unloader and an automatic horizontal bandsaw to feed each lathe. They were running 20hrs/7days with the remaining 4 hrs/day for machine upkeep. After a predetermined number of parts, each lathe would stop and signal one of the operators (2 ops for the 6 lathes). They'd remove that part, set it on a CMM sitting at about the middle of the line, and scan a tag for that job and machine. The CMM would bring up the verification program for that part, check the dimensions, and alter the offsets on that machine per the measurement results.

According to the owner, he selected the 2 man crews based on their ability to keep everything moving at the pace, and not on machining ability. Everything was automated to the point where their logistics ability was the main factor in maximizing production.

For aluminum automotive wheels, Chiron (German), Okuma, and Fanuc market a wheel production cell. The cell uses an Okuma wheel lathe, one of the super fast Chiron machining centers, and a robot and a vision identification system from Fanuc. They can run a mix of up to 4 different wheels and the vision system will identify each wheel and notify the lathe and mill accordingly. The robot is placed so that it loads from a pallet to the lathe, switches the wheel from the lathe to the mill, and removes the finished wheel from the mill and back onto the pallet. That thing can do all the turning and millwork on a wheel in just under one minute.

Obviously, both these are full bore dedicated production operations, but the significant thing about both is that they use a collection of standard machinery instead of custom built machines that can cost tens of thousands of dollars to retool when there's some design change to the product, IF they can be retooled and made to work at all.

For a long time, I've thought job shops could gain a lot of productivity by running as many jobs as possible on multi-pallet horizontals. A friend of mine runs a recurrent aluminum job, sometimes on an Okuma vertical and sometimes on a Mori horizontal. The vertical has 1400+ ipm rapids, a 15K spindle, and a scary fast toolchanger, but the part has to be repositioned to accomodate a hole at an angle midway thru the cycle. They can hit $90/hr pretty consistently, but its a real strain to gross $100. Same job on the Mori will consistently pass $350/hr because the indexing capabilty gets rid of the need to reposition the part, and the 2nd pallet gets rid of lost time for changing out parts and blanks.

He mentioned a while back that if the Mori had a 10 pallet pool, they could run the required quantities lights out most of the time by just loading up the machine at the end of each day and unloading the next morning. Of course a 10 pallet pool would also allow you to have 2 pallet setups instantly available for each of 5 different jobs with near zero switchover time.

I really don't think most people can grasp just how many parts can be cranked out of even a 2 pallet horizontal as compared to a vertical unless they've stood there a while and watched both machines making the same part. Regardless, it seems there's a lot of hesitation to even consider a horizontal on the part of many folks. We just can't get Bridgeports and the idea that parts are supposed to lay down, out of our heads

I run one part that really begs for some sort of automated loading. A spring ejector would take care of the unloading. The cycle time is less than 15 seconds, but with changeout time included, about 100 pcs/hr is all you can get, and you're chained to the lathe continuously. I've got an old Fuji 4 axis hydraulic chucker that's perfect for running the parts. Job itself only needs one slide so I've beat the idea around to somehow automate the loading via the 2nd slide. As it is, I run out several hundred of these once in a while, and they sell out fast. If the loading was automated and I kept a consistent stock of them, they'd probably net me $1000/wk or more for an hour's work per day. Meanwhile, I spend my time working on stuff that looks like it oughta net $75 but turns out more like $35. I ain't figured out whether I'm dumb or lazy, or both, but the thought of some R2D2 thing loading parts while I catch up on some golf sure sounds pretty damn good to me

Metlmuncher, I don't know what your making, or what machines you are running, but, I agree. If I could load up 2 pallets and sit down for half an hour instead of changing parts, i could get a lot more done.

Automating a job shop is a lot harder than automating full on production, and as you were talking about would require non-dedicated machines and flexible everything. I think one $500k horizontal with a pallet pool could out produce a shop with 10 verticals running 40 or even 80 hours a week if done properly and with a hell of a lot less people. I'm all for bringing jobs to America, but I'm so sick of dealing with idiots.

The Fanuc plant has one of the most impressive setups around. Next to nobody works there but it produces thousands of motors/amps/controllers a month.

Bobw said:

How hard is it to build a robot, it doesn't have to be fancy, you don't need gigantic motors, PCs are cheap, and the hardware is cheap. You could even use a coolant proof button on the table to tell the external stuff when to do its thing. You have my mind rolling now, I might not get anything done tomorrow.

Click to expand...

Scara and cartesian robots are simple enough. Easy to make, easy to program. Not sure why manufacturers haven't sold those on the cheap for parts loading. Maybe hard to adapt to current machines like turning centers due to clearance issues...

Ox said:

putzy little lathes and mills

Click to expand...

Hey, "Slick. Sport. Champ. Chief. Ace. Tiger. Slugger. Hot Rod. Killer. Big Guy,
BuckWheat..... Mo"! Don't be calling my Machine no little condescending names. Geez....... Putzy? Putzy....... What kind of name is that to call my fine little lathe and mill.......

Ox, I've wondered the same thing, just not with Haas, with Mazak, went to their little show/open house in December and they were running a couple of Fanuc robots. Mazak is all about their one and done thing, full automation, Blah, Blah, Blah, but they don't supply the full automation, wierd. They integrate it, but don't supply it.

Click to expand...

FANUC Robot is one of the largest around. Mazak has had FANUC robots available on their systems for many years. What do you mean they don't supply it? By integrating it and offering it, they are supplying it..... they just don't build the components or the architecture..... OR do they??? Fanuc Robotics do use Mazak machines in their factory.

Automating a job shop is a lot harder than automating full on production, and as you were talking about would require non-dedicated machines and flexible everything. I think one $500k horizontal with a pallet pool could out produce a shop with 10 verticals running 40 or even 80 hours a week if done properly and with a hell of a lot less people. I'm all for bringing jobs to America, but I'm so sick of dealing with idiots.

Click to expand...

You hit it there. But in this sense, you are still offering job(s) in the US, although with less people. But theres a flip side. There are no "people" with skills, and even less with motivation. This is with many industries in the US. You hire people to get them working, but they don't want to put anything into it to learn and excel. The "laziness" and the "I want it on a silver platter" attitudes have turned many things upside down. But this is a whole different story....

Metl hit it too but .... don't discount the verticals either. Most people don't realize that you can build a mutli-pallet FMS for verticals as well. Mazak has had this for years as well and I've seen a reconfigured Mori SV system this way. It's VMC set up like a HMC cell with multiple spindles, pallets and loading stations. Another neat set up is that several companies also can configure different machines in the same line. For example, on the same pallet line you could have a HMC, a 5 axis and a Multi task machine.

Robotics and upper end automation has been around for a long time and the concepts even longer. It's only now that the consumer (machine shops) are more actively looking into this. The truth is, if people were more receptive to new ideas and technology back when this stuff came out, many shops would be much further down the road about capital equipment. The industry giants were ready to change and wanted to invest in the technology..... most of "us" were not.

Muncher:

Don't know the dsign of your part - but it sounds like something that could be set up in a B&S, Davenport, or Acme "chucker"?

There is a 1" Acme Chucker with new[er] bearings and a mag loader on e-bay rat now if it would werk for you.

I have converted one of my four bangers to and chucker and back to bar for that type of job. (back in the day eh?)

------

I am Ox and I approve this here post.

Five thoughts:

First - why does automation require a "robot" meaning something with an arm, as opposed to some other material load/unload system. I've seen a DMG installation south of Seattle with two large 5-axis "horizontal" mills, both with pallet loaders. The pallets are fed by a rack system (fastem?) The machines can each store a lot of tools (200?) When fully running (may be by now) they could load all the pallets with all sorts of different parts and let the thing run unattended over a weekend. None of this looks like a "robot" as in a Fanuc or Kuka sort of thing. Of course, there are lots of real common robots that don't look like robots (dishwashers, etc.)

Second - I wonder how much intellectural property (patents) or various license/royalty bindups affect the industry. Do Fanuc or Kuka or Motoman have enough patents, etc., to keep Mazak etc at bay? (Seems unlikely, but...)

Third - I have often wondered if the answer wouldn't be one of those huge lathe/mills with a very large bar feed. Feed, oh, 6" aluminum bar in one end and turn/mill the entire part. Next! Of course, the part has to be long and skinny compared to a mills work cube, and the machines are huge and costly.

Fourth - there ARE highly motivated people - but it seems (from what I read) that nobody in manufacturing knows how to attract them. A highly automated shop might have an advantage, in that you'd be trying to hire one $120k per year automation wizard, rather than 10 $12 hour button pushers.

Fifth - there's a lot of talk about jobs lost in manufacturing. Lots of blame directed to "offshore" etc - but of course, the number one loss of jobs in agriculture, manufacturing, etc. has been vastly greater productivity via the use of machines. This trend has gone on for well over a century. Probably, the first job shops that can master this, win.

Bobw said:

Ox, I've wondered the same thing, just not with Haas, with Mazak, went to their little show/open house in December and they were running a couple of Fanuc robots. Mazak is all about their one and done thing, full automation, Blah, Blah, Blah, but they don't supply the full automation, wierd. They integrate it, but don't supply it. Too big of a nut to chew since they are both(Haas and Mazak) at full capacity?

Click to expand...

Many companies Mazak/Haas supply require a specific manufacturer of robot. For instance, we use Kawasaki and Nachi. Those are the only two makes of robot we buy for our US facilities because of parts / training / programming / automation interface / custom operator interfaces we have developed, etc. Mazak will supply whatever brand robot you specify, they aren't tied into Fanuc. Have no clue about Haas, because we don't, and I don't foresee we will ever, own any.

Bobw said:

How hard is it to build a robot, it doesn't have to be fancy, you don't need gigantic motors, PCs are cheap, and the hardware is cheap. You could even use a coolant proof button on the table to tell the external stuff when to do its thing. You have my mind rolling now, I might not get anything done tomorrow.

Click to expand...

They aren't as simple as one would think. The operating system is complex. The precision of assembly is critical, and it takes stronger components than you think.
A robot is rated by the payload it is capable of moving and stopping at full speed, at it's maximum reach. So, if you have say a Kawasaki FS30L, you are able pick up and to move a 30kg (66 lb) payload at 2269 mm (89.3" = 7.4 ft) horizontal, or at 2799 ( 110" or 9.1 ft) vertical and move and stop it at 160°/sec (JT1). And place it with a repeatability of ±0.15 mm (±0.006")

http://www.kawasakirobotics.com/PDFs/FSeriesREV06-06.pdf (for reference)

These things are pretty incredible and extremely flexible in what they are able to do, and what you can program them to do. While most types have a 'simple programming mode", where you simply move it to where it needs to be and teach it, they also have a very powerful, feature rich advanced programming language where you can do much more complicated tasking and it really increases the functionality available.
While we have many varieties of parts, by using the advanced programming, you can basically macro the program so that you only need input a few variables and the robot can calculate where it needs to go and what it needs to do based on one or more fixed, physical reference points and input signals.

I design, install and program 20-30 of these robot cells a year. The programming can be as simple, or as complex as you want. Interfacing to the machine is fairly simple for most applications, there are only a handful of signals really needed. Access to the machine plc is a must for most applications though.

Anthony

Bryan - #1 = it doesn't hafta be like a robot. More specialized fixturing will Shirely operate a lot cheaper once running, but it is likely only gunna werk for one family of parts.

A 6 axis bot is extreemly generic in design and function and can be [somewhat] mass produced at a co$t effective price and used in many diff applications. A bot Shirely aint gunna replace a barfeeder!


-----

Tony - But there are a LOT of HAAS shops out there that have never ran a bot before, and in the near future many of these shops will start to entertain the idea. If there was a unit available from the same manufacturer that they are already comfortable with - I would hafta think they would git a good foot in the door in the industry real quickly.

Not stating that those that have been using them would entertain the idea of switching to them for the forseable future. But those that don't already know how Brand XYZ operates and would be learning from the ground up anyway - I see no reason that they wouldn't keep an open mind to a HAAS bot.

???

---------

I am Ox and I approve this here post!

bryan_machine said:

Five thoughts:

First - why does automation require a "robot" meaning something with an arm, as opposed to some other material load/unload system. I've seen a DMG installation south of Seattle with two large 5-axis "horizontal" mills, both with pallet loaders. The pallets are fed by a rack system (fastem?) The machines can each store a lot of tools (200?) When fully running (may be by now) they could load all the pallets with all sorts of different parts and let the thing run unattended over a weekend. None of this looks like a "robot" as in a Fanuc or Kuka sort of thing. Of course, there are lots of real common robots that don't look like robots (dishwashers, etc.)

Click to expand...

First: In a high volume dedicated production environment, it may not make sense to use a robot, some other automation device may actually be faster. BUT... If you ever have a change in the part design, or change the part, you are stuck with a dedicated automation design that won't work anymore. With a robot, you simply change the end effector and reprogram and off you go.
Second: The cost of dedicated automation will be the same, if not more expensive than a robot, and the dedicated stuff will almost always have many custom fabricated parts that you must maintain repair inventory for, or just be down while you make another. The only custom part on a robot will be the end effector, and in most cases, this is made of off-the-shelf components, less the actual gripper pads, or other part that actually touches the component you are moving and are considered wear items (tooling) anyway.

While the end effectors I have designed for our application are custom, they are standardized per operation across many machining cells. The differences in the designs are based on hard operation criteria, such as the design of the machine the robot is loading, clearance issues, part orientation, etc. They also use standard components, all the grippers are the same brand and model, etc.

Another advantage to robots is that if you should happen to have one break down (very, very rare), unlike dedicated automation that you would just have to rebuild on the spot, a robot can simply be switched out with another and fixed offline. We can change a robot out in less than 4 hours on even the most complex applications, complete and running production.
This does take advanced planning in the initial implementation stage, though.
Almost all of our physical electrical interfaces are identical, as far as connections go. It's a simple plug & play job to swap one out.

bryan_machine said:

Second - I wonder how much intellectural property (patents) or various license/royalty bindups affect the industry. Do Fanuc or Kuka or Motoman have enough patents, etc., to keep Mazak etc at bay? (Seems unlikely, but...)

Click to expand...

See my other post about why they don't build their own.

Ox said:

Bryan - #1 = it doesn't hafta be like a robot. More specialized fixturing will Shirely operate a lot cheaper once running, but it is likely only gunna werk for one family of parts.

Click to expand...

Hang on to your horses there a min..... That isn't really the case. We have had Kawasaki robots since 1994. They run 24/7, and they average less than 1 hour of unscheduled downtime per year. That is by far and wide, less unscheduled downtime than any other piece of equipment in the buildings, including all of the actual dedicated pick and place / specialized automation.
And that dedicated automation will most likely cost the same or more than a robot.

Ox said:

A bot Shirely aint gunna replace a barfeeder!

Click to expand...

Wanna bet? It most definitely can. The most likely application is to feed the bar feeder and feed either a second op, or palletize the finished parts.



-----

Ox said:

Tony - But there are a LOT of HAAS shops out there that have never ran a bot before, and in the near future many of these shops will start to entertain the idea. If there was a unit available from the same manufacturer that they are already comfortable with - I would hafta think they would git a good foot in the door in the industry real quickly.

Not stating that those that have been using them would entertain the idea of switching to them for the forseable future. But those that don't already know how Brand XYZ operates and would be learning from the ground up anyway - I see no reason that they wouldn't keep an open mind to a HAAS bot.

Click to expand...

The problem is that you already have many manufacturers of robots that know what they are doing, have already developed the software and hardware, and they are as reliable as can be. It would be a huge, huge investment for HAAS or any other MTB to develop a robot, then you have to have some sales reason as to why the HAAS / other mtb robot is better than those that are already on the market, and the truth is, they won't be, and most likely won't be as good, especially the first generation. Kawasaki, Motoman, Fanuc, Nachi, Reis, and the other major players in the robot industry have a huge advantage and many, many years head start.
I don't see it happening.

metlmuncher, why don't you equip the second turret with a few bar pullers? You could then index the turret, grab a part, and repeat until all stations are full, then pause the cycle to unload the finished parts. There are plenty of simple, purpose built, end effectors you can make. A bar puller is just the most universal that people are familiar with. You could take a piece of plastic, slot it 6 ways, put a o-ring groove in it, with silicone o-ring, and use it to grab the ID of the part and hold it. Basically a plastic inside collet with some grab to it.

Fourth - there ARE highly motivated people - but it seems (from what I read) that nobody in manufacturing knows how to attract them. A highly automated shop might have an advantage, in that you'd be trying to hire one $120k per year automation wizard, rather than 10 $12 hour button pushers.

Click to expand...

Bryan makes some good remarks. This one though isn't actually the only reason for the problem. Geography plays a role for the cost of living relative to the local market rate of jobs... therefore the number of motivated people to learn a blue collar trade may not be as high as other areas. This also makes the attraction tough. And there are many other reasons as well....

I do agree with Automation not having to equate with robots necessarily. This is a wide open subject and to be honest, I'm glad to see that there is actual discussion amongst the many and not us "selected few" that have been incorporating it for over a decade (and longer for several places).

KEEP IT COMIN' !!!!

psychomill said:

I do agree with Automation not having to equate with robots necessarily. This is a wide open subject and to be honest, I'm glad to see that there is actual discussion amongst the many and not us "selected few" that have been incorporating it for over a decade (and longer for several places).

Click to expand...

I've suggested before that there should be a forum here specifically for automation, as it is an increasingly important part of the machining trade.

metlmunchr said:

I run one part that really begs for some sort of automated loading. A spring ejector would take care of the unloading. The cycle time is less than 15 seconds, but with changeout time included, about 100 pcs/hr is all you can get, and you're chained to the lathe continuously. I've got an old Fuji 4 axis hydraulic chucker that's perfect for running the parts. Job itself only needs one slide so I've beat the idea around to somehow automate the loading via the 2nd slide. As it is, I run out several hundred of these once in a while, and they sell out fast. If the loading was automated and I kept a consistent stock of them, they'd probably net me $1000/wk or more for an hour's work per day. Meanwhile, I spend my time working on stuff that looks like it oughta net $75 but turns out more like $35. I ain't figured out whether I'm dumb or lazy, or both, but the thought of some R2D2 thing loading parts while I catch up on some golf sure sounds pretty damn good to me

Click to expand...

Think VMC swing arm tool changer. Feeder tube / rack at the one end, out/in and swing motion. Use pneumatic cylinders for both swing and extension/retraction. Simple, efficient, fast and reliable. Spring ejection to a parts catcher as you noted. Swing arm for loading only, and you still have the second slide available for other work.

Tonytn36 said:

I've suggested before that there should be a forum here specifically for automation, as it is an increasingly important part of the machining trade.

Click to expand...

IMO Any thread that would/could be posted in an "Automation" board could be posted here in CNC - No? Same people looking eh?

------

H Ross in '08!
Now more than ever!