What's new
What's new

What will the new disruptive technology be that rocks the manufacturing boat.

kpotter

Diamond
Joined
Apr 30, 2001
Location
tucson arizona usa
I went to a shop today that was highly automated they had very few workers and the ones they had were very skilled, I walked in and there was 6 machines running and no one in the room they were all sitting out side eating donuts and smoking. This is why no one is hiring anyone they dont need them. The owner of this shop told me his dream was to have a man and a dog watching the shop, the dog gaurds the machine and the man feeds the dog., something like that.
 
Robots will never be good at custom machine work.

Whats going to rock the boat? Well, these 3D printers seem to be causing a big fuss. But, until they figure out how to print metal: I am less than impressed.
 
Where the hell else would manufacturing be headed...no people, just machines. This is the future of lights out manufacturing! Buy stock in German Shepard's and Science Diet dog food!:eek:

Stuart
 
I visited a (Danish) company a year or so ago and was surprised to see that they mass produced items that I'd often bought myself and always thought were made in a country with cheap labour. Denmark isn't exactly famous for low wages :)

There was about 1 man for every 5 or 6 machines and more or less "all" they had to do (apart from set-up) was put in mile long coils of metal of various widths and thickness once in a while. Even packaging was done by robots at 90% of the machines.

On top of that almost all the machines worked around the clock (24x7x365) and if a machine stopped a phone call was automatically made to an operator who drove in and fixed the problem.

One of the things that surprised me too was that as good as all robots were thought up by their own design department.

I suppose the moral here is "Don't work hard, work smart" as it all seemed very relaxed.

Gordon
 
I was talking about the 3D printers that people are using to print plastic parts with. you know, lithography. Not printing paper.

Like the retards that used them to make guns.
 
Robots will never be good at custom machine work.

Whats going to rock the boat? Well, these 3D printers seem to be causing a big fuss. But, until they figure out how to print metal: I am less than impressed.

They print metal fine, have done for decades, it's just slow and the machines are more complex and expensive to run than the plastic types. Great for making mold cores with complex (but efficient) internal cooling passages that would be impossible to machine conventionally.
 
Yesterday I was at the Swiss Technology Network Zurich 2013 Exibition.

Here's a couple of stunners.

IMG_20130607_093719.jpg


and

IMG_20130607_091339.jpg


The first is the Robotic boy from ETHZ and friends, in the flesh so to speak.

The second are ETEL's micropositioning systems with 7nm (nanometer) repeatability! They are also increadably fast.

Prevasive technolgies were real-time ethernet, FPGAs, cameras with embeded FPGA vision systems, and centralization of controls and monitoring for networked factories.

I think that the accuracy in production is about to jump by an order of magnetude. Heidenhain's off the shelf stuff will measure down to less than 100nm, pretty cheap too.

As an example of automated manufacturing, maxon is now offering an online configurator for their miniature high-precision geared-head dc-motors. Once you order on this web page --> maxon DCX: Configure your DC motor and more <-- your made to spec. motor sytem will be delivered in ~11 days!

They had this neat fully integrated system with 8mm ball screw, or 4mm or 2mm with ceramic screw on display.

HERO_TEASER_PRODUKTE-UEBERSICHT_SPINDELGETRIEBE.jpg


I also heard that permenent magent motors will be replacing induction motors as they are more efficient with VFDs, and all motors used in the induction-motor role will in the future will be driven by VFDs. That is news to me.

Cheers
<jbc>
.
 
According to Back to the Future Part 2, in two years time, we're supposed to have flying cars, instant pizza hydrators, and hoverboards.

Dehydrated-Pizza-1316019655.jpg


Automated manufacturing is sweet, but the human race is way behind. In 10-15 years time, we'll probably have the same CNC machines, slightly faster with higher resolution screens. That's about it.
 
The big question I would have is this... At what point will this stuff hit the private sector in force? That is to say, when will the investment cost and reward converge, and perhaps provide a healthy profit?

Let's not forget, CNC machines were around in the 50's, but it took a generation for them to really take off. No one could afford to invest in them. Aside from that, nobody was trained to use them other than a handful of very high paid engineers and scientists. How many of these old, beat up machine shops are going to be spitting out widgets with this technology when it costs millions to buy it? What bank is going to give them the loan?

I think a few places might invest in a wide variety of this cutting edge stuff. The majority are going to continue to use methods that are slightly behind. The cutting edge job shops with millions of dollars in space age technology can't take all the work off the table, leaving plenty of scraps for the little guys to eat with their used market CNC machines, or even 20 year old manual machines.
 
Excellent pictures but I can help but think of the automatic shaving machine that had rotating razor blades mounted in a helmet.

When the inventor was asked if he realized that not all had the same shape face he replied "One shave and they will".

Gordon

How bad a cut are you going to get if its off by 7nm? That is 0.2756 millionths of an inch!

Feedback loops are getting really tight.
 
I dont think it will take as long as some of you think, some parts were dropped off to my shop from a cnc shop that I use and the look on those guys faces when they came into my shop was total amazement. They are in there 20s and have never run a manual machine in there whole lives. Their shop has no manual machines it is all cnc. I turned on the shaper just for fun and they were watching it and said how could you possibly be accurate with this machine, it has no positioning system other than dials. These guys are mostly self taught and there first machine was a HAAS vmc. Technology breeds technology I think it will speed up dramatically with in the next decade.
 
Parts get done one way or another, its about finding the best way for each scenario, although people like to assume everything can be done 1 way.

I'm quite sure too much Automation is part of why fanuc has always been 10+yrs behind on technology, by the time they get all their shit running straight and producing with nobody on the floor(check their factory tour on youtube), its outdated, but now that is set up it might as well keep running to pay for itself, and so a 2009 machine still comes with PCMCIA and about the same memory as a toaster. Not that others haven't always been behind as well, but some are worse.

Although there's certainly more flexibility in automation now it seems, it still takes the right work for it to produce ROI and hopefully not fall behind the competitor by the time its paid off and time to retool. It has caught many companies. The ability to adapt and change quickly is more important than ever.


Most of my work these days is 1 to 10 parts, lots is still manual stuff. In the end, it either makes $ and pays the bills, or it does not. No shortage of fancy everything shops going out of business.
 
No shortage of fancy everything shops going out of business.

The ballpark prices I got for Heidenhain LF485 scales and EIB741 interface were not fancy but instead surprisingly economical. The recommended resolution for these scales is 0.1um. Or 100nm. That is 3.94 millionths. The accuracy grade over 1000mm is 2um.

The commoditization of such improved measurement technology will effect manufacturing. Old scales 5um, new scales 0.1um. That will change things.

Imagine you've lived your life up to now with 20/100 vision. Then you get glasses.
 
I sometimes wonder if technology will actually hit a wall at some point ,look what happened to Concorde, 50 year old technology we can no longer afford.
 
I meant the fancy machine shops with all high end and automation and what not.

Though in that regard. Most technology and equipment today more than exceeds our needs. No doubt it has its applications or they probably wouldn't do it, unless its marketing and showing off as offering something better than the other guy, even if nobody needs it. Pretty common today. I still can't tell the difference between the quality of my 17yr old TV and the new ones a the store, some of which seem worse...

If they want to give more for the same $, that's fine. But what use is .1um on a manual lathe? or even on a cnc?

Does it matter if GPS tracks my location to 1cm instead of 1 meter? nope, pretty cool if they can do it though.

Same idea that we can make really fast elevators so long as we don't mind killing everyone inside.
 
The first is the Robotic boy from ETHZ and friends, in the flesh so to speak.

The second are ETEL's micropositioning systems with 7nm (nanometer) repeatability! They are also increadably fast.

Why people build robots that look like a person is beyond me. Cute but missing the point of what they do by a mile.

Fast?? as in 4000 IPM rapids and 60+ inches of travel?
High-res scales are nice until you need to count them at high rapid speeds. A design feature many don't think about. Fiber-optic connections to your scales to handle the bandwidth?

I worked on a "factory of the future" project.
Lights out, no people, then along came a company named Toyota which had a whole different vision how to mix manual labor with automation with a whole set of lessons on automation.
End of project. Very big money sent to the scrap man but many things learned about what you can do along with a bit of appreciation for the basic human brain.

Don't get me wrong, I love this stuff but a dose of common sense needs to be used.
Computers, robots, and the like are way below the brains of your below average brain dead human and there has been no major advances in the past few decades.

Lower cost electronics have helped some but we really have hit a wall here.
In the end you need software and robust design engineering to make this stuff work and while it looks simple at first glance it is not trivial or cheap.
Bob
 
Why people build robots that look like a person is beyond me. Cute but missing the point of what they do by a mile.

The robotic boy is a research project. If you miss the point then you are not seeing humanoid kinematics, man-machine interaction, materials science and a host of other developing technologies.

Fast?? as in 4000 IPM rapids and 60+ inches of travel?
High-res scales are nice until you need to count them at high rapid speeds. A design feature many don't think about. Fiber-optic connections to your scales to handle the bandwidth?

The stage was fast enough so that I was not able to photograph it without blur. Good point about the bandwidth, I did not notice if the interface was fibre. I very well could have been. But I don't think they are using scales. I think they position directly with the motors. I didn't ask but I assume that they have a similar stage in production for wafer lithography inspection. The travels were about 400mm x 400mm x 360° using linear motors of their own design and hybrid air / linear roller bearings.

The thing is, I was so flabbergasted at the 7nm repeatability claim I lost my ability to ask sensible questions about it.

I worked on a "factory of the future" project.
Lights out, no people, then along came a company named Toyota which had a whole different vision how to mix manual labor with automation with a whole set of lessons on automation.

Indeed, the patent date of the first safty-pin post-dates the development of the marine chronometer by about 100 years. It is easy to overlook the obvious if you are concentrating on the obscure.


Don't get me wrong, I love this stuff but a dose of common sense needs to be used.
Computers, robots, and the like are way below the brains of your below average brain dead human and there has been no major advances in the past few decades.

Lower cost electronics have helped some but we really have hit a wall here.
In the end you need software and robust design engineering to make this stuff work and while it looks simple at first glance it is not trivial or cheap.
Bob

Robust software design engineering? What is that? People with the skills for that usually do something else.


Cheers
<jbc>
.
 
I actually suspect that the biggest changes will be in social processes and in software.

20 years ago, you wanted a database, you bought it from one of a few vendors, you bought an entire machine to run it, and you were happy because it was way way better than other things you might have done instead.

Now, you can download the relevent software and run it on your laptop.

But as anybody here surely knows, debugging the design is often the costly part - and simluation software has surely come quite some way. Will we, in 10 or 20 years, see software systems that allow really complicated designs which always work as expected right off the bat? So when first articles from each process are assembled the result is production unit serial #1, rather than prototype #1 of #58? Some of that has already happened.

When that becomes more widespread (strong CAD/CAM in use by more people, just as open source SQL has databases in use by more people) that will be interesting.
 
Robust software design engineering? What is that? People with the skills for that usually do something else.

<jbc>
.

Unfortunately yes, but it is kind of important in automation so that you don't make 100,000 wrong parts running lights out or worse yet your robot kills a person once every couple of years.

I had a $600K+ automated machine that was quite happy to scrap 3,500 $13.00 inserts over the weekend running by itself. :angry:
The software was smart enough to call me if the machine shut down but was not smart enough to make sure the coolant pump was running. It just kept adjusting size chasing the grinding wheel and thermal cracking everything it ran.
A very distinct "oh-shit" smell when walking into the building on Monday and the clamping centers glowing red hot. Machine just cycling away like a happy camper.
Bob
 








 
Back
Top