The fact that a fully-configured arc welding robot can be purchased for significantly less than an experienced welder makes in a year is pretty impressive. Of course, the price of robot itself is only part of the cost of setting up an automated welding cell. Also, automation imposes a whole new set of constraints on a manufacturing process, the consequences of which may not immediately be obvious.
For example, when I worked as a process engineer in highly-automated foundry, I came up with a way to reconfigure a mold which made four castings per pour so that it could make six castings per pour. This was a big deal - the number of pieces per mold was directly related to profitability, and this was a 50% increase! It would have represented over $1 million per year in increased profit.
The only problem was that I couldn't get six castings to fit into one of the totes which was used to carry the parts to the automated finishing cell. (Actually, I figured out a way to get them to fit, as long as they were loaded a certain way. But the way the process was set up required them to be randomly dumped into the tote. Loading them in an ordered way would have required an additional robot).
With human beings, this would be no problem - just train the operator how to load the parts, or else just tell him to put three of the parts in one tote, and three of the parts in another tote. But the automated system was set up such that one mold worth of parts had to go in a single tote! Re-writing the code to allow parts from a single mold to be split up between two totes turned out to be such a daunting task that none of the automation engineers - who were extremely talented guys, some of whom have since gone on to start their own company - was even willing to touch it.
Still, with the price of robots coming down, and increasingly user-friendly interfaces, it looks like automation will be making its way into more small-to-medium sized manufacturers.
A base robot that costs less than a welder for a year would almost certainly have constraints on the number of applications it could serve. But for those applications where it fits, it would seem almost impossible to pass on, especially if it offers a long-reliable lifetime.
You make an interesting point, Ann, when you mention that robot solutions may work for smaller shops. I would guess that means the overall cost of ownership is less. It sounds like it is less not necessarily because the robot is less expensive but because the software is less difficult from a user point of view. Does that mean that set-up is easier? It also sound like the new software doesn't require the programming that previous software required. Is that it?
Rob, you're exactly right, the overall COO is lower. Not only does the entire package cost less, including the robot itself, but the software gives engineers two options: program it yourself in an IDE or use the much easier point-and-click type interface for configuration and setup.
Pretty cool, Ann. That whole point-and-click figuration and set-up is becoming more popular through the automation world. It's part of the whole plant-as-a-video game trend. Smart devices has helped the effort, and now apparently robots become an easier set up as well. This approach certainly keeps costs down.
I like your "plant-as-a-video-game" image, Rob. This same trend has come into machine vision during the last few years. Development software for vision apps is usually sold with both types: a GUI and a "real" programmers' interface for writing your own code.
It's interesting to see how advance robots, better pick-and-place systems, and improved cameras (machine vision) and camera interfaces are together working in concert to improve reliability on the production line. This is not a trend that's getting a lot of publicity, but when you go and visit plants, they're clearly getting good ROI on this stuff.
I agree with Alex that many of these production floor advances are not getting the "air time" that they should. Couple the improved automation systems with tighter integration with enterprise software platforms like PLM and MES and you have a recipe for manufacturers to get tighter controls, better visibility, and wring more efficiencies out of their production systems.
There is no doubt that automation delivers tremendous advantages to shops who employ it. My point is simply that automation brings with it a whole new set of manufacturing challenges. Is it worth it? Sure. But you are going to have to operate within the constraints imposed by the automated system, many of which you won't anticipate beforehand, no matter how hard you try. And if you have smart and innovative process engineers, they will constantly be going up against these constraints and trying to figure out how to "break" them in order to get even more out of the system.
Dave, I, too, was surprised to hear that the robot costs less than an experienced welder. Of course, it still doesn't do everything an experienced welder can do. And, as you point out so clearly, the complexities of the automation process can get in the way of designing a better system.
This is a great development, and it should wind up making automating even short production runs economical. Programming these robots will wind up being the one thing that is a cost item. I have done robot programming in the manual mode, which is an intensely exacting process. Moving to each position with the required precision is quite tedious. Model based robot programming is a whole different realm, from what I understand. But it would seem that it may be better setting up a welding path on an actual part.
William, your description of programming robots by hand makes it sound excruciating, and even more demanding and tedious than coding a machine vision application. Although this package is for the low end of robotic welding applications and therefore would require simpler programming, as Chuck points out.
If anyone has any specific questions pertaining to robotic welding, programming or any other questions I am a robotic welding programmer for a large company in the Midwest. I use Panasonic, Fanuc, and ABB robots currently, and experience with Epson and Kawasaki material handling. Worked with Vision for a month, but do not have much knowledge on that other than all of the problems we had with it.
As to options for small companies who want to start up with a robotic department, a great solution is simple... BUY USED! Robot cells are like cars, as soon as they step off the car lot they lose half of their value. The robots I work with range from 2003 all the way back to 1993. The ABB was our oldest robot and ALL do what we want them to do. We actually just sold our ABB which I believe should be in a museum being that it was sooo old! To back up the programs I had to use Floppy disks! haha
I am leery of any software that claims to replace an experienced technician.
I have seen software that allowed a programmer to drive a CNC so hard that the tool melted the aluminum rather than cutting chips.
I also saw a CNC machine boring a 9/16' hole through a piece of 1/2" aluminum without a center drill or a pilot drill. An experienced tool maker or machinist would cringe at this description. I was called in because they wondered why the spindle bearings were sloppy.
And I have seen robotic welding applications programmed by 'computer geeks'.
Software cannot, in my experience, replace an experienced technician, but software can make an experienced technician more productive. It seems only a technician can appreciate the value of the skills of another technician.
Optomec's third America Makes project for metal 3D printing teams the LENS process company with GE Aviation, Lockheed, and other big aerospace names to develop guidelines for repairing high-value flight-critical Air Force components.
A self-propelled robot developed by a team of researchers headed by MIT promises to detect leaks quickly and accurately in gas pipelines, eliminating the likelihood of dangerous explosions. The robot may also be useful in water and petroleum pipe leak detection.
Aerojet Rocketdyne has built and successfully hot-fire tested an entire 3D-printed rocket engine. In other news, NASA's 3D-printed rocket engine injectors survived tests generating a record 20,000 pounds of thrust. Some performed equally well or better than welded parts.
Focus on Fundamentals consists of 45-minute on-line classes that cover a host of technologies. You learn without leaving the comfort of your desk. All classes are taught by subject-matter experts and all are archived. So if you can't attend live, attend at your convenience.