The Business Case for Automation and Benefits for First-Time Users
The main goal of automated welding — and manufacturing in general — is eliminating variation. Once you drive variation out, you can drive your costs down and produce a higher quality product.
Roughly 86 percent of welding operations with 10 arcs or more (excluding those in the automotive sector) have not yet turned to automated welding. Of that 86 percent, we estimate that roughly half of those may never pursue automation. That still leaves somewhere in the ballpark of 40 percent of 10-plus arc shops today where automated welding may be an option.
We emphasize option for a number of reasons: Many operations have always made a profit with manual welding and will continue to do so. Many parts are just too large or cumbersome to work within an automated cell. If you’re not regularly repeating parts and seeking a high throughput of identical parts, it may not make sense to invest in the infrastructure that goes with an automated welding solution. Concerns over labor relations and misunderstanding of corporate motivations may also be a roadblock.
The repeatability of an automated system, however, opens new doors that companies thinking of exploring welding automation for the first time should consider:
- Addresses labor shortage
- Increases productivity
- Allows manufacturers to more aggressively and accurately bid work
- Increases quality of production and helps identify inefficiencies elsewhere in the process
The following article looks at key parts of the manufacturing process, details ways in which an automated system may improve your operation and may shed light on whether automation is right for you.
While training is a key element to making an automated system work, giving responsibility of the robotic system to your current welders ensures someone with an understanding of welding processes and applications is always engaged.
The labor shortage was a very significant factor driving people to automate before the market crashed. The problem still exists today, although less acutely. It is more of a chronic problem now. The baby boom generation is still retiring (and taking its collective knowledge with it), and we’re still not backfilling our workforce with young, trained welders. This is one of the rare fields in this financial climate where there is still a demand for labor, and not enough skilled people to fill it.
It is critically important that we clarify the effect of automated welding on the workforce. Automated welding exists not to replace the man, but to complement him and help him produce more. It’s a method of making sure that manufacturing can be cost competitive. If you’re cost competitive and profitable, there will be more jobs, not less. When you add automated welding to your operation, you’re going to be able to bid more work with greater certainty and you will know exactly what your costs are going to be. I sold one particular company its first robot in 1997. That company now has 24 robots and their total employment in welding has increased. They have been able to bid more work and bring more work in house because of automation.
The main goal of automated welding — and manufacturing in general — is eliminating variation. Once you drive variation out, you can reduce your costs and produce a higher quality product. Many robotic installations failed in the 1980s because the customers’ parts were really bad and they did not understand the processes upstream of the robot. If the processes upstream of the robot are not capable of producing parts that have minimal variation, it will be a problem when welding robotically.
Adding an automated welding system to your operation doesn’t just help increase your welding throughput, it helps identify variation elsewhere in your production line. Think of the robot as a constant — you know it will always perform the same work repeatedly. If you start to see problems in the way the pieces are brought together, you can then identify inefficiencies further upstream.
Today, the upstream processes are considerably better than in the past, and applying an automated solution is easier than ever before. Variables such as part fit-up, size differences and gaps have typically improved, which makes it easier to implement a robot. Robot technology has improved as well. Some robots today have the ability to do things such as adjust the tip-to-work distance and keep that constant over the length of a seam. It also displays torch angle, which is another process variable that can be controlled and removed from variability. In the past, a robot operator would pick a torch angle that looked right to him. Now, he can display that on the robot’s teach screen and it can be controlled like any other variable. Arc data monitoring — which keeps track of how the weld parameters are changing over time on each particular joint — allows the operator to export the data and examine it to see abnormalities that are occurring.
Man Vs. Machine
When considering whether to automate, you really need to look at your operation and what a human does well compared to what a robot does well. Manual operators are good at making judgment calls. A TIG welder, for instance, will position the electrode and play the heat against one side or the other. He’ll press the pedal harder or softer depending on how he wants to manipulate the puddle. He’ll manually dip the filler rod when it looks right to him. The TIG process is very heavy on judgment, and a manual operator is much better at adapting to these situations.
The robot is not good at judgment at all. The robot is very good at repeatability. You have to take judgment out of the equation in order to make a robot’s work repeatable. If you can’t do that, you may simply want to look into ways to make your manual operators more efficient.
For example, can a robot weld over a gap? Yes, but only if the gap is the same every time. If the gap is the same every time, I can put a weave in to go over the gap so that I don’t blow through the joint, or I can put in a low amperage stringer to cover up the gap first, then come back and throw some heat into it to blast everything in and fuse everything together. If only one part has the gap and the next one doesn’t, however, that’s a problem. Varied gap is the automation killer. If the whole part moves, a robot can typically find it with touch sensing. If the part curves or has some distortion, that can be found through seam tracking. A part with a ¼-in. gap one time and a tight gap the next time? That’s a judgment call, which the manual operator is better at than a robot.
Eliminating Waste and Rework
Automated systems create noticeably less spatter than the manual welder, and the process can be continuously refined to the point where it is almost eliminated altogether. Anyone who is spending time to manually grind a part afterwards to remove spatter is performing rework. The welding process is creating that situation where you’ve got to do more physical work to the part. That’s very expensive and a factor that can be vastly curtailed with automated welding systems. It’s also a cost that people often don’t take into account.
A general rule of thumb is that a welder running a robotic cell can produce three times as many parts as the same welder performing the work manually.
Manufacturers may be intimidated by the idea of adding an automated welding system: Will I have to rearrange my facility? Will we have to add another building? Will we have to add infrastructure to our existing building? In some cases the answers may be yes, but manufacturers now offer pre-engineered robotic welding systems that include almost everything you need to start welding. These systems can be dropped right down into existing workflows and put into operation with training and much of the basic tooling your manual welders are already using.
Implementation: Tooling and Fixtures
A major cost of implementing an automated system is tooling. It sinks projects. Many end users get nervous and decide to hire an outside contractor to do the tooling for them. This has pros and cons — obviously, a tooling and fixture specialist knows the ins and outs of numerous systems and may even be able to help you further streamline your operation. Many customers, however, have the wherewithal to build their own tooling and save a lot of money in the process. Tooling and fixtures similar to what you’ve always used in your manual processes often can get the job done. No one knows your operation and your needs better than your people. Tooling doesn’t need to be expensive to be effective.
Don’t underestimate the importance of training. Assume that you will send your operator to training, and then send him again in six months when he has experience with the system and is ready to learn advanced techniques. This will help teach the operator some of the finer nuances of the system that may have been lost on them during their first training. Robotic system providers will also work with most first-time users to do in-house training after the installation.
While every company is structured differently, pushing responsibility for the automated cell down to the operator is a key advantage. Even if a guy hasn’t run a computer before, I want a true welder running that system. We want someone who understands how the weld reacts to torch angles and positioning. You can typically train anyone to run the robot, but automated welding systems run considerably better when the operator understands the welding process and its limitations. Putting that welder in charge of the cell also upgrades his job and gives him a sense of ownership (which helps with employee buy-in).
The Business Case: Payback
In terms of return-on-investment, calculating payback on an automated welding system is difficult to do without examining each operation individually. A general ratio to use is 3-to-1: If an operator can make 100 parts per shift, that same operator feeding a robot will make 300 parts per shift. The best productivity improvement that I have personally documented is 4.5-to-1, while the worst I’ve documented is 1.7-to-1. That was a case where there were gaps and we had to slow the robot down to properly make the weld.
Using that 3-to-1 ratio as a baseline, it means that the robot is allowing you to do the work of three people with one. It is paying you in production what you would pay two additional operators. If you’re running two shifts, then the robot pays you the equivalent of four salaries in direct labor. If an operator with benefits costs you even $40,000/year in that scenario, the automated system will pay you $160,000 in the first year. This ratio tends to hold true whether you’re looking at a basic $60,000 entry level work cell, or if you spend $250,000 for a more advanced system. It’s very conceivable that it could pay for itself within its first year, and that’s not taking into account other factors such as improved quality and reduced spatter.
There will be people who look at these numbers and see it as eliminating jobs. To the contrary, this efficiency lets companies grow their operations and add more staff, more capabilities, and help them be more competitive in the market. One of the proudest moments of my career was when I helped a customer improve the cycle time on a part with a robot. My customer had lost this part to a Chinese fabrication shop, but with the improved cycle time they re-bid and got the work back. Efficiency creates growth, and that is the benefit that an automated system provides — especially taking into account the lack of skilled welders coming into the industry today. If you take the plunge, you will find that you’ll be able to sell more work, bid more work and get more work because you can weld more cost effectively.