By Clare Goldsberry,
Associate Editor

Multiple robots can be coordinated for automated welding and fabrication.

Robots in an automated welding cell.

Automation can reduce time and costs-to-manufacture while improving quality and efficiency, but implementing automation for welding requires strategic planning, and evaluating the job and the expected outcome. And, it requires the right employees and a good robotics company or integrator if it is going to achieve the goals set out for it.

"A lot of companies are trying to maintain their business through the use of automation. Companies that previously didn't think about automating are thinking about it for survival. The rules of the game are changing, and we're looking closer at a lot of the small to mid-sized businesses and suggesting areas where automation might help," Harley Jacobson, account manager for welding for KUKA Robotics Corp. (www.kukarobotics.com), said.

However, before a company starts to automate its processes, Jacobson said there are a number of things to consider, especially if the automation includes a shop's first robotic welding cell.

Depending on the markets you serve, Jacobson said you need to look at the needs of the markets, where the business is within the market, and the impact of offshore competition including how much product from the market is coming from the Pacific Rim.

"The overriding factor is the Pacific Rim always looking over our shoulder," he said.

Additionally, Jacobson recommends looking at what the competition is doing: Are they automating? And will automating welding operations help you be more competitive?

One of the first things to consider is whether the parts you are considering for welding automation are consistent in size. Jacobson said that for robots to work well, the parts need to be consistent for a repeatable fit.

"One of the major challenges to implementing a robot on a job is high part variability. If you're producing parts by hand or using non-CNC equipment, you may be getting a lot of batch variation or non-repeatability. That would be difficult for a robot to handle," said Chris Anderson, welding technology leader for Motoman (www.motoman.com).

If a job shop is using CNC equipment to produce parts for welding, it's a sure bet that the parts are repeatable enough to automate the welding.

"Over a decade ago, a lot of job shops were using shears and break presses to produce parts, but today so many have invested in CNC equipment that we don't see repeatability issues much anymore," Anderson said.

Kuka's Jacobson pointed out that the number of jobs with which to occupy the robot is another important consideration.

"Are you doing one job? Two jobs? That determines how many sets of positioning equipment you need to keep that robot moving. Often, the cost of the robot is dwarfed by the fixturing to keep it going," he said.

Anderson added that part volumes are also an important consideration.

"With low batch volumes, if you have five of this part and 10 of that part, you may find yourself doing more in setup than in actual production, so volume is key. A thousand pieces would be a good number to consider automating a welding job, but it could be that 100 pieces would warrant automating, depending on how much welding there is on the part. You need some degree of repetitiveness with jobs so you can change over the fixture and do another job as well," Anderson explained.

Another issue, noted Motoman's Anderson, is the ratio of low weld content to high part handling or part size for a robotic cell.

"There should be a significant number of welds, for example 18 to 20 welds per part – that's pretty significant. With a small number of welds, often there's a lot of part handling and a small amount of weld time," he said.

If the part needs to be turned to accommodate the welds, a part positioner is required, and that will add to the cost of the system. Motoman produces robots with reaches to 10 ft. But, as parts get bigger, the system costs more. That is the reason that evaluating the part size is critical when making the decision to automate.

"If you're talking about something that is a meter cubed, that's one size. A part that's a meter in diameter and two or three meters long is another size," said Anderson. "Once you get bigger than four meters long and two meters in diameter, the system gets really expensive because of the tracks and other equipment required to move the robot."

Another consideration with respect to the price of the system is the number of pieces welded for each unit. A high quantity of pieces per each unit can make the positioning fixture quite expensive. For example, Anderson said, if a product has four parts and requires 18 to 20 welds, that's one thing, and it may be a good job to automate. However, if a part requires 18 to 20 welds and it has a dozen component parts, it becomes difficult to fixture.

Anderson suggested that more-complex welding jobs with a high number of parts to be welded in each component can be broken up into sub-assemblies, using the robot to weld smaller sections and then finishing it by hand.

"If you use your 80-20 rule and let the robot do 80 percent of the job and the other 20 percent is done manually, you can save money and complexity in the automation cell, and get better productivity," Anderson said.

This brings up a caveat to automating a welding job that Kuka's Jacobson addressed:

"If the part is too complex – if a human has a difficult time making the weld time and time again, then it's probably not a good candidate for automation. There's a lot that can be done out there, but there are jobs that I've looked at and decided that it can't be done with a robot."

Versatility also is an important consideration when choosing a robotic welding system. The answer to the question: How versatile today's robot systems are? is still evolving.

Kuka's Jacobson said that 10 years ago it wasn't evolving.

"If you had a guy welding a part seven days a week, 52 weeks a year, that was probably a good job to automate. Today, we're looking now at shorter parts runs, which changes the scope of things. Robots today are more versatile than years ago, and we're selling a 6-axis robot that gives us great versatility," Jacobson said.

Anderson added that while the robot is flexible enough to handle a variety of jobs in a production job shop setting, the key is to also implement an easy toolchange system, including fixture plates with dowel pins so that fixtures can be easily changed in and out of the automation cell to accommodate job changes.

Another consideration, and something that's often overlooked by companies looking to automate a welding job, particularly if it's their first foray into automation in welding, is system integration.

Through two separate divisions, Motoman offers both the equipment, through their hardware group, and systems integration. Anderson said what typically happens is that a customer might buy a standard welding cell, positioner, and safeguarding and be ready to weld.

"But the integration aspect is something the customer needs to address on his own. Often someone buys a cell and didn't anticipate how much work would be involved in system integration. You are better off contracting with a systems integrator, a consulting firm that specializes in this, or buying an integrated system from the get-go. You save a lot of time that way," Anderson added.

Jacobson said that KUKA does initial set-up for clients, and that there is some support that a client gets when they purchase a system. Beyond that, however, "the meter's running," meaning that additional services cost extra.

"We take a different view, but not unlike our competition, that as part of the package, depending on how the deal is structured, we have training classes in basic programming. It's not difficult, and we don't have to talk to robots in machine language, as ours is a PC-based system. So it's not all-together different than sitting at a computer. There are a few areas of finesse such as torch angle, and some others specific to welding, but the welder will have that knowledge, and can take that knowledge and apply it to the robot, so it's not really that difficult," he explained.

When it comes to the employees needed to help a company operate a welding automation cell, Kuka's Jacobson says that with the shortage of young people filling vacant slots in the welding trade, automation helps. But, welders will still be needed even though a company has automated certain jobs. At the Milwaukee Area Technical College, a certificate in Welding Automation Technology is being offered. With this new certificate, people are brought into the labor pool who know their way around welding and also are being trained in how to interact with a robot.

"These people are in high demand right now, and that's one place for manufacturers to go to find workers who have training and expertise in both areas. If companies have a welder on staff, he makes sparks, but he might also be a good person to send to school to get that Welding Automation Technology degree to learn how to program and operate robotic cells," Jacobson said.

Motoman's Anderson is Chairman of the Dayton, Ohio, AWS section.

"A thrust for AWS is training and education. Nobody wants to get into manufacturing anymore, so the State of Ohio is starting to get involved in manufacturing education. The need now is for students who want to get into these programs. The days of getting paid $18 an hour for pushing a button are over, and people need more education than ever to work in the factory," Anderson said.

Additionally, multi-skilled individuals are becoming increasingly valuable to manufacturing companies. "In larger companies, where welding plays a more important role in getting the parts out the door, a welding engineer is necessary, but they don't come cheap," said Jacobson, adding that usually a company can justify their salary in a couple of months.

In addition to engineering functions, they do tasks such as write the procedures on how things are done. "The staff you have on hand, if they've been welding manually for years, can help with automating the process. One place the programmer can gain expertise is to talk to the welder, find out how they're doing the specific job currently. People working on the manufacturing floor can be a wealth of information on how jobs are currently being done and can be extremely helpful in developing the automation."

How to determine if automating a job is right Simple steps for evaluating whether to automate a welding job from Chris Anderson, welding technology leader from Motoman. 1: Evaluate your parts and the current labor applied to it. How many semi-automated welders are you using currently? If you're using two to six welders for a job, that job may be a candidate for complete automation. If you're using more than 10 welders, that job is a certainty for automation. Crunching the numbers by looking at lease rates, and factoring those into a wage or dollars per hour, shows that a $75,000 to $150,000 expenditure in automation equates about $9.00 to $18.00 rate per hour (based on a 24/7, 50week year). "That can be very economical when implementing automation with semi-automatic welders. If you have simple parts with high volume, you can probably set up a robot and use a low-skilled operator to load and unload the parts. If you're doing more complex or heavier parts, you may want the cell run by a welder who can pre-tack and finish parts as he loads and unloads the cell," Anderson said. 2: Ensure that you have experienced welders on staff to program and operate the robot. Do not send your CNC operator to a welding robot programming class and think that he will be able to run an automated welding cell, unless he knows welding. "Robots are relatively easy to program, and we have a oneweek class here geared toward shop floor programming. But in the end, the robot has to weld, so someone with welding know-how will be most effective at applying it," he said. 3: Consult with the robot manufacturer or integrator on the automation of parts before you make the investment. They can provide some cell concepts and give you estimated cycle times. 4: Consider automated welding as a part of your overall growth strategy. Some shops invest in an automated welding cell knowing it won't be used 100 percent of the time at first, but they use it to promote a high-tech advantage and to get more work.