"Yeah, I Need to Automate My Welding Process": A Roundup
, by Jim Ryan, 8 min reading time
, by Jim Ryan, 8 min reading time
If it has become clear to you that you really should automate your welding process, you have lots of thinking to do before moving forward with implementation. Here is some help from the pros: a roundup of posts by people who know all about weld cells, including at the end some on pitfalls that you'll want to avoid. This is a long post, so get a cup of coffee.
I'll give you a quick summary of each of the fifteen articles, but do find time later to click through and read them all. They will get you started with the right concepts and might save you a ton of money and frustration. We'll start from a very high level (how to automate in general) and gradually move down to the very close-up view (e.g., which welding wire to choose).
The View from 30,000 Feet: Automating a Manufacturing Process in General
Wauseon Machine and Manufacturing has advice for managers on how to automate a process in Industrial Equipment News. First off, you should be clear on the benefits of automation for your business and how it can be implemented for your specific goals. But in general, automating will, according to Wauseon, offer reduced operating costs, improved throughput, profitability growth, and low-risk (more predictable) ROI. Next, you should conduct an automation audit for your business's specific position and needs, because whether and how to automate are not answered by a simple algorithm or formula. Also, you should hire a system integrator, instead of going it alone without the help of an experienced pro. Finally, assemble a team of three leaders to manage profit and loss, operations, and engineering, respectively. Finally, develop a clear plan for your specific needs, taking into account relevant laws and regulations.
DIY Robotics has advice similar to Wauseon's, but there are interesting differences. Additional benefits of automation that you should become clear about include improved safety and quality control and better scalability. Also, DIY points out certain points you should bear in mind when looking at your specific needs in order to decide what and how to automate: bottlenecks and rate limits, labor intensity, task complexity, and safety issues. DIY suggests you get your data analytics team to help here. DIY adds that your workforce should be engaged in the implementation, and the implementation should be tested and tuned before deployment and monitored afterward.
Tim Garnett has further high-level advice over at Automation World. This time the advice comes from the perspective of a system integrator. He advises that the whole workday of the employees be looked at for places where they are obliged to do non-productive tasks (such as clicking through a dozen HMIs). Garnett offers lots of details about how to shift this work over to a more productive scheme. (When you read his article, you'll be able to tell that he's a systems engineer.)
The View from 10,000 Feet: Automating a Welding Process in General
Okay, now let's look at welding in specific. Just to get the ball rolling, Yaskawa has initial thoughts. Here again we see an emphasis on what needs, in specific, your manufacturing process in particular has and how automation can address them. Yaskawa even has an app for that (linked to in the article.)
Pemamek offers an introductory discussion covering the benefits of automating your weld process and also the reasons you might still be hesitating over whether to automate, such as concerns over whether your particular process even can be automated, the difficulty of programming, and the lengthy ROI time.
Jeff Henderson helps analyze your needs in a post at the Abicor Binzel Blog. He wants you to go through the following thought process:
At THG Automation, Zenobia Weigel has additional tips: do not overcomplicate the implementation, look at which of your processes are necessary and which are not, and think about the effect of automation on your staff. Weigel mainly has cobots in mind here.
At Mechanical Farm, Dhruv Mehta points out what might be obvious to some but needs to be said:
There are several types of robotic welding machines such as Gas Metal Arc Welding, Gas Tungsten Arc Welding, Thin Gauge Arc Welding, Plasma welding, Laser Hybrid Welding etc. The first step in determining the right robotic welding machine is understanding the workflow involved in making the end product and how a robotic welder can integrate into that workflow.
He also notes something you might not have thought about: What about your upstream processes? Are they going to be ready to produce the parts that the robot needs quickly enough for it? He says, "Making sure your upstream processes are in line with the performance of the robot is crucial as failing to do so is lowering the efficiency of the robot and ultimately annulling the investment."
Up Close: Weld Cell Implementation
A few articles take an even closer look at implementation. Codinter has a lengthy overview that introduces you to the basic ideas of automated welding (parts of the robot and weld cell (arm, power source, controller, JIG, sensors, etc.), "cobot", some history of robotic welding, etc. The article also notes several important points, such as the need to train some on your staff to program the controller and dealing with team members who may fear that they are being replaced by a machine. Chiming in with Weigel's idea of keeping implementation simple, the article notes that if you have complex welding processes and simple, repetitive ones, consider automating the latter and letting the humans keep control of the former. Codinter has more implementation tips - consider heat generation from the weld cells and consider division of labor among the robots so that their programs don't need to be frequently switched - so click on over and take a look.
Robotiq's Catherine Bernier has a two-part series on getting started that looks even closer at weld cell components. Part 1 is here and part 2 is here. (These posts, though still helpful, are ten years old, so unfortunately the videos in them are no longer available.)
Warnings in Advance from the Pros
You know that feeling "I wish I had considered a few things more carefully before I started down this path," right? Two articles offer assistance in this area. Bob Rochelle over at The Welder has eight mistakes that he can help you to avoid, such as eliminating your key weld personnel, underestimating the payload reach you will soon enough need, and expecting too much from the robots. Miller Electric Mfg's web site has an article that will help you keep in mind four crucial ideas: keeping implementation simple (we keep hearing that from these pros, don't we?), getting buy-in and input from staff, determining part mix and fit-up, and analyzing your workflow.
Details about Weld Cells
When you're ready to zoom in to the even finer details, take a look at a couple of valuable articles from Tregaskiss. One is here and the other is here. These concern factors such as choosing welding wire, reducing cable wear, and establishing reliable TCP connections.
Two points. First, if you can, attend FABTECH next month (September 2023 - or if you're reading this long after that, go to the next FABTECH.) If you're short on time, just spend one day at the four-day convention. Go to as many booths as you can, especially those of the aforementioned integrators, but also the exhibits of the weld cell manufacturers. Talk, look, chat, get business cards, etc.
And please do stop by our booth. We will give you a few minutes of very fun hands-on training with sensors and cables. You'll walk away knowing exactly how your robot will see the metal it welds. Also, when you implement your weld cell, choose components that won't burn out. That means choose Weld Dynamix sensors, cables, cable protection, and mounts. You will never know the frustration that has plagued weld cell teams for decades: burnt-out components, down time, reduced production. Our sensors, cables, and mounts are coated with ACTIVSTONE so that they are impervious to weld spatter. Our mounts are robust (to put it mildly) and our sensors' performance is excellent. Don't forget about our cable protection tubing. If you're reading this after already purchased and installed vulnerable cables in your weld cell, then cover them with our cable protection. Also, definitely cover your ethernet cable with it, as well as any other cable in your cell that isn't coated with ACTIVSTONE.
UPDATE: Robots.com chimes in with a few tips on building a weld cell.