They’re Taking Our Jobs! Or Are They?

Oct 18

The Pitfalls of Welding Robots & Human Job Security in Fabrication: an Op-Ed piece by Charlie Beech

Part I: Intro to the World of Welding Robots

As the season for Tech Expos comes to a close this year, my social media feed has once again been overrun by fanboys and computer nerds posting videos of welding robots doing pretty incredible things. If you aren’t familiar with this recent development in the welding technology world, welding robots are machines that are either stationary or mobile on a confined track and can lay perfect welds on metal without the “human err” element getting in the way of structurally welded perfection. I admit that these robots are badass creatures- they can operate on multiple different axes, lay beads that most of us welders can only dream of, and they won’t show up hungover 4 days a week (as is the stigma our party-hard trade gets stuck within most blue-collar circles.)

Most of the welding channels I keep up with love to promote this new wave of technology. Anyone looking at the welding world from the outside would tell me to brush up on computer skills so I don’t get replaced in the next 20 years. I have strongly disagreed enough times in various conversations that I decided to write this essay. I just hope my laptop doesn’t shut down halfway through writing once it realizes how hard I’m slamming robots in this piece. There is a big divide in tradespeople currently- those 50+ years old who have been doing this their whole lives, and the under-30-years-old crowd who saw the downfall of a traditional 4-year degree’s value in our youth and took the trade route instead. A common belief I have heard is the older group of us tradespeople are the ones pushing back against technological updates in our field. Well, as a 25-year-old welder, I’ll be the first one to tell a robot to frig off and get out of my shop.

Yes, these robots can do it all. As long as “doing it all” doesn’t require leaving the ~10-foot diameter of where the machine is set up. Fabrication is an extremely physically demanding job (unloading materials from trucks, lifting insanely heavy projects all over the shop, prepping large pieces of metal for welding, and choosing specific abrasives for cleanup once welding is complete.) I easily hit my 10,000 steps a day just by moving around our 800 square foot shop; I’ve never even met a robot that owns a Fitbit. These robots have no clue what the idea of “custom fabrication” is. The root of our field’s job security is that people need custom work- no 2 railings are the same measurements, and what one person loves the next client will despise. We, humans, are silly in how we still crave having the fingerprint of the artist in a project, despite also begging for perfection in a final product. We crave seeing artisanal mistakes and the way they are miraculously saved, being hidden by the overall impression of a final product. When your robot goofs up it is time to start over only after realizing you just burned thousands of dollars because your robot can’t hold an angle grinder without built-in thumbs to solve a quick mistake (stupid robot).

Part II: Money and Practicality

I will preface this next section by saying a fabrication welder (who in addition can grind/polish welds, keep track of various projects, do custom work, and clean up after themselves at the end of the day) makes an average salary of $30,800-$39,200 in Ohio in 2021 [Source 3].

The barriers to entry in owning and operating a welding robot are mind-blowingly expensive. To get your hands on a single welding robot unit be prepared to drop $75,000-$175,000 [Source 1]. Then, unless you’re Tony Stark with Jarvis’s guidance on your side, good luck finding an affordable local technician to set this machine up and get it running. In the past, I have paid an electrician $400 to have a single outlet repaired in my bathroom. I shudder trying to imagine what a technician will charge for setting up a massive, newly developed piece of complicated technology (even installing the 3-phase 240v power outlet for the robot to turn on is upwards of $1,000).

Say you have your shiny new hunk of metal assembled and ready to go. Who is going to run this machine while you sit in the office and cry after looking at the receipts? The cheaper route is to hire a computer scientist who can program your silly machine to run (these guys pull an average salary of $43,000-$93,500 in Ohio, and an experienced computer scientist worth their salt makes over $100,000 easily [Source 2]). I used to live with a computer scientist and after seeing our lifestyle differences, it is much cheaper to hire a welder. Trust me, if I were better with computers I would be wiping my tears with $20 bills after tough days and my back wouldn’t hurt this much every morning. I can work with my hands like a damn masseuse once I step into a fabrication shop, and am a cheaper hire to every boss in the world. Call it passion or technological retardation- either way, I’m happy to be doing the career I’ll have forever.

But you’ve already spent this much money on a cool robot, electrician, and shop space, right? Why not hire someone who can not only program this machine but analyze welds and understand what they’re looking at when a weld is wrong? Well, buddy, you’re dishing out $80,600- $106,000 in salary after hiring a welding engineer [Source 4]. A friendly group of people from my experience, but most without practical knowledge to fabricate something themselves. This group can talk electrical jargon, metal compositions, and polarity while sounding like the smartest person in the room. [True story- this summer I had a conversation with an electrical engineer I met at a party (yes, I was polite despite my stance in this article). He was asking my advice on how to get the robot to reach the end of the joint which was outside the robotic arm’s reach while still keeping a degree of travel that allows for shielding gas on the MIG wire, asking if I have run into this issue in the ‘real world’. “Why don’t you just unhook the welder and finish the last few inches yourself,” I replied. Mind. Fucking. Blown. He hadn’t seen the forest for the trees.] At the end of the day, your goal is getting something welded, not to rely on the robot for your shop work. I have nothing against the Welding Robot Overlord role that welding engineers play these days- again, they’re good people, but the creative solutions on how to make 2 pieces of metal into one bigger piece of metal can be solved quicker by fabricators (the cheaper of the 2 hires).

All of this being said, I do recognize the benefits that welding robots can play in the approaching decades. Repetitive shop welding on heavy-duty pieces of equipment, medically safe welds, or production to the largest scale can be passed off to robots. I, and many other fabricators, have no interest in doing these jobs anyway. Besides, once you pay a welder to make perfect medical grade welds on sanitary equipment they are earning in Ohio’s upper percentile at $60,000-$70,000 a year [Source 3] and you could pay off robot overtime for that much. So let the robots have the straightforward high-precision jobs that your local hungover chain-smoking welder shouldn’t have their hands on in the first place. As these machines become more affordable, and as the technology development advances towards a lower barrier to entry cost- and education-wise, we will see more industries where automated welding is welcomed. Especially in large factories where dozens of welding robots run constantly under people operating all 3 shifts of the day and machine settings hardly change. But my point stands for the custom fabrication industry- there is no room for training a robot for what many of us love to do every day for a fraction of the cost. And that’s the last I want to hear about it at parties. I drive this point home in Part III, outlining the dirty jobs that a robot just can’t handle.

Part III: What the Job Requires

It's raining outside and the deadline can’t be pushed back any further. A robot can’t show up, but I can. My raincoat is a lot cheaper than the custom weatherproof tarp it would require.

The mud is knee-high and the last few inches of weld require laying nuts-deep in the slop and finishing the job. A robot can’t get muddy, but I simply have laundry to do when I’m done with the work. Your workpiece is 12 feet long but you saved a few bucks buying the robot that only reaches 10 feet overall. A robot can’t move any further but I have 2 God-given legs and can waltz to the other end of the table and finish your project. You’ve run your robot 10 hours straight and blew the duty cycle on the machine for the night. Your robot can’t turn on until the morning, but I’ll have a beer and a Redbull and pull that 14-hour shift you’re asking for. Shop morale is low. Your robot doesn’t know any dirty jokes to make the guys smile but I have like 6 on the top of my head that will turn the mood around instantly.

(I have a great one about a penguin driving a car, too dirty for this essay but ask me about it next time you see me -Charlie).

Your robot will never come in hungover but I will. Okay, point for the robot on this one but it won’t have any great bar stories to justify moving slowly that morning.

You want to weld a ⅛” piece of sheet metal to a 1” plate. I can preheat that metal with an oxy-acetylene torch with the precision of an ancient Red Dragon that has been breathing fire to protect its gold treasure for the past 2,000 years, and weld that shit without keyholes and with absolute penetration that would make Ron Jeremy rollover with jealousy. I’ve never seen a robot attempt to preheat metal (yet).

And finally, you have a small shop in a run-down area of the city and are tiptoeing the line of being broke while credit card bills are piling up because dumbass clients won’t pay you upon completion of a job. I’ll keep showing up to build the next projects to earn us money (and offer to collect outlying debts with the help of a crowbar if you need it). If you bought a welding robot instead, it will just be taking up too much floor space in the shop to get anything done while the $100,000 receipt, along with your mental stability, is being questioned by both AmEx and your bookkeeper. I’m the one you would keep around.

Part IV: Conclusion

By this point, you are probably thinking “Charlie, you sound like a bitter old man. There is no use in fighting the advancement of automated jobs in the rapidly approaching future.” Sure, pretty soon your CT Scan machines, automobile frames, and airplane components will have been welded by a robot, and those “Built in America” stickers that once pointed to pride in a manufacturer’s craftsmanship won’t sing with the same weight they used to. Times change and technology advances with consumer demands, so one stubborn child’s opinion in this article won’t slow the freight train of society’s onslaught. But reader, I assure you that your custom railing, firepit, home decor, or table legs will always proudly have an artist’s fingerprint of a fuck-up disguised in a place you won’t notice until 15 years later. Automated custom fabrication is not a reality on the horizon, so you can stop sending me youtube videos and articles about what computer skills I need to brush up on ASAP.

Buy your local foul-mouthed welder a 6-pack and hug them, because in the world of blossoming technology we are a minority that builds amazing products you enjoy through methods immeasurable by computers- through unwritten trade skills learned only by listening to those before us, through years of knowing how to fix & hide our fuck-up fingerprints in everyday items, through showing up to work in thunderstorms and mud only to earn less than a robot that can’t walk or tell a dirty joke. However, we are still here, and we will always be fighting hangovers while fabricating magical metal items that you may never notice yourself using every single day. It is a thankless and dirty job, and I am among the group that would not have it any other way.