Much talk, writing, and thought are being devoted to the concept of islands of automation. The idea is this: a fabrication shop is an archipelago of bits of automation. These islands could be an automated press brake or laser, a machine tending robot that can be moved from one machine to another, an automated laser welding unit—just about anything that contained and delivered automation.
We know why people are buying into automation:
- Speed (how fast—applies to machine actions as well as time between machine actions)
- Stamina (how long and how constant)
- Accuracy (how close to a given target/measurement)
- Precision (how close are multiple tries to each other—a strong factor in repeatability)
- Flexibility (perhaps it can do many jobs and is openly programmable, but automation can be tied to an idea—an example is an AGV that can only move along a wire planted under the floor and is in that sense inflexible)
- Teachable and capable of decisions/actions (think of the AI boom)
- And many more reasons.
Get into a more casual conversation, and often the answer to “Why automate?” centers around a decade-long (at least) struggle to find people to do the work. The lack of job applicants puzzles hiring execs from fabrication firms mostly because of the amount of automation we already have. It’s a much-improved work environment from the days of hauling/pushing/pulling stuff around with your body.
During this decade, we have seen great strides in automation, often tied to a machine tool—a press brake, let’s say. The rams are now either dual-geared for approach vs. bending speeds, or are adjustable in some other way. Tooling? Lots of options there, from super-high-speed shuttles that place punches and dies at just the right spot, to robot placement, with the tools located in the same plane, adjacent to the bend, or underneath the backgauge, or on racks standing tangentially to the front of the press brake. The tooling is even tagged (often using RFID).
Laser cutters have too many bits of automation to fully list here, but a cool recent addition is the ability to throw a remnant piece on a cutting bed and let the cutter figure out where (and in what configuration) to place the blanks. Automated nozzle replacement, collision recovery, beam manipulation—these are other, near-magical automated features.
And, in an important way, these “islands of automation” really are connected physically (with Ethernet) and via data (through management software as well as control/operating software).
However, the last processes to be integrated are the ones that exist between the machines. Moving material from one step to the next (from blanking to bending, for example) are right now the potential pontoon bridges between the islands. The giants of the industry—as well as startups—are populating this field with hardware and software as we speak.
Another area to connect the islands is in the robot/machine vision combination. It’s now possible to attack the longstanding problem of part sorting. The worst work, of course, is sorting very similar parts. However, the combination of robot/machine vision/processing software can very quickly sort parts that differ only in threads per inch, for example.
My prediction for 2026 is that much attention will be paid to welding. Automated welding has the advantages of speed and accuracy in the weld, totally manageable and preconfigured use of wire, and darn-near perfect repeatability. The ease of use is improving too; most of the time you set the start and stop points and you’re done. And yet, I do see a lot of manual setups—a time-consuming step. I think it is difficult even for a robot to make impressive speed gains in setups but I hope I am proven wrong with a complex demo soon.
One more point about automated welding: the welding department is an often-cited bottleneck, and is often the last step that you could call fabricating. It’s also the worst place in the plant for mistakes, mainly because it is the last step. Most of the value in the part has been put into it upstream, and any mishaps are more expensive because of the time and operations invested in it.
Once we address all the between-the-machines issues and the issues specific to the welding department, we can look at our work world as a place where materials come in to deliveries, parts go out of shipping, and we know exactly what’s happening during all the steps in between. Meanwhile, let’s appreciate the islands and work toward their connection.