AI, Machine Learning, and the Future of Metal Fabrication

WhatsApp Group Join Now
Telegram Group Join Now
Instagram Group Join Now

On the first day of the 2024 Fabricators and Manufacturers Association annual meeting, held in Clearwater Beach, Fla., Jane Marks, a speaker and columnist for Forbes magazine, gave a recent look at the costs of computer processing speed. Pointed to an eye-opening chart. Decades

During the FMA’s annual meeting, a Navy SEAL leadership consultant floated an idea that seemed a little surprising to those who had never served in the military: decentralized command. “Not only does decentralized command allow you to grow in a role, but as a leader, it allows you to step back and look at the bigger picture. If I’m constantly making decisions for my team, I have no such method.

It was veteran Carlos Mendez, an adviser to the Texas-based Echelon Front. His insight went against the popular view of the military, which was shaped by movie scenes of sergeants yelling at subordinates. The reality is that soldiers can find themselves disconnected from central command, and can be in a world of trouble if they don’t have the training or authority to think and act independently.

Lives may not be at stake in Fab Shop, but livelihoods certainly are. Most metallurgical structures occur in high product mix environments. With equipment and software juggling hundreds of jobs, some unexpected variables are bound to throw a wrench into the workday. Fabricators work to minimize exceptions, but there will always be.always– There are exceptions

On the first day of the conference in late February, which was held in Clearwater Beach, Fla., Jane Marks, a speaker and columnist for Forbes magazine, presented a cost-tracking account of computer processing speed over the past few decades. Pointed to an eye-opening chart. “Processing speed today is about a hundred millionth of what it was in the 1970s. The fastest computers in 1993 could do less than 1,000 operations per millisecond. It’s gone up to over a billion operations. Every millisecond is.”

The extraordinary power of modern computing has created all sorts of AI tools, but they are not “total” solutions. They can write emails, design presentations, and automate some email tasks. They help immensely in a thousand different ways, but they only get you 80% to 90% of the way there. Humans still need to finish work.

This scene may one day reflect life on the shop floor, although we have a ways to go. As several customizer attendees discussed during the conference’s breakout sessions, the challenge is data. The conventional wisdom is that manufacturers are swimming in it—but how good Is that data? Machines and software occupy an incredible amount. But in most fab shops, not every machine is automated, and a lot has to happen before and after each manufacturing step. Instead of paper job travelers, operators now use laptops, tablets, even their phones, but they may still be manually entering work information into an ERP system. What exactly happens at the heart of a particular task between that initial clock-in and final clock-out is often simply not captured.

It sometimes comes as a surprise when manufacturers integrate Industrial Internet of Things (IIoT) platforms. These can show how little “real” uptime the machines have—that is, when the machines are actually cutting, turning, and welding, and good parts are actually being produced. Generally, this is a fraction of the time people assume. IIoT is showing improvements in low-hanging fruit (material staging, standard procedures and work procedures between shifts, etc.), but it’s also showing, no matter how dialed-in an operation is, exceptions. will exist. Compared to previous years, the discussions at the FMA Annual Meeting really focused on how best to manage them.

Lean principles came to the fore during the conference sessions — such as improving machine utilization but not at the expense of overall plant throughput. Lesser spoken negatives also entered the discussion. During a breakout session, Caleb Chamberlain, co-founder of OSH Cut (and fellow columnist for The Fabricator), walked attendees through the customer experience he and his team designed. The OSH kit does not print. In fact, there are no prints of it, which raised some eyebrows among the audience. Users upload design files directly to the OSH Cut website, which performs a manufacturability (DFM) analysis for the design. If there is a problem, the user can make changes and then upload the design again. From there, nesting, machine programming, and various other order prep tasks are automated.

OSH Kit is not the only fabricator to do this. A few offer similar services in the US, and in Europe there is a collection of “webshops”, 247TailorSteel, a Dutch operation, being the most popular. This model won’t work everywhere, but it raises big questions about what activities in the metal fabrication supply chain really add value, and where those activities (especially DFM) should be located.

Brian Steele, CEO of Cadrex Manufacturing Solutions and a panelist at the conference, represented the other end of the metal fabrication spectrum. After acquiring several plants, Cadrex is now one of the largest contract metal fabricators in the country. The company has also adopted software that runs factory-wide simulations, weighs different production options, then suggests what should work best.

Attendees at the Fabricators and Manufacturers Association’s annual meeting took a deep dive into what defines operational excellence, expertise, and future success: good systems, good data, and, most importantly, good employees.

All of this reflects the growing importance of software innovations, the best of which aim to weigh the effects of thousands of variables in high-product mix manufacturing and, ultimately, help skilled people make better decisions. Software will not account for everything, which brings out the importance of problem solving and decentralized command. The last thing FMA Annual Meeting attendees want is to lead an automated shop where software makes all the decisions and people mindlessly do what they’re told.

Operators should not avoid using new technology or change machine programs or tools because they prefer it. But they also shouldn’t run a machine program or task that doesn’t actually work. They need to be able to identify what really is an “exception”, then have sufficient knowledge and authority (supported by good systems and procedures) to act and act. No matter what the future of software, machine learning, and AI looks like, employee expertise and curiosity will remain a toolmaker’s key competitive advantage.

WhatsApp Group Join Now
Telegram Group Join Now
Instagram Group Join Now

Leave a Comment