Perfecting sheet metal fabrication bend

In autonomous fabrication, software algorithms identify parts suited for specific cutting and bending processes, assign routings, and write programs, all on the fly and in the background.

Manufacturing is, and has always been, a perpetually evolving industry. But the rate of change seems to have accelerated in recent years, along with a growth in orders. Keeping pace with increased customer demand has posed challenges for many manufacturers. It just seems like there are not enough fabricators to meet the growing need.

During the pandemic, the trade of the traditional fabricator shrunk dramatically as many decided to retire early or move into new positions that promised wage gains. At the same time, the industry has seen a growing trend toward high individualization and smaller batch sizes. Customers demand individualization of their end products and they want to see the status of their orders in production. This has spurred higher complexity in fabrication along with the need for continuous transparency while reducing batch sizes. Making improvements to individual processes is not enough.

Robots may not have made sense in previous shop floor environments, but recent shifts in production require another look at the potential role of automation in metal fabrication. One notable area where automation has produced gains, particularly for mid-sized manufacturers, is in the cut-bend cycle.

The Impact of Indirect Processes

A study conducted a few years ago by the Fraunhofer Institute and TRUMPF looked at how much time fabricators spend on indirect processes related to organizing and delivering parts to their next step. The study found that nearly 80% of production time is spent on those indirect tasks. This leaves only 20% of production time for value-added tasks—the moneymakers—such as cutting, bending, welding, and assembly. Fabricators looking to increase revenue need to reduce that 80% by automating and optimizing those nonvalue-adding processes.

The industry has seen only limited use of automation. Certainly, fabricators have used storage towers, load/unload systems, and bending automation for several decades. Because of the limited flexibility in existing systems, however, most of that automation use has been limited to high-volume, low-mix work, usually at OEMs or large contract manufacturers with predictable demand and highly standardized products. To date, the industry really hasn’t seen the wide adoption of complex, highly sophisticated automation systems in flexible environments.

Most North American fabricators work in environments with rapid and often extreme demand changes across different machines and work centers. Some orders have cut-only parts; others have parts that are cut and formed; still others require welding. Rapid changes in parts, demand levels, and job routings make it difficult to plan steady, predictable production runs to match customer takt.

Different part geometries, surface requirements, and edge-quality expectations can influence the technology fabricators select to make those parts. For example, a part with louvers and formed features could lend itself to punching. If a customer expects that no oxygen will be introduced into the cut edge during laser cutting, then the job might lend itself to high-powered fiber lasers. Or maybe an order contains parts with complex round-bending features, which could be best produced in a scratch-free panel bending process.

Legacy Technology and Shop Floor Layout

Historically, many fabricators with high-mix environments established a process-centric layout with laser cutting and punching machines in one department and press brakes in another. Such an arrangement can offer flexibility. For instance, if a part needs to be punched and then bent on a panel bender, material handlers can easily move the work between process steps using carts, fork trucks, or other manual methods. If the contract for a particular part is lost, the fabricator can fill that newly available capacity with additional work on the panel bender. Or, if nibble-free parts are required, the work can be shifted from the punching machine to a laser or combination machine.

This approach gets the job done and has worked well for decades in sheet metal fabrication, but it also has significant disadvantages. Downsides include lost parts, lost time due to manual material handling, and the cost of forklifts. However, these factors do not actually incur the biggest expense of doing business like this. Most lost revenue can be attributed to a reduction in working capital.

Think of it this way: Each company needs a certain amount of capital to do business daily. Just like a person, a company needs a certain amount of money at any point in time to do business. This amount of money is called working capital.

An integrated cutting and bending system—in this case, a laser, a punch, and two robotic press brakes—has towers that store raw stock and a buffer of work-in-process. That buffer ensures all machines can produce at higher capacity and achieve greater throughput.

In an environment where parts are cut, separated manually, and then transported by forklift to bending, the work-in-process (WIP) parts and, therefore, the inventory of parts between process steps can become immense. In a typical process-centric model, most parts are waiting 98.6% of production time. Such operations can have thousands or even hundreds of thousands of WIP parts waiting for the next step in production. That represents a considerable sum of working capital bound in unfinished parts on the shop floor. At some organizations, that value exceeds millions of dollars.

In times of rising inflation and interest rates, high levels of WIP can create a real financial burden. Imagine you’re a medium-sized contract manufacturer with a revenue of $20 million per year. Your profit margin is around 15% annually and your average lead time is around eight weeks. Once an order is placed, it takes about two weeks until the parts are cut and another six weeks until they ship. This means that parts sit as WIP for about six weeks. Multiply $20 million by 85% cost of goods sold and your parts being on your shop floor for six weeks ($20 million × 85% × 6/52), and you’ll find you have $1.9 million in working capital tied up in parts sitting on your shop floor. That’s already a lot of money, but if interest rates are high, and if you use a line of credit to pay for that working capital, the financial liabilities can add up quickly.

New Options for the Sheet Metal Fabrication Bend

How can you reduce those buffers and reduce the working capital between those different process steps? Really, there is just one option: You must shorten your throughput time and reduce the buffers between different work centers. This is easier said than done, as it is complex to steer the processes accordingly—especially considering the human element of moving parts between cutting and bending.

Compare the high-mix, low-volume fabrication plant to a major airport. Imagine scheduling different planes, some on time and others delayed, managing different departure slots while keeping an eye on the ever-unpredictable weather. This scenario is comparable to a flexible fabrication environment with ever-changing demands and customer requests. How do the air traffic controllers manage this problem? They rely on radar and other analytical computer simulation models to help make the best decisions.

In the past, fabrication management software could be complex and prohibitively expensive. Thanks to an acceleration in software development, this has changed. Complex software specifically for the fabrication environment is now available at a relatively low cost. Using advanced programming algorithms, a manufacturing execution system (MES) can help fabricators increase throughput, lower WIP levels, and reduce working capital.

Shortening the Cut-bend Cycle

MES production dashboards display and communicate, in real time, the most pressing order to the operator. Once the operator starts running a certain part, the machine communicates back to the system that the part is in process, and software realigns or reschedules priorities on the fly across different work centers.

All of this happens with one goal in mind: the reduction of throughput time and, with that, the working capital. When orders for new parts come in, algorithms quickly check which of those parts lend themselves to robotic bending by creating programs for the bending machine as well as the robot. Once that determination is made, it is easy for someone like a system control engineer to schedule and assign the orders to either a robotic or manual bending machine.

Digitization and the Sheet Metal Fabrication Bend

To accommodate time-consuming setups involving on-machine programming and test bends, traditional fabrication shops used large buffers between cutting and bending. Thanks to software and quick tool changes, setups take less time and buffers have shrunk dramatically. High-mix operations still need these buffers to accommodate varying cycle times, but modern technology has made most unpredictable setups a thing of the past.

Of course, these process improvements are only possible using complex algorithms and artificial intelligence. Thanks to ever-improving machine learning models, the algorithms become better at making those determinations. Truly, this is what enables integrated cutting and bending solutions that make sense in low-volume, high-mix fabrication.

Algorithms enable equipment OEMs to build highly complex machines with advanced, reliable part separation. Using these algorithms, computer models work in the background, proving the success rate to reliably separate specific part geometries; based on this, they recommend the machines that would suit best.

This setup connects a punch and panel bender, with a storage tower in between acting as a buffer. Today, algorithms help software identify which process (such as punching versus laser cutting or press brakes versus panel benders) best suits a particular job.

The same thing happens with automatic robotic bending. The algorithms determine which grippers are required to run the parts on the equipment as well as the correct tooling, and write programs for the robot and bending machines, on the fly and in the background, before the release to production. This technology enables integrated cutting and bending systems where a laser or punching machine cuts and separates the parts. Those parts are then buffered in the storage system prior to being requested by an automated robotic bending machine.

Directly connected cutting and bending systems without an in-between tower have existed for some time, but they work well only on certain parts, including those that spend equal time in cutting and bending. If one process or the other runs slower or faster, then either the cutting or bending machine will have to wait for the other to finish. This creates a bottleneck and underutilizes expensive production equipment.

Systems with a tower in between stand in contrast to that. By using a tower, batch processing becomes feasible, and each machine can produce the greatest level of throughput without being slowed down or slowing another machine down. If one machine is slower, the other can process parts from the buffer tower or from equipment outside the automated system.

Technology for the Next Generation of Fabricators

In the future, autonomous manufacturing will be the key to success, particularly as fabricators across North America struggle to fill positions and find young people who are excited about a long and steady career in fabrication. By establishing autonomous processes, fabricators also can expect machines to run more since they do not require operators to be present.

As the industry continues to move toward higher levels of automation, American fabricators will become even more competitive in the world market and build the foundation for the next generation. Today’s fabricators require people with new skills. Over the past 50 years, operators adapted from running shears to programming and running CNC equipment. The next generation will focus on making processes autonomous.