Metal fabrication in Big Tech’s backyard

Martin Sullivan meets with Will Bravo to discuss the day’s production at Inland Metal Technologies, a custom precision fabricator operating in one of the most expensive places to do business in the country. Images: Inland Metal Technologies

Walk through Inland Metal Technologies, and you’ll see precision sheet metal automation on full display: a punch/laser combo with automated part removal, an automatic tool change brake with sheet followers; robotic welding cells; the works. Thing is, this isn’t Illinois, Georgia, or Wisconsin. It’s in Hayward, Calif., in the heart of the Bay Area, one of the most expensive places to live in the country.

Apple’s space-age headquarters sits less than an hour away in Cupertino; Googlers work in nearby Mountain View. Tech workers demand, and receive, extraordinarily high salaries, which in turn boosts the cost of real estate. Most manufacturers have moved to far-out suburbs or, for that matter, other areas of the country—but not Inland. Despite its name, the custom fabricator has not moved inland.

Inland has been in its current location since the early 1990s and has no plans on moving. Why? According to Martin Sullivan, COO, the answer has to do with Inland’s business model, customer mix, and approach to metal fabrication technology and automation.

“The short answer is, we go up market. The big costs are labor and real estate,” Sullivan said, “and so, we minimize their impact by doing harder work. Easy and less critical work can be done in a cheaper area, but it’d be difficult to replicate our capabilities anywhere else.”

Automation, Margins, and the Value of Labor

Sullivan emphasized that minimizing the impact of labor doesn’t mean automating all manual processes away. To the contrary, work that’s easily automated likely wouldn’t make business sense in such a high-cost area, at least in the precision sheet metal sector. Such automation often requires a lot of space. Jobs that are easily automated often end up having lower gross margins, especially as more shops adopt a particular technology. That in turn drives up the impact of labor. Yes, labor costs might be minimal, but the little labor costs that remain can make the difference between a job making or losing money.

The trick is to automate difficult jobs that few competitors are willing to tackle. To illustrate, he pointed to the company’s robotic welding cells and vast collection of weld fixtures. Conventional wisdom states that a job needs a certain level of volume to weld robotically. He pointed to an operator placing components into a fixture design with laser-cut components alongside the typical toggle clamps and stops.

“It doesn’t cost us very much to robotically weld smaller quantities,” he said. “And we apply it to very complicated work. If we approached it conventionally, each job would take a lot of skilled labor, but instead, we’ve created our own very fast setup process to make it much easier to scale.”

Inland has over three dozen custom robotic welding fixtures used for weldments ranging from small sheet metal parts to frames weighing over 1,000 lbs. The company also has strategic staging, so operators can quickly call up a program and load the fixture before the robot commences welding.

“For most parts, we make custom fixturing. For others, we could use [off-the-shelf] modular fixturing. That way is certainly easier, but you’ll always have long setup times, and your surge capacity is always capped by how quickly you can hire and train, assuming the talent required is even available when your customers are ramping up.”

If an operator continually had to fuss with every job and adjust every fixture, robotically welding such a small quantity wouldn’t make business sense—especially with the talented MIG and TIG welders Inland has on the payroll.

Alfredo Vazquez and Martin Buitron talk about the upcoming job that will flow through Inland’s press brake with automatic tool changes, then downstream to laser welding.

The trick, Sullivan explained, is in the upfront engineering the second a job is landed. “Our initial design-for-manufacturing assessment is extensive. It often includes shop floor managers to ensure no upstream process is causing their downstream work centers undue hassle. We also always ask, ‘How do we get this on a robot?’ Only if it’s completely unreasonable do we acquiesce and process it as a normal shop job.”

How do they develop their engineering capabilities, exactly? One way is to relate innovations demanded by one customer and apply them to others. To illustrate this, Sullivan recalled work Inland did during Tesla’s early days, when the EV maker operated out of an old 20,000-sq.-ft. warehouse in the Bay Area. Inland robotically welded battery boxes and other components with sophisticated fixtures designed for the higher volumes common in the automotive business.

“We were able to take the ideas behind that elaborate fixturing to design weld fixtures that would work for lower-volume components,” Sullivan said, adding that such experience leads to a virtuous cycle. The shop’s location helped it land the work; engineers gained unique fixturing experience; and that expertise helped them develop a fixturing toolbox for other customers.

Opportunity in the Highly Cyclical

Fabricators in general have a love-hate relationship with highly cyclical sectors. A Texas fabricator might, after an inevitably dramatic drop in business, vow never to overload on oil and gas customers again—but then the boom times come again, and it’s hard to leave money on the table. So they take a deep breath, accept the lucrative work, then muddle through the next boom-and-bust cycle.

“[Managing through highly cyclical businesses] is not a labor cost problem,” Sullivan said. “It’s about the inelasticity of labor during a ramp-up. When a cyclical sector is strong, it all hits at the same time, and everyone serving that sector has a lot of work. If you have a very skilled worker, that person now can go to another shop that’s willing to pay $5 or $10 more an hour. Everyone is competing for the same labor, and the labor pool cannot match the speed of the ramp-up. You end up paying a lot more labor, but you’re not getting more products out the door.

“The secret is being able to mitigate that inelasticity. You take the value creation out of the labor itself, and you instead put it into the coordination of labor, which is the definition of a company. So now, the company is creating the value, not the labor.”

To explain this, he pointed again to the company’s flexible robot welding cells. The arrangement has allowed Inland to ramp up significantly to meet the demand of highly cyclical industries without hiring more people. But the success doesn’t hinge on technology itself but instead the coordination of talent given the right tools to succeed: that is, loads of upfront engineering, custom fixturing, and programming expertise to weld lower and lower quantities, while making all of it robust enough to ramp up quickly during times of peak demand.

If demand weren’t so cyclical, the work—even if individual part attributes were challenging—would be inherently easier, and the value of that labor coordination would diminish. A challenging low-quantity job might require a specific machine or a certain skill, but the jobs could be muscled through, and even if the job faced major roadblocks, the overall operation wouldn’t suffer, since relatively little revenue was tied to that oddball job. Conversely, a consistently high-volume job could be designed for a highly automated system—blank, bend, weld—with minimal manual intervention. And the high cost for that design, and the necessary equipment investment, could be amortized easily over that high-quantity work. For highly cyclical work, the coordinated efforts of skilled people really make the difference.

Sullivan again pointed to the parts being fixtured for robotic welding. The value stream prior to welding wasn’t entirely automated. Yes, some blanks were cut in the AMADA punch/laser combo, which stacks parts automatically. But some blanks required that satin, dual-action (DA) sander finish, which meant they needed to be sent through a flat-part deburring and graining machine. Still, processes were in place so that the entire value stream—from blank to bend to weld to finishing—could be ramped up if need be. Quick changeovers allowed for smaller batch sizes, less work-in-process (WIP), and faster delivery, while the flexible automation provided the processing capacity.

“It’s a bit of a self-amplifying cycle,” Sullivan said. “The more exposed to these cyclical jobs we became, the better we got at designing our processes around them.”

Abel Garcia and Nick Beardo review a print in the forming department.

Tuning the Business for the Bay Area

Sullivan tends to look at new technology investment as something that doesn’t mitigate the need for skill, but instead augments the ability of the skilled people the fabricator already employs.

He pointed to a new TRUMPF press brake with automatic tool change. That system today complements another new technology, this one from IPG Photonics: hand-held laser welding. Certain jobs, particularly those that flow best in kits rather than batches, simply need the flexibility of a brake with automated tool changes. And many of those jobs (including one large one involving sensitive 0.040-in. stainless steel with exposed, cosmetically critical welds) suit the hand-held laser welding process best.

Hand-held laser welding isn’t a perfect replacement for all previously TIG-welded joints. Some hard-to-access welds, or those with less fit-up repeatability, are just better suited for the TIG torch’s nimble tungsten tip. In adopting the laser welding technology, Inland effectively made its existing TIG welders more valuable than ever.

The fabricator has “tuned” its processes and technology around its customer mix. That mix of work—low volumes that demand highly skilled labor combined with highly variable volumes from cyclical industries—provides enough margin to warrant the higher salaries Silicon Valley demands. In turn, its prime location gives it access to the skilled technicians and engineering talent. Today, Inland’s 200 employees generate nearly $40 million in revenue a year.

The operation might serve as a kind of harbinger for various kinds of metal fabricators across the country, many of which are benefiting from reshoring trends and a strong economy. Flexible automation continues to advance, yet the people using it really make the difference.

Competitive edges gained with equipment alone usually don’t last long before the edge erodes along with margins. Early adopters of the fiber laser know this all too well. After all, metal fabricators usually don’t rely on proprietary technology. If a shop purchases a bleeding-edge machine, there’s nothing legally stopping the shop down the street from doing the same.

A group of skilled people, both technicians and engineers, is harder to replicate, and Inland’s location has given access to plenty of them. Considering the cost of living, they earn well above industry average, but increased labor cost is really a small price to pay. Without technical talent, Inland wouldn’t be where it is today.

Slide Show | 8 Images