PragasitLalao/iStock/Getty Images Plus
When it comes to custom metal fabrication, quoting practices often shape the customer mix. A typical shop that grows to, say, beyond $10 million a year likely has a few “whale” accounts with repeat orders that bring in significant volumes and revenue. Then there’s that long tail of low-quantity jobs that the sales team works to grow into larger, more lucrative accounts. Greater volumes simplify an operation. Perhaps a larger account can have a dedicated multiprocess cell or even its own value stream.
One task makes this customer mix, which often comes with high revenue concentration, especially common at many fab shops. It’s the quoting. Every job needs it, and it’s difficult to scale. Creating and submitting a bid can take hours or days, especially when it entails complex routings and downstream processes like welding and assembly. Winning is no sure thing.
This makes staying small a natural fit for many shops that want to focus solely on low-quantity and prototype work. To truly scale such an operation, quoting needs to evolve. And at Protolabs' sheet metal division outside Nashua, N.H., the evolution is well underway.
Why Quoting Is Difficult to Scale
Data from recent Fabricators and Manufacturers Association benchmarking surveys show that the average quote turnaround time is three days, and the bid-win rate hovers around 30%. To be sure, these averages come from a wide range of responses from diverse metal fabrication businesses, but the fact remains that quoting any job remains a gamble.
Sometimes sales engineers quote only those jobs that are easy to bid on or sit squarely within the fabricator’s sweet spot, like a product family or industry. This builds focus and, often, a core group of big customers that help the fabricator scale.
But what if a fabricator wants to specialize in the low-quantity niche? Engineers might bid quickly on a wide range of jobs and take a win-some, lose-some approach. Certain jobs might flow through quickly while others might hit significant roadblocks and end up being money-losers. It’s just the price you pay in the low-quantity world, the thinking goes. Quick response is key. A bid might be carefully crafted and incredibly accurate, but if it’s submitted late, a competitor may have already won the job, so all that work goes out the window.
The quoting environment is starting to change. Enterprise resource planning (ERP) and specialized quoting software have streamlined the process significantly. Instant quoting services for cutting and bending have emerged, with customer-facing web portals and Amazon-like transparency.
Protolabs’ approach in particular has revealed how quoting automation can fit into a diverse fabrication environment, one that cuts, bends, welds, powder coats, and assembles. Call it a quoting automation hybrid, with estimators supported by customer-facing software.
Making Best Use of Talent
Custom and contract fabricators invest in technology not only to increase throughput but also to make best use of talent. The last thing a shop wants is for its best brake operator or welder to spend monotonous days on easy brackets or weldments.
Focused on low-quantity work, Protolabs’ New Hampshire plant hasn’t faced this problem. Talented welders and machine operators often work on multiple jobs a day, so tedium usually isn’t an issue. Its plant is split into two value streams. One has multiprocess cells dedicated to straightforward jobs, while the other specializes in challenging work—large, unwieldy pieces; sheet metal with complicated, numerous forms and tight tolerances; and the like. The arrangement separates the work to maximize throughput and prevents the challenging work from causing a traffic jam, with piles of easy-to-fabricate jobs waiting for machine time.
An operator in the express lane forms a part on a compact brake. He changes over quickly by pulling up the program, developed and simulated offline, and switching out tools stored nearby—not far away in a centralized tool crib.
The separated workflow also makes best use of talent. Within the lane dedicated to complex work, specialists cut, form, and weld challenging work. Meanwhile, in the lane focused on straightforward jobs, cross-trained workers focus on managing flow, staging work, and streamlining setups. You won’t find any central tooling crib for the press brakes. They’re all strategically staged near some advanced press brakes forming parts that have been simulated offline. Rapid task-switching is their forte.
This layout sets the foundation for Protolabs’ in-house scheduling software. As Chris Litchfield, estimator at Protolabs’ New Hampshire plant explained, “Calculations are made from data collected over many runs, and we create an average. That’s how we have a good idea of how many seconds per inch it will take the laser to cut a part, the average setup and run-time for an individual bend, and so forth. The data has been gathered over a very long period of time. And any improvement we make—if we alter a process, get new tooling to streamline a setup, or anything else—we capture it and put it back into our system.”
The Science of Sheet Metal
Protolabs’ sheet metal operation is no stranger to software. The plant launched in 2001 as Rapid Manufacturing, and in December 2017 it was purchased by Protolabs, a Minnesota-based manufacturer specializing in (as its name suggests) prototyping and low-quantity runs using a wide variety of manufacturing processes.
In recent years, the overall Protolabs organization has built some automated design for manufacturability (DFM) capabilities for its injection molding, 3D printing, and machining operations in Minnesota. Customers upload digital files online, view and manipulate a 3D image of the part, and almost instantly receive a manufacturability analysis—complete with a 3D animation highlighting problem areas. Most important: The DFM analysis proposes a solution. The user can accept those changes and move forward with the order or can reach out to a salesperson to discuss options.
In December 2023, the New Hampshire sheet metal operation launched something similar. The project origins go back to software, called eRapid, used at the Nashua plant before Protolabs acquired the fabricator.
“This new system is an entirely new animal [compared to eRapid],” Litchfield said. “It behaves similarly, but it captures a lot more information and is capable of identifying many more issues.”
Today, if someone uploads a sheet metal part with conventional bends, the system flags certain manufacturability issues that might seem obvious to those who work in sheet metal every day—but not necessarily to a customer’s purchasing personnel.
“For many, sheet metal can seem like more of an art than a science,” Litchfield said.
The software demystifies the process and shows customers the science. They upload a 3D file, see a representation on screen, select a material grade and thickness, then view a manufacturability analysis. If a hole is close to a bend line, it points out the issues. If an enclosure has no corner relief, it flags it and suggests adding one. If a bend radius isn’t possible with the shop’s current tooling, the software suggests an alternative radius. If hardware placement would create difficulties, the software flags the issue and can sometimes suggest a different approach.
A quote and DFM analysis of a simple bracket might be entirely automated and turned around in seconds. An order for, say, 30 different cut and bent blanks might take a little longer. A more challenging job, or one that involves a multitude of processes like welding and assembly, would be flagged and sent to an estimator.
Here, the bend radius specified cannot be air-bent with Protolabs’ available tooling. The customer can see the achievable radius that the closest available tooling can create, then can either accept the change or choose to consult with a company representative. Images: Protolabs
The software helps streamline the manual quoting work that remains, which, according to Litchfield, can take anywhere from 15 minutes to more than four hours, depending on the job complexity.
“It used to be that every job we bid on was manual, whether it was a flat part or a 25-piece assembly,” Litchfield said.
This meant that estimators spent much of their days on simple, monotonous tasks—move a hole, add a weld notch, change a radius. They also had to communicate each change to the customer with every bid. Time spent on all that took time away from more complicated estimating work.
Sheet metal DFM software has changed the customer experience, of course, but it’s also changed the estimator’s workday. The monotony is automated, and every change, no matter how minor, is communicated with the customer automatically, right on the web interface. If, say, a radius isn’t possible with the company’s available tooling, the software will suggest a new radius, which the customer can approve or deny. When a solution can’t be found, the job is sent to the estimator.
“Most customers can live with an alternative radius,” Litchfield said. “But for that one out of 100 customers that need a specific radius, we want them to know about it in advance so we can address it.”
This means estimators spend their days on more complex work—the “exceptions,” so to speak. They also work with jobs in which a 3D model isn’t available. “We still haven’t found a way for a computer to read a print, at least one that’s scanned as a PDF,” Litchfield said.
Those prints often have assembly and welding information, both of which can’t be automatically quoted. Litchfield added, however, that even for simpler parts where 3D models are available, Protolabs allows customers to upload prints. The prints might not be used for quoting, but the PDF is still there for workers to reference.
“There is so much information on prints in regard to finishes, specific material types, and other details,” Litchfield said. “For us, it just didn’t make sense getting rid of them.”
Estimators also shepherd some production-level work through the system. “When we do get a request for a production-level number of parts, that job is put in front of our manufacturing team,” Litchfield said. “Leadership will look at shop capacity and lead time, to see if it’s something we can offer. For the most part, though, we stick to our wheelhouse—prototypes and short runs.”
At this writing, Protolabs is monitoring and perfecting its sheet metal DFM software. “It’s never going to be 100% perfect,” Litchfield said. “That’s just due to the nature of what we’re trying to do. But we hope to get it as close to perfect as possible.”
The quoting tool shows a feature is too close to the bend line and may cause distortion. If the feature isn’t critical, the customer can approve and move forward with the order. Alternatively, they can reach out to a representative.
Litchfield doesn’t foresee a world in which the estimator function is completely automated. After all, many customers value the personal relationship. The focus is on making estimators’ workday more engaging. The company doesn’t want highly trained people mindlessly clicking on 3D models to change a hole location or radius. They want them to suggest a novel approach to a complicated problem—a unitized design that eliminates welding, a difficult feature that could be manufactured a different way, perhaps an alternative sheet metal grade more conducive to a specific fabrication process. Or they could see the potential for entirely different manufacturing processes, some of which might be offered by other divisions within Protolabs.
This hybrid approach to quoting automation, with customer-facing software complementing manual estimating, could be a way for that low-quantity fabricator to grow in new ways, which in turn could drive future technologies—machines and software, on the floor and in the office—that enable thousands of custom, one-time (nonrepeat), low-quantity orders to be fabricated on a massive scale.
A welder gas tungsten arc welds a workpiece. Quotes involving parts that are cut and bent are completely automated. Orders that call for other processes, including welding, undergo a manual review.