����ý

At additive manufacturing service provider , AM for production has already arrived. Owned and operated by engineers and brothers Carl, Brian and Sean Douglass, the company has been succeeding at part production through polymer 3D printing for the last several years. Over 85% of the work underway at its facility in Willmar, Minnesota, is production for medical devices, consumer goods, defense parts and automotive components.

These are some of the applications where additive today makes sense, instances where the volumes required are easily achieved on one of the company’s four full-color or six standard HP Multi Jet Fusion (MJF) 3D printers. 3D printing for these products can deliver speed, performance enhancements and the right price point.

But DI Labs has its sights set on additive’s future as well, and sees major opportunity in what this technology can bring to higher scales and mass customization. The Douglasses started an entire brand around this idea, called , through which they offer aftermarket automotive parts that can be tailored to a user’s vehicle and interests (more on this in a moment).

As the Threedom ����ý has shown, the challenge with mass customization is no longer manufacturing technology, as AM readily meets requirements of a production application where each object is slightly different. The difficulty now lies more in the realm of design, and how to accommodate all the possible variations a customer might want. Through Threedom, DI Labs is learning how to support an enormous range of potential SKUs, all manufacturable through 3D printing, and translating these lessons into benefits for its scaling production ����ý.

From Technology Supplier to User

Carl and Brian Douglass cofounded DI Labs in 2013, originally using a horse barn on Carl’s property and pursuing a much different mission: launching a printhead for fused filament fabrication (FFF) machines. The concept was initially released through Kickstarter, but when OEMs showed more interest than consumers, the brothers licensed the technology and began selling the hardware to printer builders directly.

With the introduction of HP’s Multi Jet Fusion (MJF) technology several years later, however, the Douglasses saw a 3D printing technology they believed could be applied for scale production. They purchased an MJF printer in 2017 and spent a year proving out the platform and building a customer base for additive manufacturing services.

The company today focuses on verticals that can benefit most from AM — specifically those where injection molding might otherwise impose significant design constraints, and where time to market is critical. Most ����ý today comes from the healthcare space, but the company also serves customers in defense (both verticals value AM’s time savings) as well as consumer goods, where mass customization is becoming a key offering.

“Becoming” because it has taken time for product developers and inventors to come around to the possibilities that additive manufacturing brings with regard to mass customization. Without AM, offering custom products was prohibitively difficult and expensive in most cases, and awareness of this possibility is still growing.

So, in order to explore and begin to prove out mass customization with AM, the cofounders launched the Threedom line in 2020. Threedom offers customized aftermarket accessories for Jeeps, whose owners are notorious for decking out their vehicles with unique items. The idea started with customized key fobs, and today Threedom offers nine types of accessories ranging from hinge covers to light panels to door handles.

But here’s where mass customization changes the game: Across these nine product lines, Threedom offers more than 50,000 possible SKUs. In other words, there are more than 50,000 possible results that a customer can achieve by ordering through the company’s online configurator, all of which can be manufactured at DI Labs.

This flexible variety is possible because each product line has been designed using a “feature tree” that lays out which variables can be changed, and constraints on how much they can deviate from a base design.

“We had to ask, how do we build this for design flexibility?” Brian Douglass says. That flexibility was achieved by “deconstructing” each product and validating all its constraints. For key fobs, for example, it meant ensuring that any possible design outcome could still accommodate the circuit board, RFID chip and physical key itself while maintaining proper tolerances.  In the case of the key fobs, the basic design and its branches offer a possible 10,000 “child” designs.

Equipped for Flexible Production

To meet the needs of both Threedom and DI Labs’ regular, ongoing production, the company has invested heavily in polymer 3D printing technology, still primarily MJF. At the Willmar facility, there are seven MJF 4200 machines served by about 25 swappable build units; most non-consumer parts (medical device housings, orthotics, tactical service housings) are made on these 3D printers. In a separate area, DI Labs also operates four HP Jet Fusion 580 Color 3D printers, its choice for consumer items where getting the color right matters.

(DI Labs finds these printers so important for this purpose, in fact, that when HP discontinued the platform, the company purchased several used machines to ensure its ability to scale. There are two such color printers waiting in storage for the moment they are needed.)

To support MJF production, DI Labs has several depowdering units dedicated to different materials; a semi-automatic tumbler-blaster; a VaperFuse unit from DyeMansion for vapor smoothing of printed parts; and two AMT PostPro3D machines used for smoothing unique geometries. Investing in this suite of postprocessing equipment enables DI Labs to keep up with what its printers can produce, whether that’s many identical parts for a medical device maker or a batch of Threedom key fobs of all different designs.

There are some tricks to managing batching and workflow — for vapor smoothing, parts need to be of similar color tone and geometry, for example — but the company is doing what it can to “take the guesswork out of the process,” Brian Douglass says. An enterprise-level ERP system captures the part “recipes,” parameters and workflows to ensure repeatability, and “enable rapid change without a struggle,” he says.

While MJF is the primary platform used for production, DI Labs is also equipped with a Carbon DLP 3D printer, also used for production work; two Formlabs Form 3 SLA printers used for prototyping; two Stratasys FDM 3D printers; and a Markforged M2. Other equipment, including a Keyence visual CMM, is applied for reverse engineering parts as well as inspecting work to understand variability in the design or process.

Custom Informs Scale Production

DI Labs has been working to develop tools and workflows to enable custom Threedom products for the last five years, while also delivering more standardized production parts. Ongoing manufacturing work includes parts such as housings for medical devices, orthotics, and a high-volume application in the adult industry for which DI Labs manufactures more than 120,000 parts per year.

But lessons learned from mass customization manufacturing are beginning to inform the company’s standardized production jobs as well. A recent high-volume application offers an example. Minnesota-based organic farming company Lettuce Abound recently began using aeroponics, a method of growing plants in a medium other than soil. To hold the organic grow medium and support plant roots, Lettuce Abound needed baskets that would be easy to handle and clean. DI Labs worked with the company throughout prototyping and development and now produces these cages, seen below:

In the prototyping stage, DI Labs applied some of the lessons learned from its mass customization ����ý to deliver 1,600 baskets in each of 10 designs — a total of 16,000 parts for functional prototyping. The company developed and produced these slightly different baskets quickly by applying some of the same strategies used with Threedom parts, including flexible design within constraints and batching of similar items in postprocessing.

Lettuce Abound was able to thoroughly test each design and arrive at the best basket quickly, ultimately going into production with DI Labs on a design that resulted in decreased growing times while also increasing lettuce production by 30%.

“With injection molding you have to find the common 80% of customers and serve them with one product. With additive manufacturing, we can divide customers into smaller groups and serve each of them.”

The aeroponics basket is an interesting example of both the scale that is possible with DI Labs’ additive manufacturing capacity (thousands of parts, delivered quickly) and how design variation (multiple, similar parts) need not be a hurdle to this scale when 3D printing is the production method. It also suggests a point about where manufacturing could be headed with this technology:

“With injection molding, you have to find the common 80% of customers and serve them with one product,” Carl Douglass says. “With additive manufacturing, we can divide customers into smaller groups and serve each of them.”

Threedom is a step toward that future, where mass customization makes it possible to serve a larger customer base by offering more choices. Today Threedom products account for about 10 to 15% of DI Labs’ overall ����ý, but that percentage could rise as demand grows and technology advances.

Although customers already have great flexibility in what they can order through the system by virtue of the “feature tree” architecture, the Douglasses are looking toward a future where customers could upload their own designs, or even simply describe in words the product variation that they want to an artificial intelligence tool able to generate a suitable design on the fly. An AI tool like this could, in turn, help with iterating and finessing designs for industrial, high-volume customers.

“We’re mapping a pathway toward this with products like the key fob,” Carl Douglass says. “Unless we invest in developing pathways like this, mass customization with additive just won’t happen.”