Episode 9: The Future is LAYER3D

The disruption to our supply chain during the Covid 19 pandemic was pronounced.? Ships were stacked up outside the Ports of Los Angeles and Long Beach.? Trucking firms were trying to extend the available hours their drivers can be behind the wheel to move product to the interior of the country rapidly.? The White House created a Task Force to address the problem.?

With their supply chain as backed up as everyone else’s, energy giant Chevron chose a different strategy.? Unable to get key replacement parts it required for a $55BIL project in Australia, it approached a Sydney-based 3D print startup called AdditiveNow .

The company didn’t simply deliver the parts on time – they so impressed the team that Chevron decided to license AdditiveNow’s intellectual property.? This type of need is industry wide.? Italian 3D print company Roboze fortuitously opened an office in Houston just before the pandemic hit.? It was inundated with requests from the oil and gas industry to provide parts that were otherwise unavailable.

Despite this pair of success stories it would not be too far off the mark to characterize 3D printing as the manufacturing industry’s Boulevard of Broken Dreams.? Early on the technology was described as a game changer for making products and government agencies devoted considerable resources to it.? With price as no object, these early innovators weren’t constrained by what was a good business – they instead focused on pushing the art of the possible.? And some of their developments are indeed amazing.?

The Oak Ridge National Laboratory (ORNL) outside of Knoxville, Tennessee was an early experimenter in additive technologies, with a budget and thought leadership that allowed them to be creative despite the significant hardware and raw material costs.

ORNL 3D printed a Shelby Cobra convertible – not simply pieces, but the entire vehicle (other than tires, glass, and electronics).? The car has now become ORNL’s most popular ‘eye candy’.? This is no static mockup – each subsequent Energy Secretary has driven it around the Lab’s parking lot.? But it’s hardly their only achievement.? ORNL also 3D printed a miniature submarine for Navy SEAL Teams to further demonstrate the potential for on-demand 3D print production.

?While such one-offs are impressive, they have the additional benefit of moving public perception.? When national laboratories demonstrate the art of the possible, the science of the possible soon follows.? Once that science is understood, the business of commercializing begins, and Additive Manufacturing is no different.

A couple of companies have come and gone including MatterFab and Electroloom, while a few larger firms such as Carbon, Stratasys, and Desktop Metal have grown, albeit in fits and starts.? From raw materials in polymers, resins, carbon fiber, composite, and several types of metal, the technology is adapting for multiple industries from medical devices to aviation to automotive components.?

While building 3D printers is relatively straightforward, operating them - as a business of selling the 3D printed items - can be very difficult.? The early innovators appear to be differentiating along three models:

?

Personalized Model - customized for each customer, so the necessary cost premium is justified

Performance Model - supply chain performance by decreasing shipping time, with speed overcoming the cost barrier

Preference Model - a service bureau offering low frequency parts (distinct niche markets) on a build-on-demand basis, essentially mass prototyping

?

There are very distinct structural dynamics for these three models. For the Personalized model there is a very high design mix for each customer.? Those customers are low in number, and the volume count is even lower, including one time only.? Every ‘repeat’ business means starting from scratch with a new design.

The Performance model has a similar low design mix – most of the designs are fixed, perhaps even printing on the customer’s site, eliminating shipping costs entirely.? Again, the volume count is extremely low.

Finally, with the Preference Model, where design is cosmetic vs. custom and the customer count is very low, but the aggregated niche volumes can be quite high.? A key point here is that this scaled personalization is cosmetic vs structural, allowing for mass customization that, together, reflect the more typical economies of scale.

?

Personal Model

Few startups have mastered this platform approach quite as well as Invisalign Technologies.? After breaking into the orthodontics world twenty years ago with a radical concept for replacing braces to correct teeth, they’ve become a dental unicorn and forever changed oral healthcare.? A scanning wand captures thousands of digital images a second in the patient’s mouth.?

These photos are compiled to create a three dimensional model.? A predictive software application then maps a proposed sequential movement of the teeth over time.? The models are used to 3D-print personalized casts of the patient’s teeth that are the die molds to make the aligners that move the teeth around.? The high price point reflects how each patient’s product is custom made – but twenty years of satisfied customers is a powerful selling point.? No more ‘metal mouth’ teasing.? A truly disruptive innovation.

?

Performance Model

According to CapGemini , the last mile of delivery can produce up to 41% of logistics costs. UPS sought a way to leverage their 1000+ warehouse locations worldwide.? Their strategy was (and likely remains) to build 3DP capabilities into their logistics operations.? UPS invested in a 3D print startup called Fast Radius (FR).? Fast Radius CEO Lou Rassey had an MIT engineering degree and several years running McKinsey’s manufacturing practice.? But things did not quite go as planned, despite many successes.

It has been a great learning experience for UPS.? How do they guarantee the validity of a customer’s design file?? What are the counterfeiting risks for high margin parts?? Who owns the risk, both in design and file integrity?? How would they test and validate complex products?? UPS continues to explore these questions.? But the risk uncertainties, and Covid 19, were too much. Fast Radius’s assets were acquired by SyBridge Technologies in 2022.

?

Preference Model

In 2016, Adidas launched their first robotics manufacturing site in Ansbach, Germany near their Bavarian corporate headquarters. Known as a SpeedFactory , it blended robotics and 3D print technologies in unique ways. Part of this automation was in the form of 3D printed soles by a startup called Carbon. Combining Carbon’s expertise with several separate robotic components, the SpeedFactory was a showcase.? A shoe taking several weeks to produce in Asia could be made in the SpeedFactory in five hours.

But that complexity created other problems.? Adidas makes hundreds of millions of shoes per year – at best, additive manufacturing would only be able to produce less than 1% of this total.? So while mass customization looks good on paper, in reality additive technologies remain labor intensive, the very problem Adidas was hoping it would solve.? As with UPS previously, Adidas ended the program in 2019.

?

Adidas is not alone.? Competitor Nike has several patents for 3D printed shoe components.? In 2016 the Investor’s Business Daily ran a story about Nike having more patents than Lockheed, Ford Motor Company, or Pfizer. ?Shoes are a cutthroat business, where even a minor improvement in cost structure can rapidly scale to a significant improvement to both revenue and profits.??

One thing Adidas proved was the digital platform for such an undertaking can be done at scale among multiple customers and suppliers.? The learning curve remains steep for the manufacturing world, but Adidas and Nike are making these early bets on the technology to lead the change rather than follow it.? These digital platforms are causing traditional manufacturing to catch up to where innovative startups - like Fast Radius and Carbon - are impatiently waiting.

What I think we’ll see next is less focus on manufacturing IP and more of a focus on business model design.? None of the examples here failed due to a design or engineering flaw.? Many of them were, like Apollo 13, ‘successful failures’.? What I expect to see is economists attacking this problem head on, creating new business models that are specific to 3D printing.? Whether these become business method patents or remain trade secrets will be worth watching!