The Engineering Research Visioning Alliance (ERVA) published a report on Engineering the Future of Distributed Manufacturing, identifying priorities that can revitalize U.S. manufacturing with sustainability and resilience in mind. In this podcast, Tali Rosman, startup advisor and Entrepreneur-at-Residence at Toronto Metropolitan University, shares how 3D printing plays a role in ERVA’s vision. Tali has worked at Stratasys, led Xerox’s former 3D printing subsidiary, Elem Additive, and now serves as a business advisor with expertise in advanced manufacturing.
What is ERVA, and what is the report about?
ERVA is the Engineering Research Visioning Alliance. It’s a National Science Foundation (NSF) funded initiative that works to identify future engineering research directions by enabling a wide range of voices to impact national research priorities. Think about it as bringing together the engineering community and unifying it as one voice to identify research topics that will have the most impact on the most complex problems we’re facing as a nation and as a society, such as the future of distributed manufacturing. ERVA invites experts across a multitude of disciplines to weigh in on those topics at visioning events. Finding solutions to challenges like those requires a big-picture approach. I was very fortunate to be one of 56 experts who attended such an event in March [2023], where we identified the priorities you’ll find in the report.
We identified three big challenges engineering research is positioned to address and move U.S. manufacturing forward. Those three grand challenges are (1) the materials supply chain and how to ensure they are secure and distributed; (2) enhanced tools and processes to create any discrete manufacturing product, anywhere, anytime, in any quantity, and with the desired quality; and (3) the data and quality assurance engineering research to improve the efficiency and production processes, regardless of the lot size and while optimizing inventory management, reducing the downtime, and enhancing product quality.
How does 3D printing help achieve these goals?
Regarding the second challenge of enabling discrete manufacturing at any time, anywhere, in any quantity, without compromising on quality, this is where 3D printing, or additive manufacturing, really kicks in and enables discrete manufacturing. With additive manufacturing, the variable cost per part is fixed versus traditional manufacturing, where you have economies of scale, so the cost comes down. With 3D printing, you eliminate the need to create toolings and molds and save all that upfront investment. You can make one part or 100 parts, and that cost will be fixed, unlike traditional manufacturing, where if you make just one, two, or 10 parts, that cost will be extremely high because of all the upfront costs. With 3D printing, you can have true direct manufacturing because you don’t have all this setup and can achieve it anywhere you want.
Again, think about distributed manufacturing. With traditional manufacturing, you need economies of scale, and you’re going to make 100,000 parts in one location. Now that your variable cost is fixed with 3D printing, instead of making 100,000 parts in one location, you can make 1,000 parts in 100 different locations, getting you to that 100,000 quantity in a smarter way. I’m saying “smarter” because a distributed manufacturing approach with 100 factories close to the point of consumption gives you supply chain resiliency because you’re not dependent on the logistics and shipping. We’ve all seen in recent years that there could be massive disruptions there. This is how you make things really close to the point of consumption, and you’re saving all these risks and all these overheads.
Obviously, you still need materials. That’s touching on the first challenge — the material supply chain. But you need less material because you’re only making the quantity you need where you need it. Again, you’re not making 100,000 upfront, you’re making a much lower quantity on demand at the place you need.
Industries such as aerospace, defense, and medical are still the most willing to use and experiment with 3D printing for prototyping and end-use parts. Is there more traction in other industries?
First, it’s important to understand why aerospace, defense, and medical are ahead of the curve and adopting additive. There are two big reasons for that. The first one is they often have low volumes. In aerospace, you’re not making a million of a particular part, you’re probably only making a few dozen. So, the industry has inherently low volumes, and 3D printing allows you to make these low volumes in an economically feasible way.
In 3D printing, we often say the sweet spot is low volume, high value, which means you’re willing to pay a premium to get the part in your hand faster because you’re making it in direct manufacturing at a remote factory, and you have to ship it. Aerospace is obviously extremely high-value parts, so that’s a sweet spot for additive.
The other point, which might be very important in the medical space, is that with 3D printing, you have a lot more freedom of design. You can make geometries that are not possible with traditional manufacturing. And because you have these low volumes that are made economical, you can have what we call mass customization. Mass customization means I need maybe a million of a certain part, but there’s value if each part can be slightly different. For example, in the aligners industry, an industry that’s been 100% conquered by additive manufacturing, everybody’s doing 3D printing because they’re customizing your aligners. Also, you’re looking at 3D printing starting to gain traction in hip implants, knee implants, etc., where your ability to customize something to the person’s anatomy is extremely valuable. That’s the high value we discussed, and it’s inherently extremely low volume because you’re making it just for this person’s anatomy. And really, you cannot achieve that with traditional manufacturing.
Now, when you think about why additive manufacturing is gaining traction in these industries, you can extrapolate to where else we will see it. If we’re thinking about consumer goods, for example, 3D printing hasn’t gained a ton of traction there yet. That will hinge on two things for us to do as an industry. One is, again, thinking about these new business models. Maybe you’re not just selling aligners off the shelf, you’re mass customizing them. Are innovative business models and consumer goods now being created and opened up thanks to additive? And the second thing is, obviously we’ll need to improve the economics. I’ve been in this industry for about 10 years, and I’m already seeing massive improvements in the economics.
When I joined, 3D printing was mainly for when you wanted to make one or two, maybe in great cases, a few dozen. Now, you have 3D printing technologies, especially the sinter-based additive technologies, that are really pushing the envelope. They can be used for 10s of 1,000s, sometimes even hundreds of 1,000s of units, and they’re still economically viable. And if we get to that in a few years, I mean, one could assume that in a few years from now, five to 10 years down the road, we’re probably going to get to a place where we’re making even millions of parts economically viable.
Is it possible for small and medium-sized businesses (SMBs) to realize the benefits of 3D printing in the same way that large corporations can?
If anything, 3D printing is the great balancer for small and medium businesses to compete against large corporates. This comes back to economics, where low volumes are economic. If you think about traditional manufacturing, where low volumes are not economic, and you really start seeing your efficiencies grow the more parts you’re making, and with economies of scale, there’s an unfair advantage to the plant that can make a million parts. With 3D printing, because that variable cost is constant, you making 100 parts can compete on the same playing field as somebody who can make a million parts.
If we’re talking about distributed manufacturing, you don’t need a massive capital investment to build this mega factory, which is what we’re used to. Now, with a relatively small capital investment, you can have a microfactory, as we like to call it in the 3D printing industry. And now you can compete, maybe not for all of the volume — you might not have the capacity to make a million parts — but you can compete for some of the contract.
And again, maybe this is a good time to rethink buying a million parts upfront because, with 3D printing, you can have this direct manufacturing where you make parts on demand. Now, you can change your purchasing habits and buy things as needed, open it up to more SMBs, and reduce waste and inefficiencies by making too many.
What holds slower adopters back from even experimenting with 3D printing?
It’s a combination of a few factors. Some of it is just pure resistance to change, which is common in all industries but certainly in manufacturing — a very traditional, conservative industry. I will say that resistance to change and the lack of faster adoption are not unfounded. There is a lot of merit to them. First of all, in the past, 3D printing technologies have really been more suited for prototyping and not for production. So, there might be a lack of awareness of how much progress has been made in additive manufacturing. Also, we say additive manufacturing as an umbrella term, but in reality, there are a multitude of technologies under it. Some of them are fit for production, some of them are not. Some of them are more economically viable, and some of them are not. So, there’s this education element.
The other thing that maybe ties into one of the grand challenges is really just trusting the technology — that the quality of the parts is going to be there, and the parts are going to function. In the manufacturing industry, nobody gives you a metal when the part works and functions. We just assume everything in aerospace and medical and automotive and consumer goods is going to work, but there is a lot of trouble when a part doesn’t function, especially in the regulated industries, such as aerospace and medical. So, the question is, why should I switch to additive manufacturing? It’s a novel way of doing things. How do I know that it’s going to work? And so, we have to provide these tools and give this confidence to them.
I’ll say as a side anecdote, and this is a project I’ve done a couple of years ago, we were approached by a very large transportation company. They had a part they struggled to get and had nine months of lead time because of supply chain disruptions. They came to us and said, “Can you 3D print it?” and we said, “Yes.” It took us two days to deliver the parts: one day to print and one day to post-process and send the parts over. So, instead of nine months, it was two days. Then, they needed to qualify the parts because it’s a different way of making the parts than they were used to. That took six months. So that whole benefit of “you have nine months lead time for something, and we can do it in two days,” almost all that advantage vanished because of the data qualification and certification of the parts.
Tying back, I think a lot of companies are looking into additive and say, “Additive might be able to solve my problem, but how am I going to make sure this part is going to function? How am I going to qualify and certify them? And if now I have to spend six months on that, and on labor and costs and overhead, and maybe at the end we won’t even get certified. Maybe I wouldn’t take that risk.” So, I said resistance to change, but I fully understand where it’s coming from and the barriers we need to address to overcome this resistance.
Some people use the term “alien part” because 3D-printed parts can look so different. Even if they qualify, why is it hard to accept the look of the parts aside from the functionality?
This is exactly the point we were talking about before — that additive manufacturing allows you freedom of design and making geometries that were not previously possible. But when you make geometries that were not previously possible, it looks vastly different. And it actually has a lot of benefits because now you’re building the part layer by layer, instead of machining where you’re starting with a block of material and you’re shaving it off, you’re spending a lot less material. You can get the same material properties and the same functional performance of a part but with a lot less material and a lot less waste. Think about aerospace, for example. Reducing that weight also means airplanes can be made more fuel-efficient, which reduces cost and carbon footprint. But it looks different. Therefore, people are more hesitant to adopt it.
I think the upside is with all the advancements that we’re seeing in AI and data; we’re getting closer and closer to a point where we’re going to be able to give real-time quality assessments of parts to give all these manufacturers that are looking at these weird looking parts that are made in a different way with different technology, maybe even sometimes with a slightly different material, the confidence that the part will function. There are actually quite a few companies in the 3D printing space already that are embedding real-time inspection capabilities. As an example, one of the companies I’m working with is scanning every layer of the print to ensure enough material has been placed. They can do autocorrections in real time in the next layer to add or reduce material to make sure the part comes off as it’s supposed to. Doing these real-time corrections is becoming better, easier, and faster.
Think about all the advancements we’re seeing on the AI side, ChatGPT, and so forth. That’s obviously having a massive impact on additive manufacturing. We’re seeing more and more embedding of real-time AI-driven capabilities during the print process or design process, trying to have predictable outcomes for a specific optimized design. Touching on the point of resistance to change, a year and a half ago, people didn’t use AI as much in a mainstream way. It was more reserved for the R&D labs. And ChatGPT emerged just a little over a year ago. It’s hard to fathom a year ago, we didn’t have this incredible tool. It made the mainstream extremely comfortable — all of us, literally almost overnight, very comfortable with AI. And I think we’re going to see a similar process and additive. So, I’m very optimistic that with all the advancements that are coming and all the additional investments that are going to be pouring in, we’re going to see this resistance to change reducing over time.
What types of 3D printing technology are used most today, and is that changing, especially with AI-related and other advancements?
In 3D printing, we usually divide the market into the plastics and the metals. In plastics, you have a lot of technologies that have been around for a while, such as FDM (fuse deposition modeling), which is also called FFF. You also have the SLA and SLS technologies. They’ve all been around for 30 years, and customers are certainly getting more comfortable with them. Where you really see a lot of the growth in additive manufacturing, though, is on the metal side, which makes a lot of sense. We’re moving from being an industry that focuses on prototypes to an industry that focuses on production that, again, used to be dozens of parts, and now it’s 100s, 1000s, hundreds of 1,000s. Maybe in a few years, we’ll get to millions of parts.
Metal is the fastest-growing segment. The market today is still dominated by the powder bed fusion technologies, where the feedstock is powder. Where I actually think a lot of the growth is going to come from technologies that are not based on powders. The reason I say it is because, first, the powders are expensive. Talking about economic viability, it’s harder to get to it with a very expensive feedstock in the form of powders. The other thing is that powders are toxic and explosive, which creates environmental health and safety risks, which a lot of companies don’t want to bear. It also makes the post-processing — everything that happens from when the print finishes till the party is actually usable — much more cumbersome and much more costly. So, especially as we’re going into volumes, we’re going to see technologies that are not reliant on powders gain a disproportionate amount of the market, whether it’s because they’re based on metal wire or metal paste; I really think that’s where the market is going.
In addition to scaling production with 3D printing, is a distributed manufacturing approach a viable solution for SMBs to expand geographically?
Absolutely. In the ERVA workshop, we talked about the ability to secure the supply chain — supply chain resiliency. And with that, it allows smaller companies to compete even better. Again, we talked about the fact that you don’t need to build a single giant factory anymore. You can just have, you know, two or three microfactories and get to be, not all of the manufacturing contract, but at least a share of the manufacturing pie. And the emerging metal technologies not based on powders but on wire or paste are much easier to learn and operate. So, if you think about workforce readiness, that will have a lot of impact.
When you have a microfactory and technologies that are easier to operate, smaller companies — because they don’t have all the overheads that a large company would have — are more nimble. They’re also local in the region. They know the local needs. They can be a lot more responsive and probably a lot more efficient than the big corporations in a distributed manufacturing world. And I think a distributed manufacturing future gives the SMBs a lot more importance than the big companies.
3D printing is often associated with being a more sustainable solution. Is that true? And could distributed manufacturing networks improve efforts toward net zero targets?
The point on sustainability is absolutely true. I think about it in two parts. One is just the sustainability of making a part in traditional manufacturing versus 3D printing. And the second is where you make the part. So, on the first point, with 3D printing, you’re directly making the part, and you’re not starting with a block of material that you shave off, but you’re actually just using the amount of material you need to make the part. There’s inherently a lot less material wasted. And also because it’s direct manufacturing and closer to on-demand than traditional manufacturing. You can make just the quantities that you need and significantly reduce waste. And the point that I made about the metals, while powders are toxic and explosive, I’m really seeing this emergence of technologies using wire and paste that are perfectly safe. As that gains more traction, the point will be even more true.
The second point we’ve been hammering on in this podcast is that with 3D printing, you can have distributed manufacturing instead of centralized manufacturing. And so instead of making the part, for example, in China and then needing to ship it to the U.S., well, now you can make it locally closer to the point of consumption. So, we’re eliminating all this transportation, shipping, etc., which isn’t just eliminating costs, it’s also significantly eliminating carbon footprints associated with transportation.
Anecdotally, when I was in Xerox leading Elem Additive, we were the first ones to put a metal printer on a U.S. Navy ship. That’s really taking distributed manufacturing to the extreme. Because before if a part breaks in the middle of the ocean, maybe if it’s not a critical part, you had to wait for two months until the next time you dock. Or maybe if it is more critical, you had to send in a Black Hawk with spare parts, which is not very sustainable. Well, now you can just make the part on demand on the ship. And if that’s possible, and it’s actually happening today, certainly I think we can see distributed manufacturing inland.
What will it take to realize ERVA’s vision for U.S. manufacturing in the next 5, 10, or 20 years?
First, I’d encourage everybody who’s interested to read the report we’ve published. But to give you the executive summary, number one is we need to engineer new sustainable materials for use in advanced manufacturing. There’s a need to increase the quality, supply, and sustainability of materials and to better understand the trade-offs in substituting some of the materials, so maybe we can switch to more localized materials for specific purposes. Of course, no matter what kind of material we’re using, we need to optimize the manufacturing method to guarantee quality, and the qualification and certification have to evolve to account not just for potentially a novel manufacturing method, 3D printing, but also to account for the fact that the material might not be the same. So, we have to have improved quality methods.
The other thing we talked about is enabling new business models that better manage supply chains to build resiliency, minimize disruption, and reduce wasted material. And that’s everything from increasing the supply chain viability with improved analytics, having distributed manufacturing that enables you to have a more resilient supply chain, moving from ordering a million units from a megafactory to maybe smaller batches from a microfactory, and also thinking about reusability of parts, the whole trend of moving towards a more circular economy. All this really requires a fundamental change in how we manage the supply chain and the incentive structure associated with the people managing the supply chain. So, that’s the second point.
Third, we really need to design those next-generation machines that are small, agile, and reconfigurable. If I’m looking at 3D printing and where the printers were 10 years ago, where they are now, I’m surely seeing the strength. I think we’re going to be seeing more and more of these. The machines have to be easier to operate and easier to maintain. They should be reconfigurable. They should be more multipurpose and enable you, for example, to more easily switch between different materials and different sizes. Software is definitely another area that ties into that research. How can we leverage AI or have a ChatGPT, if you will, for parts manufacturing — everything that we talked about from the design process before you can start making the parts to during making the part. That all feeds into our need to create common, effective, affordable standards for data collection, analysis, and communication.
There’s no point in every company on its own reinventing the wheel when it comes to the data collection and qualification of the parts manufacturing and the data associated with it. We need to create common standards that everybody can adhere to, which would save us a lot of time, which means we all need to work together. If we want to regain global leadership, we need to stop working in silos, and we need to work together in a multidisciplinary way across materials, data, and analytics supply chains to truly enable the next era of distributed manufacturing in the U.S.
I think a lot of people necessarily want a career in manufacturing today. I don’t think it’s perceived as the ideal career path. I understand it because we all have in our minds factories the way they were 50 years ago. I don’t think it’s reasonable to say we’re going to reshore the manufacturing we lost like-for-like; we’re not going to rebuild manufacturing the way we did 50 years ago. That’s the critical point. When we rebuild manufacturing, we’re going to do it with all the new technologies coming in. We’re not going back to landline technology; we’re going to build it 5G style instead. If you’ve visited these novel microfactories that are using advanced manufacturing, they’re using AI and all this next-gen technology. And maybe I’ll just mention, for example, the Deloitte Smart Factory in Wichita and some others — once you go to these factories, this is an environment you want to work in. It looks high-tech; it looks technology driven. It’s not the image people will have in mind. So, if we do that, it will really encourage a lot of the younger generation to pursue a career in manufacturing.
ERVA
ervacommunity.org