• Fabri Raises $13.5 Million to Create Digital Foundry

    Fabri is a startup that wants to create a “digital foundry,” and just raised some funds to help it reach this goal. There are far too few foundries in America. These legacy businesses are closing or find it difficult to recapitalize after founders retire. Money is elsewhere, chasing bytes not bits, and foundries are an ancient business. Even if firms invest in manufacturing, then CNC, precision manufacturing, or 3D printing will be at the forefront of their plans. Foundries are old world, a part of an America that made things. Now, with most products being made outside of the US, this is a decided problem for the defense industry in particular. Fabri is therefore part of a new cohort of firms that want to reinvent foundries.

    As far as I know, Skuld was the pioneer, reinventing foundry technology years ago. With a renewed focus on defense, independence, and a US military gap, many more firms are following it now. The US needs to build more submarines to shore up its nuclear triad, it needs to replenish its depleted (by half to two-thirds) precision missile arsenal, and will need to build many more autonomous vehicles in the future. This is a significant opportunity for additive, but also Fabri and firms like it. What Fabri wants to do is use the latest computational tools to advance foundries through automation. By reducing labor as a component of investment casting and speeding up lead times, the firm hopes to grow by shipping parts faster.

    Fabri is 3D printing wax molds to skip a few steps in the investment casting process. Furthermore, they hope to automate many steps and use data and software to optimize production. For now, the company is working in Aluminum, copper, bronze, C71500, and steels, and wants to expand into nickel (IN713C, Mar-M247). Casting sizes are limited to 7 x 13 x 15.75 inches. The company is ITAR registered and has the CMMC Level 2 cybersecurity validation, as well as JCP Enhanced Validation, which also looks into compliance and security. 

    In some cases, foundry lead times can exceed two years, so this is a real opportunity with real money behind it. Fabri will now have real money as well. The firm has raised $13.5 million from the likes of Lavrock Ventures, Balerion Space Ventures, RTX Ventures, Lockheed Martin Ventures, Marlinspike, Tenon Ventures, and SBXi. The latter is a club comprised of Accel, Polaris Partners, GETTYLAB, General Catalyst, Pillar, Danaher, Underscore VC, and Glasswing Ventures, all working together to back MIT founders. Balerion is a space VC that has stakes in Anduril, Vast, and Relativity. So their money means a lot, given their exposure to firms that need to build stuff. Marlinspike, which has the coolest name of any VC, invests in AI, robotics, aerospace, and cybersecurity, and is also in Anduril. So having Anduril as a prospective client should not be too much of a problem. And having both Lockheed and RTX on this is just awesome for Fabri. Fabri also received $5 million in 2024, and RTX and Lockheed were a part of that as well. 

    Fabri says on its site that people should “Join us and rebuild American manufacturing. The DoD named castings a top-four defense priority. The foundry is the front line of reindustrialization.” Boom, no ambiguity there, no talk about golf clubs or spinal cages. This is a company with focus. And to go out and be able to just focus so fully on the DoD opportunity is a real benefit for the firm.

    The firm, which is led by ex-Relativity staffer Steven Davis, ex-Velo3D software engineering manager Pieter Coulier, and ex-Inkbit VP of Operations Tom Cole, benefits from real experience in wanting to do something old completely new again. Hopefully, they’ll develop a tool to fit a purpose, become reliable, and stray away from complexity. Fabri also hopes to leverage “our exclusive high-throughput additive manufacturing process and AI-driven design software, we can deliver casting in days, saving our customers critical time and money.” It is working on FoundryOS, which “interprets part geometries, generates tooling-free patterns, and automates process control from shell building through pour and inspection, with full traceability from melt to delivery.” And the firm is very ambitious, saying that, “the plan is one new foundry a year until American casting capacity is back where it should be.”

    The really cool thing is that they’re not going to (as far as we know) sell the software. The intention is also not to sell the equipment. Instead, they’re going to, like Seurat and Vulcanforms, use their technology to build a service. So they’re actually in the business of buying, building, and retooling foundries. We would expect more firms like Fabri to emerge over the coming years. Given the critical needs of US defense and the vast amounts of money at play, this is a very lucrative space to invest in. What I also like is that in Precision Castparts, they have a natural predator in that $10 billion behemoth. Smaller defense-oriented foundries such as Waukesha Foundry, Barron, and Signicast will also be more than a little curious. So they have exit opportunities aplenty in large and huge firms. Fabri may very well quietly and in the background build a very competitive significant business that is defensible to boot.

  • Largest Publicly Announced, Single Order in EOS History: Beehive Industries Spends $50M on M4 ONYX 3D Printers

    Earlier this year, Beehive Industries received a $29.7 million contract to produce its Frenzy 6 and Frenzy 8 engines for the US Air Force. The metal additive manufacturing (AM) user with facilities in Colorado and Knoxville, TN claims that its process is both faster and 60 percent cheaper than the conventional methods used to make engines for uncrewed systems.

    Beehive Industries leverages its large fleet of EOS 3D printers to support its workflow, a fleet that’s about to increase by more than double: EOS just announced that Beehive has ordered another 30 machines to be delivered over the next 12 months, which will bring Beehive’s total EOS capacity to 50 printers.

    Specifically, Beehive has ordered the EOS M4 ONYX, the company’s newest, most sophisticated metal AM system. The original equipment manufacturer (OEM) launched the machine at last year’s Formnext, with shipments to customers beginning in Q1 of this year.

    Relevant to the Beehive purchase, the American Center for Manufacturing Innovation (ACMI), an organization that supports US manufacturing enterprises in adopting new technologies to support strategic sectors including defense, was one of the first M4 ONYX customers, validating the machine’s use for the US military aerospace supply chain. Service bureau Incodema3D, a defense sector specialist, also committed earlier this month to buying four additional M4 ONYX machines.

    The US military reportedly will need years of work to replenish its weapons stockpiles following the conflicts in Venezuela and Iran, driving demand for domestic defense contractors. Additionally, new weapons systems in-development in response to the rapidly shifting global combat environment require a ramp up of AM-centric product development strategies.

    Anyone interested in learning more about that context should register for the UAS Additive Strategies webcast, sponsored by EOS and HP and presented by 3DPrint.com and AM Research, on June 30 from 11 AM-2:30 PM Eastern.

    In a press release about Beehive Industries’ order of 30 M4 ONYX printers from EOS, Beehive’s COO and CFO, Darius Ehshetami, said, “Beehive is experiencing unprecedented demand for our Frenzy 8 engines driven by major  defense programs and the urgent need for affordable, high-rate production of uncrewed systems. Our expanded  collaboration with EOS and this substantial investment in best-in-class 3D printers will  significantly increase our production capacity while reinforcing our commitment to delivering  scalable, American-made propulsion solutions that strengthen warfighter capabilities.”

    Marie Niehaus-Langer, CEO of EOS, said, “Beehive Industries’ unprecedented investment demonstrates how additive manufacturing has become a foundational production technology for the next generation of advanced propulsion systems. The success of the Frenzy engine program highlights what is possible when innovative design and industrialized additive manufacturing come together. We are proud to support Beehive as they expand production capacity and  accelerate the delivery of high-performance technologies to customers around the world.” 

    While I’m not naive about how much one company in the AM industry tends to view the next company with contempt, I still think that EOS is exactly the kind of company that the rest of the AM industry should want to see succeed. It’s a pure-play company that has been and will be in it for the long haul, and it has achieved its reputation the old-fashioned way, by providing a quality product and providing excellent customer service.

    The reason that matters is because that’s the only business model that can be reliably replicated in a machine tool market: you can’t replicate hype, or ride coattails, and even if you can, the lifespan for that method is inevitably short. That may be boring, but it’s also reality.

    Anyhow, what we can see from Beehive Industries, and Incodema3D, and Ursa Major, etc., going all in on EOS machines, is that eventually, building a business the right way can benefit an entire industry, not just one company. Going forward, I think this will become more and more apparent in terms of how the winners and losers get separated.

    That’s not to say there won’t be another hype cycle; in fact, the next one has already started! The point is that it should now at least be obvious which companies will outlast each successive hype cycle and make it through more unscathed than the competition. For the most part, they will be the companies that are okay with doing things the boring way.

    Images courtesy of EOS

  • As California Debates AB 2047, New York’s Law Targeting 3D Printed Guns Prepares to Take Effect

    California’s controversial AB 2047 is still making its way through the state legislature. Meanwhile, in New York, lawmakers have already moved forward with similar restrictions for preventing the production of 3D printed firearms.

    Governor Kathy Hochul signed New York’s legislation into law on May 27 as part of the state’s FY27 budget package. The law is expected to take effect later this summer and is designed to address so-called ghost guns, including firearms and firearm components produced using desktop 3D printers.

    The New York law focuses not only on the firearms, but also on the technology used to make them. Under the legislation, 3D printers sold in New York will eventually be required to include technology designed to prevent the production of illegal firearms and firearm parts. The law also directs the state’s Division of Criminal Justice Services to lead a task force that will recommend the standards manufacturers must meet. Once those regulations are in place, New York will be able to take action against companies that sell non-compliant printers in the state.

    Police agencies will also be required to report recoveries of 3D printed firearms to the state. The legislation criminalizes the unlawful possession, sale, or distribution of blueprints used to print illegal firearms and firearm parts, as well as the manufacture of certain 3D printed firearms.

    Hochul first introduced the idea in January as part of a broader gun safety package, targeting ghost guns and 3D printed firearms. The package included plans to encourage safeguards on 3D printers and limit access to certain gun-design files online.

    In fact, Hochul’s office says the measures are intended to keep state laws aligned with the changing technology. The administration argues that ghost guns and 3D printed firearms can be difficult to trace and that states need new tools to address how some firearm components can now be made outside traditional supply chains. When announcing the legislation, Hochul said New York was taking steps to close what she called the “plastic pipeline” and limit the spread of untraceable firearms.

    So the New York law and California’s AB 2047 are both trying to regulate the printers, not just the guns. In fact, California’s AB 2047 would require 3D printers sold in the state to include firearm-blocking technology and would create a state approval process for compliant machines. Beginning in 2029, non-compliant printers could no longer be sold or transferred in California.

    But, critics say it is not clear whether the required technology can actually work, and they worry the rules could affect schools, businesses, makers, and other lawful users of 3D printers.

    19 guns seized by members of the New York Drug Enforcement Task Force REDRUM Team. Image courtesy of the U.S. Drug Enforcement Administration.

    Among those raising concerns about the proposal is David Tobin, Executive Producer of 3D Printing Nerd and Executive Director of the Community Manufacturing Initiative. In a recent interview with 3DPrint.com, Tobin argued that lawmakers are focusing on the machines, rather than the criminal activity they are trying to prevent.

    “The things they’re trying to make illegal are already illegal. You can’t make them more ‘illegaler’,” Tobin suggested. “The focus should be on enforcing existing laws rather than creating regulations that affect everyone who uses a 3D printer.”

    Tobin has also warned that more states could begin exploring restrictions aimed at 3D printers themselves. New York’s recently enacted law shows that the debate is already extending beyond California.

    New York Has Been Working on This Issue for Years

    New York has long been working on this issue. In 2019, the state made sure 3D printed firearms were illegal. Since then, lawmakers have introduced additional proposals aimed at ghost guns, firearm blueprints, and the technology used to make them. In 2023, Assembly Bill A8132 proposed requiring background checks for certain 3D printers capable of producing firearm components. That same year, U.S. Senator Kirsten Gillibrand introduced the federal 3D Printed Gun Safety Act, which sought to prohibit the online distribution of digital firearm blueprints. The law signed by Hochul this year takes that effort a step further. Instead of focusing only on the firearms, it also examines the printers and digital files involved in their production.

    Backers of the law say it is a response to the growing number of ghost guns and 3D printed firearms being recovered by law enforcement. They argue that states need new ways to address weapons that can be difficult to trace. For example, Manhattan District Attorney Alvin Bragg has argued that regulations targeting both digital gun files and manufacturing technology could help create additional barriers for people seeking to produce illegal weapons.

    Not everyone is convinced the approach will work. Groups such as Adafruit and the Electronic Frontier Foundation (EFF) have questioned whether 3D printers can reliably identify firearm parts in the first place.

    For example, Adafruit argued that printers simply follow instructions and do not actually understand what they are making. The EFF has also raised concerns that these types of requirements could eventually lead to printers scanning or restricting user files. What’s more, critics say the challenge becomes even greater because the 3D printing industry relies on many different machines, software platforms, and open-source tools, which means firearm-detection systems could be very difficult to apply across the industry.

    3D printed ghost gun parts recovered from a March 8, 2023, search warrant by District Attorney Alvin L. Bragg. Image courtesy of Manhattan DA’s Office.

    New York has already taken a first step. California may be next. But for the 3D printing industry, the question is how far future regulations will go, and whether they can address illegal firearms without affecting the broader uses of 3D printing technology.

  • Divergent Declares that German 3D Printers are Superior, And Plans Massive LPBF Expansion

    Divergent has announced a new version of its Laser Powder Bed Fusion (LPBF) printer and a new site. The company aims to do nothing short of “further accelerating its mission to build the new industrial age.” It also says that it has “built the most advanced industrial metal 3D printer in the United States.” This implies that German and Chinese 3D printers are more advanced than American ones. This is quite an admission by the firm; after all, if it really thought it had built a better machine than EOS and BLT, it would have said it had made the world’s most advanced 3D printer. But this is sure to be welcome news to the Krailling crowd. However, I’m not sure if this implies that Divergent thinks that it has made a more advanced 3D printer than the Nikon SLM Solutions NXG. Or at least those units made in the US.

    Divergent has always excelled at marketing, partnerships, and PR. The 3D printing market, always curious about Divergent’s prospects, is now wondering what the firm is doing with its over $1.1 billion in investments. After developing large-scale glue robots, the firm then diverged to an LPBF machine. It diverged, developing a design and manufacturing service to help companies produce AM parts. High-profile partnerships followed, diverging it further from making a car.

    What is it doing to provide itself with new capabilities that others do not have? What can it do that an Incodema or Sintavia can’t? What’s the difference between Divergent’s offering and what a service does? What exactly is Divergent? Is it a late-out-of-the-gate unicorn designed to disappear, like a cohort of other firms? Perhaps there is a kind of duality between those entrepreneurs who can get money and those who excel at shipping products? Or will Divergent be able to make good on its promises to “build the new industrial age?” And why does it keep claiming to be the world’s first end-to-end software-hardware production system for industrial digital manufacturing? Clearly, Materialise had this decades ago, and lots of services have this now. What does that claim even mean? Does the Oak Ridge National Laboratory (ORNL) not have any software hardware production systems more advanced than Divergent? Has no one at ORNL or elsewhere ever built such a system? I’m done with mollycoddling everyone; it’s just not a good strategy for promoting the common good. It’s time to put up or shut up.

    The Monolith One.

    The company says its printer will have an output eight times higher. The system, called Monolith One, has 24kW and a build volume of 700 x 700 x 835 mm, while measuring 6.5 x 6.5 x 8.2 m.

    Large-Format Metal LPBF Systems Comparison

    Manufacturer & Machine Model [1, 2, 3, 4, 5] Build Volume (X × Y × Z mm) Build Volume (Liters)* Laser Array Configuration Individual Laser Power Total Cumulative Laser Power (kW) Size Status vs. 700×700×835mm
    Eplus3D EP-M3050 3058 × 3058 × 1200 11,239.5 L Custom Multi-Laser Array 500 W / 1000 W Up to 64.0 kW 27.47× LARGER
    Eplus3D EP-M2050 2058 × 2058 × 1100 4,658.9 L Custom Multi-Laser Array 500 W / 1000 W Up to 64.0 kW 11.39× LARGER
    AddUp “MASSIF” Project 1500 × 1500 × 2000 4,500.0 L Custom Multi-Laser Array 500 W / 1000 W Variable 11.00× LARGER
    Farsoon FS1521M-U (Tall) 1530 × 1530 × 1650 3,862.5 L 16 or 32 Fiber Lasers 500 W 8.0 kW or 16.0 kW 9.44× LARGER
    BLT BLT-S1500 (Max Z) 1500 × 1500 × 1500 3,375.0 L 18 or 26 Fiber Lasers 500 W 9.0 kW or 13.0 kW 8.25× LARGER
    Eplus3D EP-M1550 1558 × 1558 × 1200 2,912.8 L 16 or 25 Fiber Lasers 500 W (1kW optional) 8.0 kW to 25.0 kW 7.12× LARGER
    BLT BLT-S1500 (Base) 1500 × 1500 × 1200 2,700.0 L 18 or 26 Fiber Lasers 500 W 9.0 kW or 13.0 kW 6.60× LARGER
    Eplus3D EP-M1250 1250 × 1250 × 1350 2,109.4 L 9 to 16 Fiber Lasers 500 W (1kW optional) 4.5 kW to 16.0 kW 5.16× LARGER
    Farsoon FS1521M (Standard) 1530 × 1530 × 850 1,989.8 L 16 or 32 Fiber Lasers 500 W 8.0 kW or 16.0 kW 4.86× LARGER
    BLT BLT-S1000 1200 × 600 × 1500 1,080.0 L Up to 12 Fiber Lasers 500 W 6.0 kW 2.64× LARGER
    AMCM (EOS) M 10K 1000 × 1000 × 1000 ~1,000.0 L Multi-laser Custom Array 1000 W (1 kW) Variable 2.44× LARGER
    Nikon SLM NXG XII 600E 600 × 600 × 1500 540.0 L 12 Fiber Lasers 1000 W (1 kW) 12.0 kW 1.32× LARGER
    AMCM (EOS) [M 8K] 800 × 800 × 1200 768.0 L 8 Fiber Lasers 1000 W (1 kW) 8.0 kW 1.87× LARGER
    Eplus3D EP-M825 825 × 825 × 1100 748.7 L 8 Fiber Lasers 500 W 4.0 kW 1.83× LARGER
    Monolith 700 × 700 × 835 409.2 L 12 2 Kw 24 kW
    Velo3D Sapphire XC 1MZ Ø 600 × 1000 (Cyl) 282.7 L 8 Fiber Lasers 1000 W (1 kW) 8.0 kW 31% Smaller
    Nikon SLM NXG XII 600 600 × 600 × 600 216.0 L 12 Fiber Lasers 1000 W (1 kW) 12.0 kW 47% Smaller
    EOS [M4 ONYX](1.3.3, 1.3.6) 450 × 450 × 400 81.0 L 6 Fiber Lasers 400 W 2.4 kW 80% Smaller
    Nikon SLM SLM 500 500 × 280 × 365 51.1 L 4 Fiber Lasers 400 W / 700 W 1.6 kW to 2.8 kW 87% Smaller

    The 12-laser system will have 2 kW lasers and work with aluminum, nickel, steel, and titanium. The system was developed in-house across 28 months. The company also, rather interestingly, says that the Monolith One can produce “multi-material structures.” That is very interesting, and I’m very curious. Do they have the Aerosint multiple recoater technology, or how are they doing this? This could be a real advantage in some aerospace and armor structures. Especially if it could build these efficiently without too much change around. The Monolith One is entirely made in the USA. The company also implies that all components come from there. This should be a great relief to defense users. But, you are not allowed to buy one.

    The Monolith One.

    The company is also expanding to a 430,000 sq.ft facility in Long Beach. There are six Monoliths running, which implies that they’re at around a tenth of the capacity of several other firms in the Long Beach area. What is very exciting is that the company aims to install 64 more machines over the next two years. That would make them one of the largest operators of LPBF machines in Long Beach. Their total capacity over two years will therefore be approximately 10% of BLT’s current capacity.

    Divergent CEO Lukas Czinger said,

    “The Monolith One is the first metal 3D printer designed ground up for scaled production of critical hardware. Importantly, its design encompasses the years of operational insights we have earned delivering production structures to the defense and commercial sectors. Monolith One is an American machine with an American supply chain. We are building them at rate today and our Long Beach factory will house 64 more of them. With annual output in the tens of thousands of munitions airframes or hundreds of thousands of critical piece parts, our second factory represents the new industrial age at scale.”

    Divergent Factory in Long Beach.

    The company hopes to work in defense and use “4-axis scanners with spot-size zoom capability” to improve throughput. Divergent also says that a 1700 cm2/min gas-flow unit allows for fewer optical window washes and longer print runs. Powder handling is closed-loop. The company has worked on build plates, with “heating and novel cooling controls up to 200°C to enhance reliability, dimensional stability, and repeatability.” It has optimized turnaround times through software and interchangeable build volumes. We can’t be sure whether it’s kind of like a Farsoon continuous-build thing or if it works differently.

    Divergent’s CTO Brian Erhartic,

    “Every feature of Monolith One was engineered to maximize reliability, scalability and control,By starting from a clean sheet, our team has built an additive manufacturing solution that expands the overall performance envelope of DAPS, particularly to serve a wider customer landscape and drive efficiency into downstream operations. It’s only because we custom engineered the printer specifically for integration into DAPS, that we were able to realize a significant increase in operational efficiency, quality control, and build volume.”

    Cruise missiles for CoAspire.

    The company hopes that with the new printers in place, in two years it can make (either/or): over 30,000 missile airframes, 60,000+ warhead casings (100lb class), 25,000+ automotive subframes, or 30,000+ automotive suspension systems. I’m not sure if it’s economical to run parts like that in LPBF. But perhaps the firm can really lower costs significantly.

    Divergent states that it is valued at $2.3 billion and has raised over $1 billion, while it has impressive clients such as “CoAspire, Saab, Triumph Group, Bugatti, and McLaren.” The in-house 3D printing approach, using a software-enhanced large-format LPBF process coupled with design and sensor integration, seems promising. Indeed, Seurat, VulcanForms, and others are trying it as well. Doing this at a time when the US needs to rearm and make complex parts quickly is auspicious. The firm is also ambitious, and good at socializing and spreading its mission.

    But we’ve lived in a land of claims for far too long. I want to see Divergent succeed. I want to see those tens of thousands of parts coming out of there. But I want to see parts; I want to see volume. I’m done riding shotgun on other people’s dreams; I’m done being led to la-la land by unicorns. It’s put-up-or-shut-up time. I want to see parts. I want to compare parts and know how much is being made. It’s time to get real.

    Images courtesy of Divergent

  • Zellerfeld Buys Volumental

    Volumental is a Swedish 3D scanning company that has created, over the past decade, a scan-to-fit solution for use in stores. The company has worked with sports retailers and shoe brands such as New Balance and Hoka to make a shoe-fit solution. You can go to a store, scan your feet, and you’ll have the absolute right size. Then, afterward, you can always order from that brand to get the right-sized shoe. I used the engine once in a store, Bever, to buy hiking boots and was impressed with its speed and efficiency. The tool pointed my salesperson to narrowing down specific Meindl boots based on my foot shape. Others would not be comfortable given my high arch, saving the salesperson and me a lot of time. I really liked that as an experience. Their FitEngine uses AI to generate the right fit and can also be used at home through phones. And that sizing solution I used once to buy a pair of Hokas online. This helped me go through with my purchase because I would be more confident that they would fit, and they did.

    Now, Volumetric has been acquired by Zellerfed, the German 3D printed shoe production platform. On the face of it, this could be a revenue-generating opportunity and cost-saving project in one for retailers and shoe brands. But the firm is now being taken over by the 3D printed shoe movement, Zellerfeld. Zellerfeld could simply be interested in a technology that would be useful to itself or its market. But by becoming a YouTube for shoes, the company will probably differentiate itself not only by having many designers work on its platform, but also by offering well-fitting shoes. That, therefore, could point to Volumental itself or at least its offering being integrated into Zellerfeld.

    Beyond this, things could get very interesting. If Zellerfeld can make truly individual shoes that can conform to your foot and optimize comfort and walkability, they could have further differentiation. And Volumental can be instrumental in unlocking this potential.

    If Zellerfeld could engineer its shoes to work better on your feet, it could grow much bigger. Volumental reportedly has scanned over 66 million feet and is in over 3000 stores. So that data lead could also aid Zellerfeld. Perhaps it could, through Volumental, extend its 3D printed shoes into shoe stores or develop a fit-to-print solution for store chains?

    At first, Zellerfeld’s centralized 3D printing approach seems more likely. But perhaps an in-store printing solution, provided in part by the company, could give it an unassailable lead. Even if it does not provide access to the geometries, the overall total understanding of the variety in human feet is very valuable. And if it can accurately size your feet from home and give you better-fitting shoes, then Zellerfeld could see more customers.

    Zellerfeld has really pushed, grown, and developed its own offering and the 3D printed shoe market. I’ve always been skeptical of 3D printed shoe developments on the whole. But given Zellerfeld’s progress, we can now see a complete offering emerge. If the company can develop comfortable shoes and find the right designers, it could grow its business towards higher-volume customers.

  • Limitless Labs Raises $20 Million in Series A Funding for Agentic CAD

    Limitless Labs has raised an additional $20 million for its agentic CAD/CAM platform, bringing its total funding to date to $27.3 million. Investors in the Series A round include Dell Technologies Capital, Square Peg, Grove Ventures, Meron Capital, and Kinetica.

    Limitless Labs CEO David Priev said,

    “The manufacturing world doesn’t just need more automation, it needs a better way to capture and scale the expertise that still lives inside the heads of a relatively small number of experienced machinists. We built Limitless Labs to work inside the CAD/CAM systems manufacturers already use, helping teams standardize best practices, reduce programming bottlenecks, and free senior programmers to focus on the hardest work, without giving up control. We believe the next major AI platform will be built for the physical world, and that starts with giving manufacturers a way to scale their best knowledge across every new part and every new engineer.”

    Yair Snir, Managing Director at Dell Technologies Capital, stated,

    “Limitless Labs represents the next wave of enterprise AI, moving beyond digital workflows and into the physical world of precision manufacturing. Their unique foundation model and the caliber of their production deployments gave us conviction that this team is building the defining platform for AI in manufacturing.”

    The company says that its AI model is a,”Physical AI Foundation Model, trained not on text or generic code, but on the physics of metal cutting, CAD geometry, and the operational constraints of real machines.” The firm has a CAM Agent that can recommend tools, prioritize operations, and generate tool paths. This is similar to Toolpath.com, but in this case, the idea is that the CAM Agent works with or in established tools such as Creo, Siemens NX, and Mastercam, which saves the machinist time. Limitless Labs thinks that it can save half of the work.

    Lior Handelsman, General Partner at Grove Ventures, added,

    “Eighteen months ago, we backed Limitless Labs’ vision that agentic AI could transform the factory floor. What the team has achieved since then has exceeded expectations. They are combining deep technical innovation with practical software in a way that could reshape how the world’s most critical parts are made.”

    The company is targeting defense, aerospace, and motorsports applications, and reporting pilot programs in place with companies like Cadillac, Blue Origin, and Sandvik. Blue Origin is notable since the astronautics firm is usually less than talkative about any manufacturing information. Limitless Labs also says it can deploy in ITAR compliant set-ups. The funding it’s raised will be used for a sales ramp up, help growing headcount, improvements to its model, and new versions automating more CAM operations.

    To be fair, I expected a whole lot more of these firms a year ago. So far, we have seen dozens of CAD and authoring-based startups, and around 50 additional startups in broader CAD/CAM automation using AI. They all promise similar things: just by existing in the space, AI will save operators and companies time. There is also a lot of talk about physical AI and real world models, or models based on machines and operations. If so. where are these models being created? How are they being created? What are the datasets like? And does this mean that this particular model will be fed from your CAD/CAM files? Will your G-code, your CAD model, and your manufacturing data be used to power this startup and others? How are they getting this data, and how can they firewall my data? Because if they take tool paths or CAD, then I can reverse engineer model data from them.

    There is a huge amount of risk around these models sucking in training data, which is then reverse engineered to the individual file or machine. G-code reversing and just general usage data for specific installations would also be very valuable. But I think this is a huge security risk and would not be comfortable having any sort of AI CAD/CAM tool in any kind of professional environment at the moment. Sure, I’ll play with this stuff at home, but if it comes time to invent things, I’d never use them. With their originality and their intrusiveness, there is just too much risk associated with these models currently. Some of them may, in some distant future, save operators time, but will you risk your firm’s future or your job on this today? I wouldn’t. The AI CAD/CAM companies must explain how they’re, on the one hand, getting real world data that is relevant, while on the other hand, not sucking up their user data. And the fact that this startup can’t even manage to secure a picture of a factory floor or CNC machine doesn’t exactly make me warm and fuzzy inside.

    Images courtesy of Limitless Labs

  • Inside Haddy: Jay Rogers Wants 3D Printing to Build Real Products, Not Just Prototypes

    A warehouse from the outside, but step inside Haddy and it shifts quickly: finished pieces up front, clean and minimal, furniture you can touch and sit on. Walking through the factory, the machines take over — they are massive, loud, constant — robots stretching across the room, printing objects that don’t look like they should exist in one piece. That’s where Haddy really comes into focus, not just in the furniture, but in how everything is made. 

    That became clear when I visited Haddy in St. Petersburg, Florida. The focus here is production, getting parts made reliably and on time. That came through in my conversation with founder and CEO Jay Rogers, who explained that the company was built with one goal in mind: to produce large products at a commercial level, reliably and on time. 

    “Haddy was created to print big things at an industrial or commercial scale,” he said. Just as important, he is not interested in stopping at prototypes. The goal is finished parts. 

    That may sound simple, but it gets at one of the biggest shifts happening in additive manufacturing right now. For years, 3D printing was often treated as a way to make mockups, one-off concepts, or early design versions. Rogers believes that stage is over. 

    Haddy’s 3D printed seating and side tables, showcasing the range of forms possible with robotic extrusion printing.

    “We’re long past the day when printing something is just cool or when we can rely on doing a prototype where people don’t need to do it in production. Our end result is to take something that we can print and then take it into production.” 

    That line may be the clearest way to understand both Haddy and Rogers himself. He is not talking about 3D printing as a novelty. He is talking about it as a manufacturing method. 

    That way of thinking goes back to his earlier work at Local Motors, the company he founded before Haddy. Local Motors became known for its microfactory model and work in large-scale 3D printing, including vehicles like the Strati and the Olli autonomous shuttle, both built using polymer-based additive manufacturing. 

    But Rogers said his motivation was never just cars, and it was never just 3D printing for its own sake. 

    “I didn’t do it because I love vehicles. I didn’t do it because I love 3D printing. I did it because there was a need, and 3D printing offered a great opportunity to solve that,” he explained. 

    That same logic now drives Haddy.

    At first glance, the company may look like a furniture business. Rogers talks a lot about furniture, fixtures, and equipment, and the company’s work fits naturally into those categories. But he said that is only the entry point. 

    “We’re not just building furniture. We’re building things fit for the age of robotic production. We’re building a capability to make things better than they were done before, a way to produce things that matches how machines work today,” Rogers noted, comparing Haddy’s approach to how Amazon started with books but was really building a broader capability to sell and distribute products differently, better than the way they were done before. 

    That is a smart distinction. Rogers is not presenting furniture as the final destination. He is presenting it as a practical market where Haddy can scale. In his view, furniture is large, global, and relatively unregulated compared to aerospace or medical products. That makes it a good place to prove a model. 

    He even put a number on it, saying the global furniture market is around $700 billion, roughly the same size as the aircraft parts maintenance and repair (MRO) market, but “much easier to enter because it is not burdened by the same level of regulation.” 

    That helps explain why Haddy did not begin with aircraft parts or medical implants, even though those are the areas that tend to get the most attention in 3D printing. Those applications are complex, highly regulated, and slow to scale. Rogers sees it differently. Many engineers are drawn to the most advanced or impressive use cases first. Meanwhile, his approach is more practical: “Start where the need is big, where it’s easier to get to market, and where you can actually build a business.” 

    The company’s other big idea is local production. Rogers spent much of our conversation talking about how manufacturing moved over the decades to wherever labor was cheapest. 

    “In America and in other countries, we have relied on low labor rates to make things. Furniture is a clear example where production has shifted across regions and countries over time as costs changed. You would see furniture move from the U.K. to North America, then to the Midwest, then to the South, then to Taiwan, then to China, then to Vietnam,” he said, describing how production has shifted over decades as companies chased lower costs. “Now, that model is reaching its limits.” 

    Haddy’s answer is to bring manufacturing closer to the customer. 

    That thinking is visible on the factory floor. During my visit, I learned that the company has eight robots. One of the original robots is used for research and development. As Rogers showed me, materials are dried and then moved through enclosed channels under the floor, feeding directly into the robotic systems. Some jobs can run continuously without a person standing there the whole time. The team showed me large print beds, heated systems, molds in production, and examples of large-format output, including projects measured in feet, not inches. 

    Haddy’s material library includes a variety of recycled plastic feedstocks, colors, textures, and surface finishes.

    Clearly, this is not desktop printing scaled up a little; this is manufacturing infrastructure. 

    Rogers also emphasized that Haddy chooses to use polymer composites. People regularly ask whether the company prints in metal, ceramic, concrete, or other materials. But he said polymer composites open the broadest commercial opportunities and make more sense for the kinds of large-scale production Haddy is targeting, including work with partners like The Walt Disney Company

    “Other materials are often harder to control at scale, especially when heat, warping, and processing issues become more severe. For example, printing metal is not as useful as printing polymer composite, it’s higher heat, it has a lot more of a thermal gradient, so it tends to warp a lot more, and it’s harder to process control.” 

    That choice points to how Haddy is trying to build a practical system rather than chase every possible printing category. 

    And the same applies to automation. Rogers said the robot arm itself is not Haddy’s invention. The company buys robot arms and uses them as a foundation. What Haddy builds is the larger system around them, including the process automation and control needed to make the machines perform the same way every time. 

    That consistency matters because, as he explained, there’s no point talking about AI if the machine can’t first do the same thing the same way every time. Today’s 3D printers often just follow instructions. If something goes wrong, they don’t always catch it. Haddy is working to change that. 

    He also believes progress has not come just from better machines, but also from better sensors and controls. With simple vision systems and better data, machines can spot problems, like material building up on a nozzle, and react at the right time, instead of following fixed routines whether they need to or not. 

    “That same practical thinking shapes how we plan to grow. Instead of building one large, centralized factory, the idea is to expand closer to where things are made and used, closer to customers, partners, and local markets,” noted Rogers. “That matters for speed, but also for sustainability. Everything Haddy makes is recyclable, but once a product is shipped across the world, it becomes much harder to bring it back and reuse it. If you put your product 17,000 miles away, you will never recycle it.”

    That last thought gets at something larger. Haddy is not just selling a product. It is trying to build a manufacturing model. One where design, production, delivery, and even reuse happen much closer together. 

    3D printed tables, seating, and architectural elements produced by Haddy using recycled polymer materials.

    It’s an ambition that may sound big, but after visiting the site in St. Petersburg, it no longer felt like just an idea. What stood out most was not just the machinery, which, let’s face it, is pretty impressive in and of itself. It was the clarity. Rogers has a very specific view of what 3D printing should do next. He is not arguing for hype. He is arguing for production. He is arguing for local capacity. And he is arguing that additive manufacturing works best when it solves real commercial problems. 

    Haddy stands as one of the clearest examples yet of what happens when large-format 3D printing grows up, far beyond a furniture startup or a robotics story.

    3DPrint.com’s Vanesa Listek sits in one of Haddy’s large-format 3D printed chairs during a visit to the company’s headquarters.

    All images courtesy of Vanesa Listek/3DPrint.com

  • 3DPOD 302: Digital Inventory for AM with Mikhail Gladkikh, Würth Additive Group

    Mikhail Gladkikh has worked in oil and gas for many years. With this background, we obviously talk about energy market turbulence and the adoption of AM in oil and gas. We talk about what oil and gas companies want, how they are adopting 3D printing, and how it is being used. We also talk about MRO more generally, digital supply chains, and resilience, and manage to squeeze in a little about Würth Additive, where he now works.

    This episode of the 3DPOD is brought to you by FacFox, where your next product starts as a file and ends as a custom-made reality. With instant quoting, rapid design feedback, and on-demand 3D printing, CNC machining, injection molding, and more, FacFox makes it easier to test out ideas, fine-tune every detail, and manufacture with confidence. Curious what your design could become? Upload it and find out.

     

     

  • Student Research Raises Questions About Patient Privacy on 3D Printing Platforms

    A student researcher at Indiana University Indianapolis has uncovered what she believes is a significant patient privacy issue involving medical anatomy files shared on public 3D printing websites.

    Salma Kherallah, a junior studying at the School of Health and Human Sciences, recently won first place at the university’s Undergraduate Research Conference after examining thousands of anatomy-related files posted online for download and 3D printing. Her project, “The Availability of Potentially Unethically Sourced 3D Anatomy Models on Peer-to-Peer Websites,” was selected among 200 others.

    Working with Andrew Cale, an assistant professor at the university’s Department of Anatomy, Cell Biology and Physiology, Kherallah reviewed roughly 3,000 files from peer-to-peer 3D printing platforms. The goal was to determine whether some of the models were derived from real patient scans.

    According to the university, the researchers found examples that appeared to originate from medical imaging data, including MRI scans. Some of the files depicted human anatomy, including bones and organs, and could be downloaded by anyone with access to the platforms.

    “We were investigating whether people were uploading patient scans or patient bones or human remains up to these websites to be printed, which has a lot of ethical concerns and HIPAA [Health Insurance Portability and Accountability Act] violations,” Kherallah explained in a statement released by the university. “We think that by doing this research and proving that this is out there, maybe stricter regulation could possibly be enforced in the future.”

    The project began after Kherallah noticed anatomy models circulating online and wondered where they had come from. Using her anatomy training, she worked with Cale to evaluate the files and identify signs that they may have been generated from actual patient scans. The researchers say some of the files could contain patient information that was shared without permission.

    Salma Kherallah presents her project at the IU Undergraduate Research Conference. Image courtesy of Indiana University Indianapolis.

    Kherallah’s work was praised by Indiana University Indianapolis Chancellor Latha Ramchand, who described Kherallah as an “incredible changemaker” and said the project highlighted how undergraduate research can help address real-world challenges involving patient privacy.

    “Her work advocates for patients by identifying real MRI scans published on 3D printing websites,” noted Ramchand.

    Many medical 3D printing applications begin with patient scans. Hospitals and researchers use CT and MRI scans to create 3D models that can help doctors prepare for surgery, teach students about anatomy, or develop new medical devices. But patient information is usually protected, and personal details are removed before the actual files are shared for research or education. But once a file is uploaded to a public website, other people can download it, copy it, and share it again.

    “My biggest takeaway from this experience is that technology is allowing people to do whatever they want,” she said, posing the question, “How far are we allowing this to go? How are we regulating this type of technology to regulate the ethics and morals of society’s standards on patient privacy?”

    Kherallah’s project looked at whether patient scans, bones, or other anatomy models were being uploaded to public 3D printing websites without permission. The researchers said their findings show that questions around privacy and consent should get more attention as medical 3D printing continues to grow.

  • Scientists Use BMF to 3D Print Seal Whiskers That Track Prey Long After It’s Gone

    Seals use their whiskers to hunt. Not Navy Seals, although they may in some way also, but this article is about lowercase seals. Not Seal the musician either, as far as I know, he doesn’t even have whiskers. This is about different seals. Pinnipeds are marine mammals with flippers and blubber.

    One of the ways whiskers, or vibrissae, are used is as a flow sensor. In humans, vibrissae are nose hairs that act as filters, while in cats, they’re used to hunt in the dark. Vibrissae can be used to sense air currents, changes in pressure, or to sense things. They’re embedded in the sensory system and are very sensitive. They can tell you that a crevice is too small for you to fit in, sense the air flow of a running prey animal, help an animal in flight orient itself, and act as signals of intent or mood. Cats have around 200 nerve cells on every whisker, while seals may have 1500.

    Biomimetic seal whisker-inspired fully printed MEMS sensor. Image courtesy of Tekin et al., Microsystems & Nanoengineering (2026).

    These whiskers not only let them orient their bodies in the dark or sense rock formations, but also let seals track where prey has been. Through being sensitive to hydrodynamic trails. These trails are pressure changes and swirling vortexes in the water left behind by swimming animals. They’re also called wake-induced vortexes. These trails can show direction, speed, size, or even what kind of animal was there. So it’s kind of like tracking specific water footprints. When hunting in the dark, it’s easy to see how valuable these whiskers could be. Certain seals have evolved specialized morphologies to better track their prey.

    Seal species, whisker morphologies and sensing mechanisms. Image courtesy of Tekin et al., Microsystems & Nanoengineering (2026).

    Now researchers have made a 3D printed “artificial follicle–sinus complex flow sensor” mimicking the setup the seals have. They made a novel elastomeric resin and studied the whiskers of harbor seals, gray seals, and sea lions. They then designed and printed a compliant structure and used Boston Micro Fabrication‘s (BMF) system to 3D print the entire device in one step. Overall, the resolution was less than 10 μm. The team then put graphene nanoplatelet ink into the printed channels, turning the device into a piezoresistive sensor. Tests showed that these sensors could work for at least 3000 cycles, sensing strain in vortexes just like the seals do. The team used a GOM ATOS III Triple Scan 8 M to scan different seal whiskers to get the target geometries. They found that the harbor and gray seal whiskers were better at sensing and differentiating than the sea lions’ whiskers were. Later, an 8-cm-long whisker was tested for 3000 cycles.

    The researchers worked at the Department of Bioinspired MEMS and Biomedical Devices (BMBD) of the Engineering and Technology Institute (ENTEG) at the University of Groningen in the Netherlands. Engincan Tekin, Ming Cao, and Ajay Giri Prakash Kottapalli got their work published in Nature Microsystems and Nanoengineering.

    PμSL printed MEMS mechanosensory artificial follicle sinus complex sensory base. Image courtesy of Tekin et al., Microsystems & Nanoengineering (2026).

    The whiskers in the tests were ~80 mm in length and were made using the BMF microArch S240, the team’s own material, a blend of 3DResyns PDMS-like resin and BMF’s UTL resin mixed in a 70:30 volume ratio. Mixing resins like this is widely done in research and by hobbyists alike to try to achieve the right mix of properties. You should be very careful and orient yourself thoroughly if you’d like to try this at home or at the office.

    This is great news for people working in bioinspiration. Also, for the soft robotics crowd, this could be an exciting proximity or action sensor. This kind of sensor could also point to a working mechanism to let a temperature gradient or current follow a soft robot descend or ascend appropriately. In fact, if you would cover an entire robot in these whiskers, I bet that you could use it for sensing and navigation.

    At the same time, this work points to more possibilities for integrated 3D printed MEMS devices. This is something I’m super excited about. MEMS are great but require significant up-front investment, especially for low-production-run MEMS, or indeed for my concept of Macro MEMS and 3D printed MEMS devices. Indeed, as we wrote in 2022, a class of entirely new devices and MEMS could be created much more rapidly and inexpensively through Additive. Here, where 3D printing produces almost the entire device from a single material in a single step, we can begin to see the outline of a world in which tiny devices could power sensing, navigation, actuation, and other functions across billions of devices.