My Printer

  • Bridging the Gap: 2D to 3D AI in Manufacturing

    For decades, the early stages of manufacturing have been defined by a simple, frustrating trade-off: you can have it precise, or you can have it fast. AI just broke that rule.  Manufacturing has never lacked data, but it has consistently lacked time at the earliest stages of decision-making.

    Across engineering, procurement, and sourcing teams, critical information still arrives as technical drawings, blueprints, scanned documents, images, or even photographs shared over email. While downstream workflows are increasingly digital, early-stage decisions often depend on incomplete inputs and manual interpretation by the expertise trapped in engineers’ minds.

    This is where AI-assisted interpretation of 2D drawings and images into 3D geometry is beginning to change how teams work. Not by replacing CAD, but by removing it as a bottleneck when speed matters more than production-ready precision.

    Why 2D Still Dominates Early Manufacturing Decisions

    Despite decades of CAD adoption, many manufacturing workflows still begin without a 3D model. A supplier receives a dimensioned drawing but no STEP file. A procurement team needs a cost estimate before engineering resources are available. A sales engineer must respond to an RFQ having only a PDF attached.

    In these situations, the objective is not production-ready details. It is speed, feasibility, and direction. Can this part be manufactured? Which process makes sense? Is the cost even in the right range?

    Traditional CAD workflows are not designed for this stage. Creating a fully parametric, production-ready model, can take hours, sometimes days. For early estimation, that effort is often disproportionate to the decision being made.

    What 2D-to-3D Conversion AI Actually Means

    Recent advances in AI now make it possible to convert 2D inputs into usable 3D representations in minutes. Importantly, this does not mean generating perfect CAD models. Modern systems now automate the leap from flat drawings to 3D meshes. While these meshes are only an approximation, they capture the proportions, shape and volume accurately enough to drive immediate cost estimation and decision-making.

    3D Spark bridges the gap before traditional CAD is even necessary, this AI driven conversion is positioned as a pre-CAD tool. The goal is not to replace engineering work, but to eliminate unnecessary delays in early quoting and feasibility analysis.

    Image 1: 3D Spark’s 2D drawing to 3D feature

    Image 2: 3D Spark’s Image to 3D feature

    Input Flexibility Reflects Manufacturing Reality

    One of the most practical aspects of this approach is input flexibility. Rather than relying solely on clean technical drawings, AI-assisted systems can work with:

    • Technical drawings
    • Standard images and legacy photos
    • Hand sketches or basic text descriptions

    This matters because real-world inputs are rarely ideal. By interpreting different 2D sources and converting them into AI-generated 3D geometry, teams can move forward faster and more efficiently.

    The Output: 3D Geometry for Cost Estimation and Production Technology Comparison

    The resulting scaled 3D-mesh is not suitable for direct CNC machining, tight-tolerance manufacturing, or toolpath generation. When users ask, “Can I machine from this file?” the correct answer is no. But that is also the point.

    Where the Real Value Appears: Cost and Process Estimation

    Image 3: Automated costing analysis using the 3D Spark platform

    Once 3D Spark converts the 2D input into a 3D mesh, it doesn’t stop there. The platform immediately uses that approximate geometry to predict material usage, production time, and process costs and therefore turning a static image into a calculated business case.This allows teams to perform cost estimation based on the 2D input in minutes rather than hours.

    Instead of delaying RFQs or relying on assumptions, teams can quickly assess whether a part should be machined, additively manufactured, cast, or sourced externally by identifying the cost drivers early, before committing engineering time to detailed design work.

    While AI-assisted 2D-to-3D interpretation accelerates early estimation, more detailed costing and feasibility analysis are often required as decisions progress. Platforms like 3D Spark extend this workflow by supporting accurate cost calculation, and manufacturability assessment based on full 3D data and production-specific parameters, allowing teams to move from initial direction to validated decisions without restarting the process.

    This continuity is particularly valuable in MRO and spare parts workflows, where early decisions must translate directly into execution without rework or lost time.

    Why This Matters Now

    Manufacturing teams are under pressure to move faster, quote faster, and make better make-or-buy decisions with less information. AI-driven 2D to 3D conversion does not solve everything, but it solves a very real problem that has existed for decades.

    This reflects a broader shift in manufacturing, where AI is proving valuable not only during design work, but by accelerating the decisions that happen before design even begins.

    3D Spark is a Bronze Sponsor for Additive Manufacturing Strategies (AMS), a three-day industry event taking place February 24–26 in New York City. The conference brings together industry leaders, policymakers, and innovators from across the global AM ecosystem. Registration is open via the AMS website.

  • StoneFlower 3D Launches Laboratory-Scale 3D Printer for Construction Materials

    StoneFlower 3D has launched a new 3D printer designed for laboratory-scale research and development with concrete, mortars, clays, and other advanced mineral materials. The system is intended for researchers, designers, and engineers who want to work with real construction materials in a controlled laboratory environment, without moving straight to large industrial machines.

    StoneFlower 3D’s laboratory-scale 3D printer.

    Based in Munich, StoneFlower 3D says the new printer is built to bridge the gap between small-scale laboratory testing and full industrial production, allowing users to develop, test, and refine real construction materials before scaling up. Unlike many compact systems that rely on simplified materials, this printer can process real concrete mixtures, mortars with aggregates up to 6 mm, fiber-reinforced materials, foamed concrete, clays, porcelain, earth, and even certain biomaterials.

    “This system enables researchers to test real concrete and mortar formulations using professional pumping and mixing equipment, while maintaining a compact and flexible laboratory footprint,” noted Anatoly Berezkin, founder of StoneFlower 3D.

    Anatoly Berezkin next to a laboratory-scale 3D printer.

    One of the key features of the new printer is its customizable build volume. Instead of offering a single fixed machine size, StoneFlower 3D lets customers choose the printing area size to match their research needs and available lab space. Printing volumes can range from about 50 cm to 300 cm, depending on the configuration.

    At the center of the system is a mixing print head that can handle both single-component and two-component materials, such as cement combined with an accelerator. The print head mixes the material continuously to keep the flow stable and can deliver up to 3 liters per minute, with printing speeds of up to 150 mm per second. This makes it possible to work with fast-curing and more complex materials that are difficult to process on simpler lab printers.

    StoneFlower 3D’s laboratory-scale 3D printer in action.

    The printer uses industrial-grade hardware and standard G-code, making it easy to operate and compatible with common slicing software. Users can control the system through a touchscreen or web interface, with a ready-to-use Cura setup included. Also, the system works with different material pumps, including a high-capacity mortar pump and a smaller ram extruder. This allows users to choose the setup that best fits their materials and printing process.

    StoneFlower 3D expects the printer to be used across several areas. In research, it can support the development of new construction mixtures, composite materials, and biocompatible cements. In architecture and design, it can be used to create complex models and prototypes of façade elements or structural details. The system is also suitable for prototyping and small-batch production of functional concrete parts.

    Pricing for the base configuration, which includes the printer frame and mixing print head, starts at an estimated €13,900. Final pricing depends on the selected build volume and pump system, and each order includes operator training.

    StoneFlower 3D’s laboratory-scale 3D printer.

    StoneFlower 3D is not alone in targeting this space. Over the last few years, a small but growing group of companies has begun offering laboratory-scale concrete and mineral 3D printers, mainly aimed at universities, research labs, and early-stage product development teams. These systems exist between small clay or ceramic printers and full construction-scale machines from companies like COBOD or ICON, which are designed for printing buildings and large structural elements.

    What makes this segment different is the focus on real material testing rather than final construction. Researchers and designers want to test real concrete mixes, fibers, aggregates, and fast-curing materials without having to invest in large industrial systems. Companies such as Deltasys E-Forming and Eazao also serve the laboratory-scale concrete 3D printing market, with systems aimed at research, education, and early-stage material testing. Meanwhile, companies best known for construction-scale concrete 3D printing, such as AC3D, have also introduced smaller platforms for research and testing, alongside their larger building-focused systems.

    Applications for the laboratory-scale 3D printer.

    For StoneFlower 3D, the emphasis is on flexibility. By offering customizable build volumes, support for industrial-style pumps, and both single- and two-component material processing, the company is positioning its system as a practical research tool rather than a scaled-down construction printer. As interest grows in printed concrete and other construction minerals, laboratory-scale tools help researchers, material developers, and designers explore and refine materials and processes before they move into full-scale industrial production.

    Images courtesy of StoneFlower 3D

  • Low-Temperature 3D Printed Shape-Memory Stents Activated at Body Temperature

    Researchers from Waseda University, the University of Tokyo, the University of Tokyo Hospital, Southeast University, and the South China University of Technology have worked together on developing low-temperature 3D printed vascular stents. Published in Advanced Functional Materials, “Adaptive 4D-Printed Vascular Stents With Low-Temperature-Activated and Intelligent Deployment” is important.

    A made-to-measure functional stent geometry that can change shape would be easier to implant and could make stent procedures safer and easier, reducing surgical risk and the need for surgery. In this case, the team used my favorite material, polycaprolactone (PCL) and DEP (diethyl phthalate) as a plasticizer. PCL is often used for applications such as trachea fabrication; it is extruded somehow, but in this case, they used a micro-stereolithography (PµSL) machine. The material was optimized to have a class transition temperature at body temperature, which is smart.

    The stent they made is one meant to be used as a vascular stent. The stent is strong, very elastic, and biocompatible, and is meant to evolve into an implantable device. Used in the treatment of coronary heart disease, stents are a major business. What’s more, PCL is bioabsorbable, so the material could, in future iterations, be made to be completely reabsorbed.

    Now the team, led by Waseda University’s Shinjiro Umezu, is making a stent that heats and changes shape once it’s in the body through body heat. I love this. The team used Blender to slice the stent, which is a powerful tool but would not be my first choice for this application. I couldn’t find the name of the 3D printer used.

    The 10 mm-diameter, 10 mm-high stents were designed with 1 mm-diameter channels. First, the stents were put in 70°C water, then put in cold water for 5 minutes, and fixed. After they were reheated in water at 37 °C. The team also did in vivo experiments in mice as well as in vitro studies on umbilical cells.

    Professor Shinjiro Umezu explained,

    “Our work provides a robust platform for next-generation adaptive vascular stents with programmable mechanics, intelligent deployment, smoother integration with human body, and reduced need for complex procedures, offering significant potential for personalized treatment in anatomically complex vascular structure. Consequently, our research could contribute to future vascular stent technologies used in minimally invasive procedures, potentially simplifying deployment and reducing the need for additional equipment. The same approach may be applicable to other implantable medical devices that are designed to respond to the body’s natural environment.”

    I think this is a great development. 3D printed PCL components could be a very safe alternative to a lot of therapies and devices currently. The material is very safe and has some wild properties. If you wanted to make a stent that could then subsequently bioabsorb once the treatment is done, then this could make things even easier and safer for patients. Just by making the stent change back to its shape in the body, the team has made a step forward. This could, if proven out, lead to a huge industry around shape memory stents. We do not yet know what industrial acceptance will be like, but if these could be coupled with simple, relatively safe procedures, they could have a lasting impact on the medical market.

    Polymers in the body still scare a lot of companies. But in this case, we’re talking about a material that is safer and more temporary than others. I can really see a whole host of treatments emerge around these kinds of systems. For treatments that have to do with muscles and tendons, something like this could make a lot of sense as well. For oral and other cancers, such custom structures can also provide relief. I really think that this team is on to something, and I hope that much more research will follow.

  • CASF: A Green Surface Finishing Technology for AM Hard Metal Alloys and Fatigue Improvement

    Sugino Machine Ltd has recently completed development of a highly specialized surface-finishing technology capable of removing partially melted particles, debris, and alpha case left behind by additive-manufactured (AM) laser powder bed fusion (LPBF) of titanium alloys such as 6Al-4V. Cavitation abrasive surface finishing and peening (CASF) goes beyond line-of-sight processing methods such as tumbling or grit blasting, since the powerful shock-wave action occurs wherever the imploding water-cavitation vapor bubbles can be directed to enter and activate. Because of this unique omnidirectional capability, CASF can treat very long circuitous internal passageways, as well as the walls of drilled holes, bores, cornices, cut tunnels, tubular channels, return flanges, overhangs, and other deep trapped chasms.

    CASF was derived through high-speed camera studies of the water cavitation effect created by the powerful club that is used as a weapon by the Mantis Shrimp.

    Cavitation water jet peening was originally derived from high-speed imaging studies of the highly evolved Mantis Shrimp, whose specialized club generates an energetic cavitating cloud in seawater as it accelerates at a rate equal to that of a bullet fired from a gun. When the Mantis’s punch wave blast strikes the outer shell of its prey, the result is instant obliteration, and that allows the creature easy access to the delicious meal that it craves.

    Cavitation abrasive surface finishing and peening (CASF) is created by shock waves that are generated by a nozzle acting through a slurry of water and abrasives upon a workpiece.

    Sugino’s CASF process is conducted inside a fully automated CNC machine that contains a fluid chamber holding an agitated slurry mixture of water and abrasives. The commercially available cutting media types used are environmentally benign ceramics, such as garnet or alumina. Inside the processing tank, the slurry is energized by a cloud of thousands of tiny water cavitation vapor bubbles, which violently implode as they undergo a phase transformation from vapor back to liquid. The wave action created by cavitation also kinetically energizes the abrasive particles – giving them motion and ballistic power.

    The abrasives’ sanding action slices and blows off the partially melted AM particles very quickly. As the clean, bare metal is exposed, the water constituent of the slurry strikes the surface, imparting compressive residual stresses. Wave action has similar results to shot peening, but at the molecular scale of water (H2O), with a 1.7 Angstrom size vs. the typical shot peening media diameter of ~1 mm +/- ½ mm, and molecular water about 1/10,000,000th the size of shot media for comparison. CASF does not leave behind the crater-like impressions associated with shot peening, and because metal is not displaced to the same degree, dimensional distortion of AM parts is virtually eliminated.

    The inside of tubular AM LPBF Ti 6Al-4V test samples built with various diameters, necks, and angles of attack were treated with CASF.

    Recent testing conducted on AM LPBF titanium 6Al-4V tubular parts with partner universities and industrial partners has shown that CASF will yield an internal and external surface roughness average Ra around 3-4 um, when starting from Ra 10-26 um hot isostatic pressed (HIP) test samples. Various additional prototype parts representative of production hardware have also been successfully processed.

    The CASF method has been shown to work equally well on other alloys, such as Inconel, stainless steel (CRES), high-carbon steels, and aluminum.

    AM surface quality varies on as-built parts due to the effect that gravity has during the solidification of melted particles.

    Compressive stresses, beneficial for avoiding the initiation of fatigue crack starts, have been observed to be in a range of -300 to -550 MPa when starting from nearly 0 MPa up to +250 MPa, even when measured at points that are deep inside trapped areas. Fatigue test data show significant cyclical life improvements when CASF is compared to as-built and HIP, and it is approaching parity with surfaces finished with conventional methods such as sanding, machining, and chemical milling using legacy substances.

    One of the fundamental challenges associated with AM has been post-process finishing, which can involve the extensive use of not-quite-effective automated abrasive machines, line-of-sight grit blasting, micro-machining, hand working (repetitive motion) methods, and/or the use of toxic synthetic mixtures such as nitric hydrofluoric acid or etchant solutions for post-printing surface finishing.

    CASF is completely benign and is a green alternative to using harmful chemicals. The slurry used by CASF is non-toxic, is safe for direct human contact, and does not require the care and disposal of industrial waste. Process water is filtered and recirculated within the machine. A clear-water spray and hot-air drying are all that are required for post-CASF cleaning of AM parts.

    Sugino is currently seeking collaborative opportunities within the AM industry to test hardware and parts and demonstrate the capabilities of CASF as an alternative to conventional methods.

    About the Author:

    Dr. Daniel G. Sanders is Vice President of Research and Technology at Sugino Machine Ltd. and an Affiliate Professor of Mechanical Engineering at the University of Washington. His work focuses on advanced manufacturing processes, surface finishing technologies, and the mechanical performance of engineered materials, with particular emphasis on applications in additive manufacturing and high-performance metal components.

    Sugino Machine Ltd. is a Silver Sponsor of Additive Manufacturing Strategies (AMS) 2026, which will take place February 24–26 in New York City. AMS brings together industry leaders, policymakers, and innovators from across the global additive manufacturing ecosystem. Registration is now open.

    Images courtesy of Sugino

  • Material Hybrid Manufacturing is 3D Printing Conformal Batteries for Drones

    Since the beginning of the decade, it seems like at least once a year, there will be a story about VC funds pouring money into some previously unknown startup that has figured out a new way to 3D print batteries. So far, from what I can gather, there still hasn’t been a ton of commercial success for any of the companies claiming to have hit the nail on the head with their respective proprietary processes. But it’s easy to understand why the dream persists: if you could use additive manufacturing (AM) for batteries, it would open up a wholly new frontier for supply chain autonomy.

    Yet, there’s good reason to hope that the latest company in the spotlight for its AM battery process may have differentiated itself with its core value proposition. Material Hybrid Manufacturing was co-founded by Gabe Elias — also the company’s CEO — whose work for both legacy (Mercedes) and disruptor (Rivian) auto brands taught Material what it shouldn’t be trying to print.

    After initially planning to target the EV space, Elias and the rest of the Material team quickly realized that car batteries don’t provide the best opportunities for leveraging AM’s advantages. The conformal geometries that can be achieved with the company’s Hybrid3D platform simply aren’t necessities for spacious automotive bodies. On the other hand, objects that tend to come in much smaller packages, like drones and wearables, represent the perfect product-market fit for what the Hybrid3D can do.

    At the beginning of January, Material raised $7.1 million in a seed round, not long after receiving a $1.25 million Air Force contract. The company will work with Performance Drone Works to demonstrate a proof-of-concept that Material claims can increase energy density by 50 percent, enabling users with the flexibility to either increase flight range or decrease the weight of the battery pack.

    In an article in IEEE Spectrum, Elias explained, “Things are shrinking, so we’re shrinking around it. Electronics are becoming embedded, consolidated, optimized, and batteries are the only part of the equation that’s being left behind.

    “We’re turning energy storage into a subsystem, just like all the other subsystems…The more complex the pack, the more value we capture from part consolidation and system integration, so those applications actually carry higher margins for us.”

    What Material Hybrid Manufacturing is doing reminds me a bit of Kupros, Inc., whose founder, Ian Ramsdell, I interviewed at the end of last year. As Ramsdell told me last year about his company’s unique metal filament that’s compatible with cheap desktop machines, “The best part of what we do, for me, is that by opening up the design possibilities, we’re ultimately opening up the design imagination of the end-users, as well.”

    Image courtesy of Material and Nimble

    Material seems to be doing much the same with its tech, and it’s worth pointing out that the company is already succeeding with some commercial applications, including foldable chargers that Material made in partnership with tech accessory brand Nimble. This is precisely the sort of activity the Pentagon wants to see from the emerging generation of dual-use startups garnering R&D contracts.

    If Material is able to translate its tech into a deployable platform, the company could provide one of the last missing pieces of the puzzle needed to truly scale the U.S. military’s autonomous frontline drone production ambitions. Even without ruggedized Hybrid3D systems, though, Material’s business model has the potential to significantly enhance the Pentagon’s ability to build up the capacity for domestic drone output.

    Going back to the idea of changing how product designers think, the greatest changes in the drone market in the near future may come on the civilian side. Given how untapped this market still is, we can expect new ideas to come and go at a rapid pace throughout the rest of the 2020s, as consumer preferences determine the trajectory of a new industry in real time. The ability to print conformal batteries at scale could become a pivotal factor in deciding how that story unfolds.

    Featured image courtesy of Material Hybrid Manufacturing

  • Takeaways From MILAM 2026: Defense’s Growing Role in Driving 3D Printing – Part I

    The annual Military Additive Manufacturing Summit & Technology Showcase (MILAM 2026) once again brought together the defense sector’s top technologists, military leaders, and additive manufacturing (AM) innovators for three days of industry discussions about the role of 3D printing in shaping the future of U.S. and allied defense capabilities.

    Held at the Tampa Convention Center from February 3-5, MILAM emphasized the Department of Defense’s push to “operationalize additive manufacturing,” from weapons systems and sustainment to logistics and expeditionary readiness.

    RTX booth at MILAM 2026. Image courtesy of 3DPrint.com.

    What the Defense Sector Is Asking For

    The message from MILAM was clear: the defense sector isn’t just looking at additive for low-volume or pre-production parts. It wants additive manufacturing that can scale, deploy rapidly, and integrate into existing production flows.

    And in 2026, this comes as no surprise, as industries from aerospace and energy to industrial equipment and automotive are asking for the same thing: faster qualification, repeatability, and real production output from additive technologies.

    Stratasys‘ Vice President of Industrial Business, Foster Ferguson, said the past year has marked a shift in how companies are using additive manufacturing. While low-cost printers have found a place in basic prototyping, he said organizations focused on qualification and scalable production continue to rely on industrial-grade systems. That shift, he noted, is beginning to drive consolidation across the industry.

    That evolution doesn’t mean the challenges are solved. Nikon AM Synergy’s Pedrum Sodouri, VP of Business Development, outlined what’s holding adoption back:

    “Defense wants additive manufacturing to move a lot faster than it’s progressing today. AM has potential barriers around part applicability and qualification that still need to be broken down before the technology can truly serve the heavy lift of thousands of parts needed across the Army, Air Force, Navy, and other services. I consider that tools like AI-driven evaluation or material substitution specifications could help shrink the evaluation timeline for what’s printable.”

    Velo3D team at MILAM 2026: Eric Cohen (Sales Director), Michelle Sidwell (CRO), Brice Cooper (VP of Defense).

    Similarly, Michelle Sidwell, Chief Revenue Officer (CRO) at Velo3D, said the focus is now on production: “We’re now at that tipping point where we’re really getting into production and how do we go faster? And scaling repeatable, qualified processes needs to happen faster than before to meet defense needs in sustainment and field use.”

    Brice Cooper, Velo3D’s Vice President of Defense and Government Relations, added, “It’s encouraging to see the level of attention defense leaders are giving to additive manufacturing and advanced manufacturing more broadly. There’s growing interest in how the industrial base can help move modernization faster. The focus on autonomous systems is especially important because it gives additive manufacturing a chance to be used and proven in real operating environments with less risk. That kind uinds of experience helps build confidence in the technology over time.”

    Velo3D’s booth at MILAM 2026.

    Production Applications on Display

    Companies at the event pointed to applications that are already in use. At Azoth 3D, Mechanical Engineer Luke Bristoll pointed to a fuel manifold that shows how additive manufacturing can replace complex assemblies with a single part. Bristoll showed me the part, which was originally made from 43 separately machined pieces, but can now be produced as a single component using metal binder jetting. The result is a lightweight part designed for use in the field, capable of powering soldiers’ electronics for about a week without the need to carry large numbers of batteries, which add too much weight.

    Munition components by Azoth 3D showcased at MILAM 2026. Image courtesy of 3DPrint.com.

    Meanwhile, at the REM Surface Engineering booth, CEO Justin Michaud described work done with the U.S. Air Force to address another production challenge: fully blocked internal channels in complex parts.

    “We developed a way to selectively target the powder with minimal wall removal, allowing parts such as heat exchangers to be finished or recovered without damaging their internal structures. Together, these examples show how defense needs are pushing additive manufacturing beyond prototypes and toward production-ready solutions that could also apply to aerospace, energy, and industrial equipment,” Michaud said.

    REM Surface Engineering booth, CEO Justin Michaud.

    If there was a central theme at MILAM 2026, it’s that defense is accelerating additive manufacturing from innovation into implementation. Instead of focusing on what might be possible someday, defense is using additive manufacturing for real applications today. That sets a high bar for the technology and helps show where it can work in other industries as well.

    Images courtesy of 3DPrint.com

  • Hardware is Dead. Here’s What Actually Wins in Additive Manufacturing.

    Hardware is rapidly commoditizing across additive manufacturing. Specifications have converged. Price competition has intensified. Margins have compressed. For companies attempting to scale additive manufacturing beyond prototyping, this shift has profound consequences.

    Yet within this competitive landscape, some companies are building durable advantages that grow stronger each year. The differentiating factor is no longer the machine itself. It is the software layer that transforms commoditized hardware into intelligent manufacturing systems.

    The Bambu Lab Proof Point

    No example illustrates this shift more clearly than Bambu Lab. In roughly 3 years, the Shenzhen-based startup captured a significant share of the global FDM market. The conventional narrative credits aggressive Chinese pricing. This interpretation misses the point entirely.

    Bambu Lab did not win by inventing novel hardware. Stepper motors, linear rails, and heated beds are widely available to anyone and easily replicable. What they built was a superior software experience based on automatic calibration, AI failure detection, and seamless cloud integration. Setup that once took hours now takes minutes, and the software makes 3D printing effortless.

    The consequences for established Western OEMs have been severe. Companies that had dominated for decades saw their positions collapse. Better kinematics and superior thermal management provided no defense against a competitor whose software simply worked better. Hardware differentiation alone proved increasingly insufficient.

    Software Compounds. Hardware Depreciates.

    Hardware businesses face a structural challenge: every machine shipped begins depreciating immediately. Competition drives prices down, components commoditize, and this cycle repeats.

    Software operates under fundamentally different economics. Each deployment generates data. The data improves models and processes. Improved performance attracts more customers, which in turn generates more data. The flywheel accelerates.

    A competitor can reverse-engineer hardware in 18 months. They cannot reverse-engineer ten years of compounding process data.

    At AMT, our 650+ systems across 40 countries continuously generate proprietary process intelligence. Thermal profiles, chemical concentrations, cycle parameters, and failure modes are captured across different production environments. Every edge case our systems encounter makes the entire platform smarter. The machines matter, but they increasingly serve as a means of data collection and intelligence deployment rather than the primary source of value.

    From Selling Machines to Selling Outcomes

    The software layer also transforms commercial models. Traditional hardware sales force customers to bear all risk: CapEx purchase, maintenance contracts, downtime costs. The vendor’s incentive ends at the point of sale.

    AI-enabled, data-driven systems change this equation. Real-time monitoring and predictive analytics allow vendors to offer outcome-based models such as pay-per-part pricing, guaranteed uptime SLAs, pricing that flexes with actual usage and performance. The vendor can confidently underwrite these models because AI predicts failures before they occur and optimizes processes continuously.

    This shifts the total cost of ownership dramatically in the customer’s favor while creating recurring revenue for vendors. Hardware-only companies cannot offer this because they lack the data infrastructure to understand how their machines perform in the field. The software layer enables commercial models that hardware alone never could.

    Our Mission: Finish Manual Finishing

    At AMT, we have made this transition. We think of ourselves now as an AI company with a hardware delivery model. Our mission is to deliver intelligent surface finishing for autonomous manufacturing. Our vision is simple: to finish manual finishing.

    And we practice what we preach. Internally, AMT runs on custom AI systems, from customer service to HR to operations. We’re not just selling AI-enabled products. We’re an AI-enabled company. Top to bottom.

    This perspective influences everything we build. The same intelligence that optimizes customer processes informs how we operate internally, reinforcing a feedback loop between deployment, learning, and improvement.

    Three Questions That Reveal Real Value

    When evaluating any manufacturing company, as investor, customer, or competitor, hardware specifications now provide diminishing insight. We should ask ourselves instead:

    1. Where does the intelligence live? In hardware that can be copied, or in software and data that compound over time?

    2. What data compounds over time? Every hour of operation should make the system smarter.

    3. Could Shenzhen replicate this in 24 months? If yes, there is no durable advantage.

    The Path Forward

    Manufacturing has always been about outcomes: parts that meet spec, delivered on time, at a cost that works. For decades, better hardware was the path to better outcomes. That era is ending.

    The companies that will dominate the next decade are those building software platforms that guarantee outcomes, not just technically, but commercially. The machine becomes a node in an intelligent network. The data becomes the moat. The software becomes the product.

    The question for every company in additive manufacturing: will you recognize this shift early enough to adapt, or will you be the next cautionary tale?

    ****

    This article builds on ideas explored by Pawel Slusarczyk in Hardware alone is not enough (RECODE.AM #31). His analysis of software-defined manufacturing crystallized a thesis I’ve been developing since watching Bambu Lab reshape our industry. I recommend reading his original piece.

    ****

    Joseph Crabtree is the founder and CEO of Additive Manufacturing Technologies (AMT), which he established in 2017 to enable additive manufacturing at scale through AI-driven automation and robotics in post-processing. With a background in Materials Science and Engineering and more than 20 years’ experience in aerospace, defense, and manufacturing, Joseph has led AMT’s growth into a profitable global hardtech company. Its patented PostPro technologies are now deployed in over 50 countries, helping manufacturers transition additive manufacturing from prototyping to true industrial production.

    At Additive Manufacturing Strategies (AMS) 2026, Joseph will participate in a panel on “Advances and Trends in Software and Automation for AM” on February 24th, and give a talk about “The Commoditization of Hardware and the Rise of AI in AM” on February 25th. These sessions are part of the broader AMS 2026 conference, which brings together industry leaders, policymakers, and innovators from across the global AM ecosystem. Learn more and register here.

  • Uptool Emerges from Stealth

    Uptool has come out of stealth mode. The company has raised $6 million from prominent investors and hopes to become an indispensable tool for manufacturing. For now uptool is providing quoting software for small to medium manufacturing services.

    By creating faster workflow software that automates calculations that have to be done by people, the company hopes to make small businesses more efficient. Small businesses are often bogged down by a lack of people, a few key people being nodes where a lot of processes come together, and lots of stupid work like responding to RFQ´s again and again. The small and versatile small business often struggles to cope with being overloaded or repetitive, tedious workflows. If those workflows also have to be carried out by your key finance person, Tom, or your CEO, Jane, and only they, critical people can be kept from sales, improving the business, and interacting with customers. Instead, they are bogged down by invoicing, filling out forms, or other rote work. This is painful for them and the business. Not only is it annoying, but it is also making a flexible firm less flexible and less able to improve itself.

    Based in San Mateo and founded in 2024, Uptool has raised $6 million from notable VC’s Eclipse, Kleiner Perkins, Bessemer, and Khosla Ventures. The founder of Uptool is Velo3D founder Benny Buller, who said,

    “We are building an AI platform to boost their business and bring their entire operations into the modern era. It’s been incredibly rewarding to see our initial customers rapidly grow their sales and free them from their desks so they can spend more time with their customers or on the shop floor manufacturing parts.”

    The company gives people a tool that they can sign up for and reportedly set up in an hour. Quoting, revisions, and communications are available through dashboards with AI tools organizing CAD and other files, BOMs, and other key bits of data. At the same time, calculation tools are supposed to make quoting easier.

    Velocity CNC CEO Nathan Dillon noted,

    “I would spend hours a day quoting — time we don’t charge for, Uptool has enabled me to cut my quoting time 10x in 2025, contributing to my biggest sales year to date – more than double the previous year.”

    The other founder is Alex Huckstepp, previously of Machina Labs, Arris, Digital Alloys, and Carbon.

    Eclipse partner Charly Mwangi explained,

    “Nearly 98% of U.S. manufacturing firms are SMBs — representing over 40% of the industrial workforce — but their capacity is fragmented and largely invisible, Uptool digitizes this long tail of manufacturers, injecting speed and transparency into the supply chain. That’s the only way reindustrialization actually scales.”

    An AI-powered Salesforce, but for manufacturing, sounds like a tempting play. There is real pain, waste, and annoyance at small firms around quoting and communication management. Often, many RFPs/RFQs receive fewer responses than they should, and firms are often unable to quote on or respond to all opportunities. By making CEOs themselves more efficient and lightening their workload, the firm places itself closer to money and the decision maker. Paul will decide if he wants to pay this fee. If Paul’s workload decreases, he decides to purchase the tool and picks up the phone when someone calls. This means that Uptool’s uptake could be quick. The tool could also be highly cost-effective since it is meant to be a bridgehead to more complex value-added products that will connect the client to more tools.

    The company can afford to be a great tool at low cost as long as it becomes the connective tissue for machine shops and manufacturers. Once it is, it can become very sticky and difficult to dislodge. Then it can slowly offer more sophisticated tools at higher rates and make a mint. It’s beautiful, a great play. I really hope that this tool goes on to make it easier for small manufacturers to become more efficient and win.

  • Lockheed Martin Ventures Make Strategic Investment in Perseus Materials’ Large-Format Composite 3D Printing Vision

    The VC shift towards increased funding in geopolitical bottlenecks looks less and less like a fleeting fad and more like a tectonic shift in where global investment dollars are placing their bets for the long haul. SWISSto12’s €73 million haul at the end of January is an excellent example: as Joris Peels noted, the satellite component disruptor is attracting the sort of investor who looks at the global manufacturing order and asks themselves, “…what if you could use 3D printing as a lever to change the world? What if you can own an application and, in so doing, help nations determine their own fate?”

    Another good example was Caracol’s $40 million Series B round last year, which reinforced the momentum that’s carrying a wave of large-format, robotic-arm systems to a position of higher stature within the additive manufacturing (AM) industry. We’re seeing this wave continue in 2026, and the latest proof is that Lockheed Martin Ventures has made a strategic investment in large-format robotic arm composite AM company Perseus Materials, a Knoxville-based startup that has been backed since its early stages by Roadrunner Venture Studios.

    Roadrunner Venture Studios in fact epitomizes this environment in which VC is moving from “virtual,” software-driven plays to “physical,” hardware-centric investments. The studio’s co-founder, America’s Frontier Fund (AFF), ambitiously states that its mission is nothing less than “…to build the capacity needed for America to endure as the world’s best place for innovators to reach for new frontiers.”

    Perseus Materials certainly fits that description, with the company’s co-founder and CEO, Daniel Lee, telling 3DPrint.com’s Vanesa Listek in a recent interview, “We’re not trying to make 3D printing a little better. We were asking why some of its core limitations exist in the first place.” Specifically, as Listek describes, Perseus leverages a principle known as ‘frontal polymerization’ to accelerate the resin’s drying process without a need for costly peripheral infrastructure like curing ovens.

    Perseus Materials plans to use the Lockheed investment to begin expanding both its physical footprint and its personnel, as the startup begins to fulfill its first orders. In addition to Perseus’ focus on wind turbines, the company has also been exploring the viability of its tech for naval applications.

    In a press release about Lockheed Martin Ventures’ strategic investment in Perseus Materials for an undisclosed amount, Lockheed Martin Ventures’ VP and general manager, Chris Moran, said, “Our work at Lockheed Martin Ventures supports promising companies that expand the U.S. industrial base and advance innovative technologies for the future of national defense. Perseus’ innovative composite production process can help accelerate design and prototype manufacturing while reducing costs and eliminating tooling, helping Lockheed Martin accelerate its ability to meet the needs of the Department of War and our nation’s warfighters.”

    Adam Hammer, CEO and co-founder of Roadrunner Venture Studios, said, “Perseus is exactly the kind of company Roadrunner exists to build — a breakthrough technology born from deep science with clear implications for security and competitiveness. Dan and his team are solving a foundational manufacturing bottleneck that has held back entire industries. This is the kind of innovation that can reshape how America builds at scale.”

    The most intriguing angle to Lockheed’s investment here is that there’s no need to view it as a sign that Perseus will be pivoting from wind to defense. First off, Lockheed has extensive experience in providing clean energy solutions, including offshore wind energy, which would combine both of Perseus’ key areas of interest, maritime and cleantech.

    Secondly, via Lockheed’s GridStar Flow technology, the defense prime has already demonstrated an exemplar of the fact that there is increasingly no distinction between defense and energy resilience. At the end of 2024, for instance, Lockheed completed installation of a GridStar Flow system at Colorado’s Fort Carson. This aligns with the argument I made in a recent post about how AM should be utilized as an enabler of sustainability-as-security.

    Thus, by focusing on grid resilience, Perseus is already leaning into the contemporary national security imperative in its truest form. If that technology can also at some point be used for structural components of U.S. naval vessels, that will be icing on the cake.

    There are many companies trying to do similar things to what Perseus is doing in terms of its core tech, but there aren’t nearly as many as you’d think who are applying it to the precise areas of the economy that Perseus is targeting — and that’s what counts. If you factor in the proximity of Perseus’ headquarters to both Oak Ridge National Laboratory (ORNL) and the Institute for Advanced Composites Manufacturing Innovation (IACMI), the company is perfectly positioned to capitalize on the emerging public-private consensus surrounding the need for enhanced grid stability.

    Images courtesy of Perseus Materials

  • The Real ROI of Personalized 3D Printed Medtech in Oncology

    Discover how patient-customized 3D printed devices like Stentra™ significantly reduce high toxicity-related treatment costs and improve workflow efficiencies to handle more cases more effectively overtime.

    Introduction: The Economic Paradox in Oncology

    For years, the Additive Manufacturing (AM) industry has battled a persistent myth: that customization is an expensive luxury. While this may hold true in consumer markets, the opposite is often the case for healthcare. In oncology, where precision, consistency, and efficiency directly affect outcomes, a one-size-fits-all approach can quietly cost health systems billions.

    For radiation oncology, generic solutions frequently generate failure demand: the downstream clinical and operational burden of managing avoidable complications. The return on investment (ROI) of personalized 3D printing is therefore not limited to improved clinical accuracy; it represents a financial strategy that converts inefficiency, rework, and toxicity into measurable savings. By shifting from generic tools to patient-specific solutions, hospitals can improve patient experience, support clinicians, and strengthen their bottom line.

    The Hidden Cost of “One-Size-Fits-All”

    As an example, in Head and Neck Cancer (HNC) radiation therapy, the standard of care often involves rudimentary tools like cork and standard bite blocks. While inexpensive to purchase, these devices are costly in practice. Inconsistent immobilization and inadequate tissue displacement introduce variability between fractions, increasing unnecessary radiation exposure to healthy tissue.

    The downstream consequence is financial toxicity. Patients exposed to unintended radiation frequently develop severe oral mucositis (SOM)—a painful, debilitating complication that extends far beyond discomfort. Studies show that mucositis and pharyngitis in HNC and lung cancer patients are associated with approximately $17,000 in mean additional cost per patient, driven by unplanned hospitalizations, feeding tube placement, and intensive supportive care (Elting LS et al.).

    For patients, this means avoidable suffering during an already difficult journey. For clinicians and administrators, it translates into resource strain, unpredictable workflows, and escalating costs.

    The Value of Custom-built AM Solutions

    How can scalable customization reduce hospital costs? By preventing the complications that drive high-acuity spending.

    Kallisio’s Stentra™ platform illustrates how patient-specific AM solutions can be integrated seamlessly into real-world clinical workflows. Using a fast, standard intraoral optical scan, patient anatomy is captured with minimal burden on staff. Design is automated, manufacturing validated, and a customized device can be delivered in as little as 72 hours.

    Because each device is engineered to match a patient’s unique anatomy and treatment plan, Stentra consistently immobilizes and displaces tissue across therapy sessions. Published clinical data indicates this approach can reduce the incidence of severe oral mucositis by 77.6% (Journal of Oral and Maxillofacial Surgery). Preventing these severe cases helps hospitals avoid the cascading $17,000 per-patient cost associated with toxicity management—demonstrating that modest upfront investment yields substantial downstream savings.

    How can personalized devices improve operational velocity?  By saving an estimated 3–4 hours of expensive machine time per patient course.

    Linear accelerators (LINAC) are among the most capital-intensive assets in oncology. Every minute of delay or rework erodes throughput. Generic immobilization devices often require repeated setup adjustments and repositioning, introducing unpredictability into tightly scheduled treatment slots.

    Patient-specific solutions such as Stentra fit reliably and reproducibly, reducing setup time and variability. Data shows that 3–7 minutes per fraction are saved by using Stentra [Kallisio Value Analysis]. Over a standard 30-fraction course for head and neck cancer therapy, this accumulates to 3–4 hours of LINAC time saved per patient. The resulting efficiency releases capacity to treat more patients without adding shifts or staff.

    Similarly, poor fit with standard devices contributes to simulation re-scan rates approaching 10%. Custom 3D-printed solutions reduce this to <1% [Kallisio Value Analysis], minimizing delays, patient inconvenience, and unnecessary imaging costs.

    Conclusion: The Business Case for Personalized 3D Printing in Oncology is Clear

    1. Clinical Effectiveness

      • By improving patient compliance and treatment accuracy, hospitals mitigate the risk of expensive complications like mucositis.

    2. Operational Efficiency
      • Minimizing errors, interruptions, replanning and expensive treatment time allows centers to increase patient volume on existing infrastructure.

    As value-based care models continue to penalize complications and reward efficiency, AM technologies like Stentra demonstrate that personalized medicine is no longer a luxury, it is the most fiscally responsible path forward.

    Kallisio is a Gold Sponsor for Additive Manufacturing Strategies (AMS), a three-day industry event taking place February 24–26 in New York City. The conference brings together industry leaders, policymakers, and innovators from across the global AM ecosystem. Kallisio’s CEO Rajan Patel will also participate in a panel on “3D Printing for Oncology.” Registration is open via the AMS website.