When Lunn reached out recently about Depix’s latest project, ImageLab, he took the hyperbole even further.

“Last year we made designers cry. This year we will make them cry and fall out of their chair and shout with elation,” he wrote in an email.

If you’re the crying type, consider yourself warned. Here’s a hands-on look at Depix ImageLab and what it can do.

What is Depix ImageLab?

Depix ImageLab is a web-based tool for generating AI images. You’ve probably already used something similar. But ImageLab explicitly targets professionals, like product designers, looking for high quality renders.

“What we’re doing is making it very simple to generate really high quality images from anywhere,” Lunn told Engineering.com during a demo of ImageLab.

“Anywhere,” Lunn specified, could be “from a very early sketch, to a polished CAD image from any CAD manufacturer. It doesn’t matter what the CAD input is. It could be architecture, it could be products. It could be a circuit board. It could be a widget. Doesn’t matter. It just makes a great marketing photo.”

Anyone can sign up for an account at DepixImageLab.com to get 10 free credits. Each generated image costs 1 credit. You can also create five second videos for 9 credits apiece, though we didn’t test that feature.

When you sign up, this is what you’ll see:

The interface is self-explanatory. You start by importing an image (or picking a sample), and then you tell ImageLab what to do to it. If you don’t feel like typing a prompt, you can hit the glowing Voice button to dictate one instead. You can also attach other images to reference in your prompt, like an image of a texture you want to apply to the target image (we’ll show an example of that later).

Let’s see how it works.

Depix ImageLab in action

For our first foray with ImageLab, we picked one of the available sample images, a sketch of a car:

A designer might want to see what a real version of this car would look like—so we asked ImageLab to generate one.

Prompt: Turn this sketch into a photorealistic rendering on a desert highway.

After about 30 seconds of processing, Depix ImageLab generated this image:

Once you’ve generated an image, ImageLab provides a toolbar that lets you save it, crop it, compare it to the original image, select a specific region to manipulate in your next prompt, and apply other adjustments. There’s also an extensive gallery of presets that will apply automatic prompts to your image (more on that later).

Let’s try again. Here’s the image we just generated, with a new prompt being applied:

Prompt: Change the color of the car to matte black with red accents.

Prompt: Change the angle of the camera so that the car is viewed from the front, and change the car’s color to blue.

Prompt: Change the background to Times Square, and put a crowd of gawking admirers around the car. And change the color of the car to red.

Prompt: Put the camera inside the car so we can see the scene from the driver’s perspective.

ImageLab, like other generative AI tools, doesn’t always give you the result you’re expecting. In this instance, rather than giving us an inside view of the original car, ImageLab has instead put us behind the wheel of a second car.

Perhaps a better prompt would have made a difference. To see what might have been, we reverted to the previous image and used the “Enhance prompt” button to give us a helping hand.

Original prompt: Put the camera inside the car so we can see the scene from the driver’s perspective.

Enhanced prompt: First-person driver’s view through a car windshield during sunset, showing the steering wheel and dashboard lit by orange light.

The same problem persists, but at least there’s an orange glow.

We could keep trying increasingly specific prompts, as Lunn did throughout his demo when unexpected problems arose. It’s easy to get lost in a generative rabbit hole, playing with prompts until every pixel is perfect. So we’ll restrain ourselves to two final examples of ImageLab.

Let’s go back to the original sketch of the car and apply one of ImageLab’s built-in presets: Amalfi Coast.

Prompt (added automatically): Place this car on the Amalfi Coast in Italy with dramatic coastal cliffs, Mediterranean sea views, colorful Italian villages, and coastal road atmosphere while adhering to the shape and position of the car as close as possible while adhering to the original image

To demonstrate how ImageLab uses reference images, we uploaded this image:

Prompt: Apply the texture in the attached image to the car.

Is Depix ImageLab a “KeyShot killer”?

We’ve shown off several examples of ImageLab’s capabilities, and you can freely test them for yourself.

What’s obvious from even limited testing is that the AI tool varies in the quality of its output. It can quickly create impressive renders from hasty sketches, and it’s adept at adjusting certain elements of a picture—like colors—without affecting others. But it can also misinterpret prompts and create odd images that don’t hold up to scrutiny.

Like all AI tools, ImageLab can be good, bad and ugly. That’s AI for you; sorting the results is up to the user. But crucially, it’s incredibly fast and easy to use.

“We’re trying to be the thing for product designers and product marketers and CAD users… who really would love to be able to make a nice image, but just can’t invest the time in learning all the settings and tweaks you have to do for rendering,” Lunn said.

Lunn has plenty of experience with the settings and tweaks of traditional rendering, having previously founded a company called Bunkspeed which sold a rendering program called Hypershot. In 2010 Hypershot was taken over by Luxion and rebranded as KeyShot. It remains a popular rendering program to this day.

With AI-based rendering, is the writing on the wall for KeyShot and other traditional rendering programs?

We asked Thomas Teger, former VP of marketing for KeyShot and a former colleague of Lunn’s at Bunkspeed.

“While you can create truly amazingly realistic images with KeyShot, it requires a significant amount of skills and expertise with the tool itself, and even then it takes often a long time to get to the final image,” Teger said.

Uneven as it may be, generative AI requires no expertise and barely any time.

“ImageLab is a tool that lets you create images and movies the way you “think” about it, not based on what is available to your or how well you master various tool sets… I firmly believe that this will be the product that has the potential of becoming the “KeyShot Killer” as it redefines rendering as we all know it in a completely new way,” Teger said.

(Teger describes himself as an “independent advisor” to Depix. He’s not paid by the company, but he receives free credits “in return for cool content, testing, insight and promotion.”)

It should be noted that if there is writing on the wall, KeyShot has also read it. The company offers AI-based rendering features of its own through a product called KeyShot Studio AI.

What’s the cost of Depix ImageLab?

Creating AI images can be addictive. And with the ever-present element of randomness, it’s necessarily iterative. You’ll quickly burn through your 10 free credits.

When you do, you can top them up at $19.95 for a pack of 250. That works out to eight cents an image, though Lunn said Depix is still debating the right pricing model.

“This whole thing costs quite a bit to run, you know, plus all the labor and humans and all involved… you can get it as low as pennies per image, basically,” Lunn said. “It depends on the use case and the user.”

Regardless, Lunn has no doubt that ImageLab and its ilk represent a radical change. The AI tool can do near instantly what used to require highly valued software skills. At one point our conversation turned to Adobe Photoshop, the pre-eminent photo manipulation software. Lunn reflected that professional Photoshop users used to earn hundreds of dollars per hour for their professional services.

“But now,” Lunn said, “for 25 cents, you can make a perfect marketing image.”

The post Is this generative AI rendering tool a “KeyShot killer”? appeared first on Engineering.com.

This 25+ page white paper explores practical ways to simplify operations, accelerate throughput, and maintain quality in today’s most demanding production environments. It shows how modern execution strategies can close the gap between as-designed and as-built products while reducing costly delays.

Inside, you’ll learn about:

• Streamlining production flows and enforcing consistency across operations
• Advanced tracking, tracing, and defect management for full visibility
• Digital instructions and data capture that eliminate paper-based inefficiencies
• Scalable approaches for multi-plant execution and collaboration
• Integration across planning, quality, and execution for better outcomes

Your download is sponsored by Siemens.

The post Your Guide to Managing Complexity with Confidence appeared first on Engineering.com.

Leo AI, developer of a mechanical engineering AI assistant, has announced a new partnership with TraceParts, an online library of part catalogs and 3D models.

According to Leo AI, the goal of the partnership is to make it fast and easy for engineers to find the right parts.

“Instead of juggling filters,” the press release says, “engineers can simply type: ‘Find me a ball bearing with a 25 mm inner diameter, a lifespan of 10,000 cycles, and a speed of 8,000 RPM.’”

Leo AI’s so-called large mechanical model (LMM) will interpret the request, ask the user for clarification if necessary, and retrieve matching results from TraceParts. Here’s a quick video demo from Leo AI:

Last week I wrote about Leo AI completing a $5 million seed funding round. I gave an overview of the company’s AI platform based on an interview with co-founder and CEO Maor Farid, who told me that Leo could, among other capabilities, search through online catalogs to find parts for a given design (a spring for a suspension system, in his example).

“Leo looks for both online vendor catalogs and on your PLM, and it provides the springs that match the parameters that you specified or Leo calculated,” Farid said.

So what’s new here? I asked Farid to clarify how the TraceParts partnership will change Leo’s capabilities. This is his emailed response:

“Leo has always combined two things: doing the calculations to figure out exactly which part is needed, and then searching across PLM or vendor catalogs to find it.

What’s new with TraceParts is the scale and reliability it brings. Instead of searching across scattered catalogs, engineers now have direct access to one of the world’s largest, supplier-certified libraries – 112M+ parts and 2,100 catalogs – right inside Leo.

So the core capability isn’t new, but this partnership makes it far more powerful: faster searches, fewer mistakes, and much more confidence that you’ve got the right part. That’s why we’re so excited about it.”

Top Workplaces for Engineers 2026

In case you missed it in last week’s lengthy newsletter, Engineering.com is hosting its second annual Top Workplaces for Engineers program, and nominations are now open.

Here’s more from Engineering.com Editor-in-Chief Rachael Pasini:

“To be eligible, participating companies must employ at least 35 engineers or have an engineering workforce comprising 10% or more of their total workforce. The award is based on employee feedback captured by the confidential, research-backed Energage Workplace Survey. Participating companies will be evaluated against the industry’s most robust benchmarks based on more than 18 years of culture research.  

The award will honor companies that create exceptional workplace environments for engineering professionals across various industries, and we will publish the list of winners in the spring of 2026.

If you believe you work at a company that deserves such recognition and meets the criteria, nominate them at: engineering.com/topworkplaces. The nomination period runs through mid-January, but submit your nomination much sooner, before the busy end-of-year season kicks in with full force.”

3D Systems updates its software strategy

3D printing hardware and software developer 3D Systems has decided to double down on its proprietary polymer platform, 3D Sprint, and back off from its vendor agnostic software, Oqton Manufacturing Operating System and 3DXpert.

Why? Engineering.com senior editor Ian Wright weighs in with the unsurprising answer:

“If you’re going to talk about software these days, it seems to be a requirement that you use the phrase ‘artificial intelligence’ at least once… That certainly holds true for the latest announcement from 3D Systems, which led the press release that it would be selling off its Oqton Manufacturing Operating System (MOS) and 3DXpert business by emphasizing its decision to focus on its proprietary polymer software 3D Sprint as a response to ‘the transformative potential of artificial intelligence in additive manufacturing.’”

There’s more to the story than just AI, so check out Ian’s full article on Engineering.com: 3D Systems shifts its software strategy to focus on 3D Sprint.

One last link

For a long read on how engineering and design software could be rebuilt around AI, check out Patrick Hebron’s essay An All-Around Better Horse: AI and the Revolution in Design, Engineering, and Problem-Solving Methodology.

Got news, tips, comments, or complaints? Send them my way: malba@wtwhmedia.com.

The post Leo AI teams up with TraceParts for AI part search appeared first on Engineering.com.

If you’re going to talk about software these days, it seems to be a requirement that you use the phrase ‘artificial intelligence’ at least once. (Not so long ago, you could have said the same thing about ‘blockchain’. How’d that one turn out?)

That certainly holds true for the latest announcement from 3D Systems, which led the press release that it would be selling off its Oqton Manufacturing Operating System (MOS) and 3DXpert business by emphasizing its decision to focus on its proprietary polymer software 3D Sprint as a response to “the transformative potential of artificial intelligence in additive manufacturing.”

The more brand-agnostic software platforms are being sold to Hubb Global Holdings, a “strategic investment group” that doesn’t seem to appear on the web outside of this particular announcement, including on the LinkedIn pages of its apparent principals, Steve Lokam and Kalyan Yenneti.

“This strategic investment is aimed at significantly expanding [Hubb Global Holdings’] core capabilities, strengthening sales and service infrastructure, and driving broader industry adoption,” reads the press release. Given the evident lack of information about Hubb Global Holdings, those first two points seem inevitable. Whether broader industry adoption also obtains as a result remains to be seen, but I’m not optimistic.

So, whence the divestment?

It’s not the first move of this kind from 3D Systems this year, with the sale of Geomagic to Hexagon going through earlier this year. The expectation that this latest sale will close in Q4 of 2025 suggests that 3D Systems is pulling back from a strategy of pushing the broader adoption of 3D printing technology to focus on driving revenue and share price via a closed ecosystem of hardware, software, materials, and services.

That’s more or less the message from CEO, Jeffrey Graves: “Our Company is focused on enabling customers to fully leverage the advantages of additive manufacturing in their production environment,” he said in the press release. “We make this possible by providing fully integrated additive manufacturing solutions comprising 3D printing hardware, materials, software, and services to customers worldwide. We believe it’s critical to continue to invest in R&D to drive innovation in all elements of our solutions—focusing these investments where they can make the biggest impact for our customers and shareholders.”

Reading between the lines, it seems as though 3D Systems could be backing away from metal additive manufacturing (AM) entirely for the sake of concentrating on polymers. Whether from a technical or business standpoint, that’s not a bad bet: metal AM is significantly younger and less standardized than its polymer counterpart, and 3D Systems already has an enormous polymer materials portfolio, roughly three times the size of its metal offerings.

That leaves the fate of Oqton MOS and 3DXpert uncertain, especially in the hands of an unknown investment group whose principals have no evident experience with metal AM. It’s disheartening to see printer-agnostic software platforms fall by the wayside when the additive industry seemed to be coming around to the mindset that success can’t come from closed ecosystems and a rising tide lifts all boats.

The post 3D Systems shifts its software strategy to focus on 3D Sprint appeared first on Engineering.com.

In turn, these tools and the techniques needed to use them cost-effectively are spurring product developers to find and maximize real-world value and are working to incorporate them into their products and services.

The resulting gold being mined by these new tools and techniques is not by itself some long-sought breakthrough but an intermediate step in end-to-end product lifecycle management (PLM). And the counterpart is just as critical. Before gold is forged into something of value, it’s meticulously assayed to determine its purity.

The valuable insights from data gold mining are often suspected, or even known to exist, yet they remain buried. Managers are unwilling or unprepared to sift through this data and put it to use. Amid resource scarcities and time pressures, few organizations know how to do this effectively or affordably—the perennial headaches of data mining.

Mining for the data gold is often seen as too iffy to promise an ROI. And yet this gold is often related to the biggest challenges facing every product and system: obsolescence and failure, ever-evolving user wants and needs, marketplace disruptions, competition, sourcing, service, and rising costs.

The challenge in data gold mining is no longer just about getting data; it’s about the strategic application of new technologies to transform that data into actionable insights and value. The following areas are key to this challenge:

• Tightly integrating the most common forms of generative AI and agentic AI into PLM platforms
• Better techniques for generating AI queries
• Better analytical tools…descriptive, predictive, and prescriptive…to make sense of the returns from AI queries.

These tools and techniques seek to find gold in the form of hidden similarities among seemingly unconnected and unrelated phenomena. This includes feedback from customers who have little or nothing in common and are using dissimilar products and services.

Gold is also buried in the incongruities in sales orders and rejections, field service, warranty claims, manufacturing stoppages, and supply chain disruptions. Whether the data is structured or unstructured no longer matters. Ditto for whether the data is internal or external. For example, incorporating parts of the Industrial Internet of Things (IIoT) and their connected data-generating devices into the Large Language Models (LLMs) on which AI is trained.

What is also new is the size and depth of databases searched, as well as how these new tools and techniques overcome the disconnects that plague every organization’s data. These disconnects include bad and useless data formats; errors, ambiguities, and inaccuracies; data buried in departmental silos; legacy data with unsuspected value; and data that is misplaced or lost.

All this is aided by digital transformation in all its myriad forms. Digital transformation is increasingly vital to gold mining because it helps users gather terabytes of data into more complete, more accurate, more manageable, and more focused LLMs. Digital transformation can also help users pinpoint what is (still) needed for timely/effective decision making (e.g., data that did not get into a given LLM and should be for subsequent queries).

CIMdata itself is adapting by:

• Broadening its PLM focus to work with clients’ AI projects, creating an AI Practice with Diego Tamburini as Practice Director and Executive Consultant. He has held key positions at Microsoft, Autodesk, SDRC (now part of Siemens Digital Industries Software), and the Georgia Tech Manufacturing Research Center.   
• Expanding its work and capabilities in digital transformation that enables PLM—and is enabled by PLM in turn. The ultimate goal is to bring together engineering technology (ET), information technology (IT), and operations technology (OT) at the top of the enterprise.

Managers and staffers who receive these AI and analytics findings have a similarly daunting agenda.   They must learn how to discern and understand what the gold is telling them. And they must learn how to weed out what has already been simulated, designed, or engineered for production. And they must learn how to choose the most viable of these tools and techniques and how to manage them.

Effective data governance is crucial to gold mining. I strongly recommend a review of an AI Governance webinar written by Janie Gurley, CIMdata’s Data Governance Director, and me. Posted July 8, 2025, by Engineering.com, available at https://www.engineering.com/ai-governance-the-unavoidable-imperative-of-responsibility/

Further insight is available in my most recent Engineering.com posting, available at https://www.engineering.com/in-the-rush-to-digital-transformation-it-might-be-time-for-a-rethink/

Getting the gold into product development offers many potential benefits by uncovering:

• Causes of product and system failures
• Unexpected obsolescence in products and systems
• Users’ new wants and changing needs
• Early indications of marketplace disruptions
• Insights into competitors’ strategies
• Pending shortfalls in sourcing…and alternatives
• Likely cost increases and finding options
• Unrecognized service needs
• Better methods for production and operations…thanks to AI’s new ability to handle streaming data in LLMs.

Over time, diligent searchers may turn up dozens of connections and correspondences in these nine bullet points. Some will be simple, random coincidences. But many will turn into gold that reveals both opportunities and challenges.

Managing resistance

I urge readers to push the envelope, to find new approaches to everyday tasks, and try new things. While there is always pushback from staff and managers who are already overworked, without encouragement, change will never happen.

The usual response is, “Yes, you’re right, we’ll get to it later.” We caution that they are letting routine tasks get in the way of potential game changers. Ignoring these issues will not make them go away.Yes, obsolescence and failure, user wants and needs, marketplace disruptions, and all the rest (see above) will eventually surface, gain urgency, and become tasks that everyone must address.

Inevitably, this new gold will have to be dug out from data we assumed was useless and then painstakingly engineered into products, services, and forecasts. These values will mandate changes to the organization’s facilities, systems, processes, business partners, and suppliers. And they will have to be communicated to the sales force and distributors.

And there will be resistance to change and its huge costs, which are the underlying themes of this article. Those costs are part of addressing newly uncovered digital gold in a mad scramble amid fears that errors and oversights will place profitability, competitiveness, and job security at risk.

In summary, the quest for digital gold is our modern-day equivalent of the myth of Jason and the Argonauts and their search for the Golden Fleece. Like Jason, we must embark on a perilous journey and overcome countless challenges to seize a prize that promises untold value—transforming raw data into profitable products and services.

The post The data gold rush: how to uncover hidden value in your data appeared first on Engineering.com.

Add in the recent interest in the nuclear power industry (driven by the incredibly power-hungry data centers supporting artificial intelligence), and it should be no surprise that interest in AM applications for nuclear power appears to be growing at a commensurate rate. The latest example of this enthusiasm comes from Argonne National Laboratory, where researchers have been investigating the viability of using 3D printed components in the next generation of nuclear reactors.

More specifically, they’ve been using X-ray diffraction and electron microscopy to investigate steels made using laser powder bed fusion (L-PBF). In one case, they examined L-PBF 316H as a candidate for structural components in nuclear reactors and, in a separate study, looked at Alloy 709 (A709), which is specifically designed for reactor applications.

According to the researchers, both studies revealed important differences between the 3D printed steels and their conventionally forged counterparts, as well as how the former respond to heat treatments typically used for wrought materials.

“Our results will inform the development of tailored heat treatments for additively manufactured steels,” said Argonne materials scientist Srinivas Aditya Mantri, a co-author on both studies, in a press release. ​“They also provide foundational knowledge of printed steels that will help guide the design of next-generation nuclear reactor components.”

In comparing the microstructure of 3D printed 316H, Mantri and his colleagues used X-ray diffraction from Argonne’s Advanced Photon Source (APS) to reveal that the recovery and recrystallization of the steel were inhibited by nano oxides.

“Nano oxides act as a sort of barrier to the movement of dislocations and the growth of new grains, causing some dramatic differences between the response of L-PBF-printed and wrought steels to heat treatment,” said co-author and Argonne materials scientist Xuan Zhang, in the same release. ​“For example, the printed samples started to recrystallize at temperatures several hundred degrees higher than their wrought counterparts.”

The more recently developed alloy, A709, is designed for high-temperature environments such as those inside sodium fast reactors, a next-generation nuclear technology that operates at higher efficiencies compared to conventional designs. According to Argonne, this is the first experimental investigation of the material properties of additively manufactured A709.

They also studied the strengths of the heat-treated samples under tension. At both room temperature and 1022 F (550 C) – a temperature relevant to sodium fast reactor applications – the printed A709 displayed higher tensile strengths compared to the wrought A709. The Argonne researchers believe this was likely because the printed samples began with more dislocations, which also promoted the formation of more precipitates during heat treatment.

“Our research is providing practical recommendations for how to treat these alloys,” said Zhang, ​“but I believe our biggest contribution is a greater fundamental understanding of printed steels.”

The studies on 316H and A709 are published in the journals Materials & Design and Materials Science and Engineering, respectively.

The post Is 3D printed steel ready for next-generation nuclear reactors? appeared first on Engineering.com.

Additive manufacturing (AM) has advanced healthcare and the medical device industry is numerous ways, from orthotics and prosthetics to 3D printed scaffolds for organ and tissue transplants. Aside from the aerospace industry, you’d be hard-pressed to find a better use case for AM than in medical. The ability to 3D print highly customizes devices, often using exotic or difficult-to-machine materials, makes the technology a natural fit for an industry where every custom is truly unique.

One of the latest advancements in this vein comes from more than half a dozen researchers based in China, representing Chonqing Medical University, Dalian University of Technology, and Henan Polytechnic University, among others. Together, they’ve developed a 3D printable bio-active glass that can serve as a replacement material for bone tissue. According to the researchers, in their experimental trials with rabbits, the 3D printed glass sustained bone growth even better than a commercially available bone substitute.

The reasons for using glass in this particular application are that its crystalline structure exhibits similar material properties to bone and its primary ingredient, silica, can be liquified and printed. What sets this material apart from other 3D printed silicates is its biocompatibility and lower fusing temperature.

The researchers combined oppositely charged silica particles as well as calcium and phosphate ions – both known to induce bone cell formation – to form a printable, bio-active glass gel. After shaping the glass with a 3D printer, they hardened it into its final shape in a furnace at 1,300 F (700 C). They then tested the new bio-glass against a 3D printed plain silica glass gel and a commercially available dental bone substitute by repairing skull damage in living rabbits.

Although the commercial product grew bone faster, the bio-glass sustained growth longer; after 8 weeks, most bone cells present had grown on the bio-glass scaffold. The researchers say that this work demonstrates an easy, low-cost way to 3D print a bio-glass bone substitute, which could have wide-ranging applications across medicine and engineering.

While titanium is often the go-to additive material for bone implants due to its light weight and biocompatibility, its higher strength makes it more difficult to use as a direct substitute. As a result, biomedical engineers may need to use considerable computational resources to model and simulate adjustments in the replacement titanium part’s geometry, in order to avoid causing damage to the patient via post-surgical activity. If this alternative proves viable, it could make an enormous difference to the use of bone substitutes via additive manufacturing.

The research is published in the journal ACS Nano.

The post 3D printed bio-active glass shows promise as bone substitute appeared first on Engineering.com.

Let’s start with the news from Autodesk University, which took place last week in Nashville, TN. I wasn’t available to attend this year, but my colleague Jim Anderton, executive editor of Engineering.com, was live on scene. Ever loquacious, Jim’s already published five podcasts from the event. Have a listen:

• The future of artificial intelligence isn’t what you think: Autodesk senior director of AI research Tonya Custis on the real impact of AI.
• An AI first: building electrical layout: Augment co-founder Aaron Szymanski on using AI for this difficult engineering task.
• AI powered mass collaboration for engineering: Autodesk workshop XR senior director Nikolas Fonta on AI in AEC.
• One tool, to make everything? Autodesk Fusion Community Manager Jonathan Odom on a sea change in short run manufacturing.
• Drone helicopters with multiple applications: SwissDrones’ Pol Victor Gisquet on medium payload VTOL for commercial applications.

Notice any common themes? Keep reading for plenty more about AI.

Autodesk announces “neural CAD”

CAD-wise, the most intriguing news I saw from AU 2025 was Autodesk’s introduction of something called “neural CAD.”

Neural CAD is “a category of generative AI models trained to directly reason both about CAD objects and industrial and architectural systems,” according to an Autodesk blog post by Mike Haley, leader of the machine intelligence group at Autodesk Research.

For Fusion, that will take the form of “neural CAD for geometry,” which will allow users to generate BREP geometry from text prompts that can then be edited within Fusion. Here’s an image from Haley’s blog post (adjusted by me to make the components larger):

Text-to-CAD isn’t new among CAD developers, including Autodesk, who first showed off the tech in a research project called Bernini last year. I asked an Autodesk rep whether neural CAD for geometry is based on Bernini, and here’s the response I was given:

“Yes, the science, machine learning infrastructure, training data curation, and training of the models we announced at AU 2025 were directly influenced and in some cases based on the Project Bernini research we announced in May 2024. That experimental proof-of-concept was a critical step in Autodesk’s decade-long journey of machine learning research and development.”

What is new, however, is having a text-to-CAD system integrated directly in a major CAD platform, and it looks like Fusion will be the first to do it. But when? And how will users access it? Will it be built into Fusion or accessed through a paid add-on? Here’s the same Autodesk rep:

“We expect the generative AI technology that Autodesk announced this week as neural CAD to be available soon. We have not yet announced the timeframe or shared how the functionality that this technology offers will be shared in our commercial offerings.”

Last AU, Autodesk talked a lot about Bernini, even though it was never presented as anything more than a proof-of-concept. A cynic might say Autodesk was eager to convey their commitment to a hyped new technology, regardless of whether that tech was ready for end users. Could neural CAD simply be the next act in that play?

Only one way to find out. In the AI arms race, I’ve seen developers announce features that have yet to materialize after years, so I’ve adopted the stance of I’ll-believe-it-when-I-see-it. The technology of text-to-CAD is clearly feasible (Bernini proved that, as have many others), but whether it’s useful in an engineering workflow has yet to be seen.

Hestus updates Fusion Sketch Helper

In more Autodesk-related AI news, developer Hestus has significantly updated its Sketch Helper Fusion add-on. Sketch Helper is an AI tool that automatically offers and applies constraints while Fusion users create sketches.

In the previous version of Sketch Helper, Fusion users would choose from constraint suggestions in a window to the right of the Fusion workspace. That window is still there, but now Sketch Helper automatically shows a preview of what it thinks is the most likely set of constraints directly on top of the sketch. Hitting Enter will apply the constraints immediately. Users can also use the arrow keys to switch between previews of alternate suggestions.

“Now we are entering the territory of coding copilots,” Sohrab Haghighat, CEO of Hestus, told me during a demo of the new update. “Not only [do] I have recommendations, I actually know what the best recommendation is.”

Haghighat says that on average, Sketch Helper can make Fusion designers two-and-a-half times faster. That number is based on Hestus’ internal testing with experienced Fusion users.

“I think Sketch Helper is just scratching the surface,” Haghighat said, describing what he sees as the enormous potential of AI in design and manufacturing.

“More than 50% of [the] time engineers spend on designing parts goes into redesigning the same part,” he said. “There are about 8 million engineers and designers working on designing these products… and if they spend 50% of their time or more redoing what they have already done, that boils down to more than $400 billion in productivity loss on a yearly basis. To me, that is the promise of Hestus.”

Leo AI gets more funding for its mechanical engineering copilot

Leo AI announced earlier this month that it has completed a $5 million dollar seed funding round, bringing its total funds up to $9.7 million.

If the name sounds familiar, perhaps you’re one of the 20,000 engineers (according to the announcement) that uses the “world’s first domain-specific AI for mechanical engineers.” Or perhaps you, like me, keep seeing Leo mentioned in the wilds of online CAD discourse—such as the email I got from a reader not long ago who wanted me to know that “Leo AI is building a very useful engineering copilot.”

I decided it was high time to speak with Maor Farid, the co-founder and CEO of Leo AI, to learn more about the company.

“Think about [Leo] as the first AI made by/for engineers,” Farid told me. “It’s the first AI to understand CAD, not only text. It’s not a large language model. It’s the large mechanical model.”

Leo integrates into a company’s PLM system and is trained on engineering books, articles, standards, and an organization’s own design data. Farid compares it to “ChatGPT that understands CAD” and claims it achieves a 96% accuracy rate for engineering questions.

Farid showed me a demo of Leo. He pulled up a partial CAD model for a suspension system and asked Leo to help him determine the spring constant, length and other parameters that would satisfy his design requirements.

“It provides you with the calculation,” Farid said. “Every calculation or piece of information is accompanied by a source, and when you click on the source, Leo opens up the right page in the right source. And again, it can come from your guidelines or external ones. You don’t need to write Excel spreadsheets anymore.”

Once you know what you need, Leo can go even further and find the part for you. “Leo looks for both online vendor catalogs and on your PLM, and it provides the springs that match the parameters that you specified or Leo calculated,” Farid said.

I didn’t record Farid’s demo, but this quick video from Leo AI shows the tool’s ability to answer engineering questions and reference technical sources:

Farid estimates that Leo can save mechanical engineers around five hours per week. Regardless of the time savings, at least one user has found it helpful. When I asked the reader who emailed me in praise of Leo (who asked not to be named) to elaborate on his use of the tool, this is what he wrote back:

“So far, I’ve used Leo for structural and material analysis. Leo has been spot-on with static stability assessments and even showed me a more in-depth way to estimate material strengths. The fact that it walks you through the steps to solve a problem as opposed to just spitting out an answer is a huge plus. As I said to Maor in an email: I would have picked up Statics far quicker had I access to Leo while I was in school.”

I’d love to hear more opinions from Leo users, so if you are one, share your thoughts with me at malba@wtwhmedia.com. Farid hinted that lots more is coming to Leo, so stay tuned for further coverage.

Turning up the heat on the simulation revolution

Malcolm Panthaki has been trying to spread simulation for decades. The co-founder of Revolution in Simulation, a non-profit coalition of simulation experts, is convinced that the technology is being held back by its complexity, perhaps even hoarded by its most experienced practitioners.

Revolution in Simulation’s driving goal is the democratization of simulation. Panthaki wants the powerful technology to be more widely accessible, and for the experts to help make it happen rather than stand in the way.

“[T]hat power needs to be put safely in the hands of others, and the simulation experts need to put in the time and the effort to do that packaging and to figure out the best ways to do it,” Panthaki told me in a recent interview.

He admits that democratization hasn’t made much headway in his years of proselytizing. The good news is that something may finally be catalyzing a change. Guess what.

“AI seems to have changed it, because it’s just simply everywhere, and it’s incredibly sexy, and nobody’s taking no for an answer,” Panthaki said.

For more about how the simulation revolution is heating up, read my Q&A with Panthaki on Engineering.com: How AI is supercharging simulation democratization.

Top Workplaces for Engineers 2026

Engineering.com is hosting its second annual Top Workplaces for Engineers program, and nominations are now open. Here’s more from Engineering.com Editor-in-Chief Rachael Pasini:

“To be eligible, participating companies must employ at least 35 engineers or have an engineering workforce comprising 10% or more of their total workforce. The award is based on employee feedback captured by the confidential, research-backed Energage Workplace Survey. Participating companies will be evaluated against the industry’s most robust benchmarks based on more than 18 years of culture research.  

The award will honor companies that create exceptional workplace environments for engineering professionals across various industries, and we will publish the list of winners in the spring of 2026.

If you believe you work at a company that deserves such recognition and meets the criteria, nominate them at: engineering.com/topworkplaces. The nomination period runs through mid-January, but submit your nomination much sooner, before the busy end-of-year season kicks in with full force.”

Quick hits

• Altair has announced the winners of its 2025 Altair Enlighten Award honoring sustainability and lightweighting advancements. The winners across seven categories include Vortex CAE, CompositeEdge GmbH and ATA Mute B.V., Syensqo and Geely, Marelli, NIO and AkzoNobel, and Lucid Motors in two categories. Check out the details here.
• CAM kernel developer ModuleWorks announced that it’s developed GPU-accelerated simulation, and the technology will debut in the next releases of both Cimatron and Mastercam. ModuleWorks says GPU simulation can provide up to ten times the performance of CPU-based approaches, adding that it will particularly benefit large and complex 3-axis and 5-axis parts.
• Siemens has added AI-powered Lifecycle Assessment (LCA) capabilities to its PLM software Teamcenter. Teamcenter Sustainability Lifecycle Assessment will “empower engineering and manufacturing teams to assess a product’s environmental compliance, supply chain risk and cost early and throughout the product’s lifecycle,” according to Siemens. The tool was developed with product lifecycle intelligence developer Makersite in a collaboration first announced in July.

One last link

For your daily dose of Dassault drama, read Ralph Grabowski’s The Backlash to Dassault Systèmes New Fees on Solidworks Third-party Developers.

Got news, tips, comments, or complaints? Send them my way: malba@wtwhmedia.com.

The post Is Autodesk’s neural CAD worth getting excited about? appeared first on Engineering.com.

So, it should come as no surprise that the newest generation of machines from Virginia-based MELD Manufacturing are aimed squarely at aerospace applications, with modifications to give them the capability to deposit titanium and aluminum alloys using MELD’s proprietary metal AM process.

The company’s unique solid-state process eschews the melting seen in powder bed fusion (PBF) and directed energy deposition (DED), and instead uses a combination of friction and pressure to deform metal feedstocks plastically. As a result, MELD machines can operate in open atmospheres and, according to the company, the parts they print are fully dense.

While MELD has previously claimed that its process is compatible with aluminum and titanium (as well as magnesium, copper, nickel, and steel), the new DragonForge series is being positioned as being even more capable than previous generations, printing AA7075 (for example) without graphite or other lubricants.

“This equipment is truly next generation capability for rapid 3D printing of large aerospace parts,” said MELD Manufacturing CEO Nanci Hardwick in a press release.

According to the company, the machines’ new hardware configuration is paired with software features intended to enhance machine autonomy and the user interface, as well as new simulation and monitoring tools. MELD is also highlighting the addition of a digital twin to help plan and implement a printing strategy before actually starting production.

From a business perspective, the strategy here seems to be positioning the DragonForge series as enabling the on-demand production of “printed forgings” to support the repair or replacement of legacy components, as well as new parts, without the need to rely on the (currently anemic) domestic forging industry.

Between the push to reshore industrial production for the sake of national defense and the difficulties of rebuilding an industry that has largely relocated to India and China, this strategy seems like a safe bet indeed.

The post MELD Manufacturing targets aerospace with new line of machines appeared first on Engineering.com.

Manufacturing engineers call it the “death zone”. Some call it the “scaling conundrum”. Once a design is finalized, and a prototype made, it’s frequently expensive and difficult to create the pilot runs and small volume production that’s essential to test market and validate a new product. It’s far too expensive to tool up for mass production of a part or device based on only a single prototype, but how can an innovator bridge that gap?

At AU 2025 in Nashville, Tennessee,  Autodesk Fusion Community Manager Jonathan Odom demonstrated a short run manufacturing platform that allows innovators to program and control multiple production technologies, from multi-axis machine tools to 3D printers, and significantly, to share designs with contract manufacturers that offer both manufacturing capacity and useful design expertise of their own.  Is this the universal tool for making anything and everything? Odom explains all to Jim Anderton. 

For the audio only version:

***

Catch up on the latest engineering innovations with more Industry Insights & Trends videos and podcasts.

The post One tool, to make everything? appeared first on Engineering.com.