1. Points
2. Lines, arcs, edges, also known as wireframes
3. Faceted surfaces
4. Smooth surfaces
5. Solids

All 3D methods have advantages over 2D drawing, chiefly:

No confusion

A part can be difficult to picture from its orthographic views. When this happens, the drafter in a 2D world is called upon to make an isometric view. Since this takes special skills, special tools and extra time, isometric views are rarely created, and their presence on a drawing is more the exception than the rule.

However, with the advent of 3D, surface and solid modeling in particular, it is common to show 3D objects with hidden lines removed and even shaded. You can spin a part around and look at it from any angle. You can zoom in and out. With surface and solid models, hidden lines are removed automatically (or shown dashed) and in real time. All this allows easy visual understanding of the shape of the part.  There is no confusion.

Accuracy

3D models, as opposed to 2D drawings, are accurate by nature. 2D drawings rely on the creation of orthographic views. Each view is measured and created individually and independently of the other. Every line is made manually. At any point, you can make a mistake. Suppose, for example, a detail is moved in one view, but the other views in which the detail appears are unchanged. Or the detail is omitted in one or more views. Each part and all its details have to be considered in all the views and drawn accordingly. If it is hidden in a view, it should be shown with dashed lines. Each time it is drawn, it is a mental exercise and a source of potential error. An accurate and complete set of views is a tribute to a skilled and practiced drafter.

By contrast, a model made in 3D is made only once and then viewed from different sides. You don’t create the view, the software does. It’s like a physical object with a camera taking photos from different sides. Views are still often required on 2D drawings, but with a 3D model, view creation is done practically at the push of a button. And the isometric view, once a laborious exercise, is now a breeze.

Faster to the finish (almost always)

Creating a 2D drawing of a part may at first appear to be more expedient than making a 3D model. But it is a trap. Don’t fall for it. From the start, 2D requires extra time to do mental gymnastics to visualize the part, time to correct mistakes and time to make each view manually. After all this, a stack of drawings is dumped on tables in machine shops or construction trailers. Then the mental gymnastics are done in reverse — with the 2D drawings turning mentally into 3D, often after much head-scratching, time and too often mistakenly and inaccurately. But now, the mistakes are made in metal or concrete.

A 3D model, on the other hand, is a cinch to visualize. A machinist can spin it around and view it from all angles, as they would with the physical part in hand. The 3D model has the added benefit of being imported more naturally to CAM.

The puzzle that never had to be

With only paper to draw on, designers, engineers and architects became proficient in converting objects into their orthographic views — and machinists and builders became experts in unraveling the blueprint into mental images of the part or building. It worked well — except when it didn’t.

The post What are the advantages of 3D CAD? appeared first on Engineering.com.

In this article, we continue the discussion of AI in design and manufacturing software with Stephen Hooper, VP of software development for Autodesk’s Design and Manufacturing division. Part two can be found here.

Engineering.com: I’ve anticipated levels of automation for various design software AIs — like SAE levels that classify autonomous-driving capabilities and Level 5 indicating full autonomy. Design software with such capability would get a prompt such as, ‘Hey AI, design a car’ and would design and builds a car. Level 0 is where we’re now. We design and building everything. The geometry is a little smart but mostly dumb. In between, levels abound. At the first level might be what Mike Haley of Autodesk talked about — a natural language UI. That might be the low hanging fruit. That would eliminate the dependance on traditional icon-based menu-based system.

Hooper: Some vendors say, and some startups have tried this. You’ll see a lot of these new startups where this text-based input leads to maybe a skateboard. It’s a little naive to believe that we could do much more than that for a couple of reasons. Let’s use 2D graphics as an example. Let’s suppose I write a prompt that creates an image of a dimly lit nighttime street scene in San Francisco. It’s a back street with neon lights, and there’s a car parked on the curb on the sidewalk. AI: create that image for me. It will accurately create that image for you. The trouble the large language model can get the same prompt three times and yield three different results. With a specific idea in mind, you’re going to have to start to expand the prompt. You’re going to have to say “I want a green neon sign, and I want the green neon sign to say Al’s Bar and I want the Al’s Bar to be six feet off the floor on the right-hand side of the image. And the car should be a Chevy pickup truck. And make it red. The problem is that for a precise output, the prompt will be so big and take so long to define that one may as well create the image manually. This is true with parametrics, too. If I say I draw me a flat plate that is 200 by 400 mil and it has six equally spaced holes in the middle of a flat plate and those equally spaced holes are going to be drilled with six mil diameter all the way through. It’s almost faster for me to draw a rectangle, put the holes in and dimension it. I think a pure text-based product that delivers a whole product definition is highly unlikely. I expect we will move towards what we would call a multimodal prompt by which one may provide an equation for the performance characteristics of the product. An engineer might provide some hand sketches, a little bit of a text description, and a spreadsheet that includes some of the standard parts to be used. I would call that a prompt package that is multimodal. You’d give back to an AI that’s able to accept multimodal input. From that it would derive a range of options with which to interact, edit and refine procedurally to get to the target output. There might be some things one can have produced purely from a prompt — for an M5 screw with a pitch of 1.5, for example. But to get to a product definition, it’s going to be much harder.

Engineering.com: There may be certain things that I’m used to doing, certain shapes that I’m used to using, or certain components. What if the AI could anticipate them? Say I’m a bike designer and in the habit of using round tubes. Could AI sense from the line I am drawing that it will be tube and start drawing a tube? Can it use the shapes I am familiar with? That’s what I’d call Design Assist rather than fully automatic design.

Hooper: I think at the moment people’s mental model of this is that it’s static and asynchronous. I think for it to be truly useful; it will be interactive and synchronous. With bicycle example, one may draw a layout sketch, and it comes back with 16 options. One could say, “I like that option.” It’s not actually right now so I’m going to tweak it a little bit and then it’s going to come back and say “Okay, based on how you’ve tweaked it, I’m going to optimize it so one can make it with carbon fiber in a mold.

Engineering.com: That’s been my frustration with what’s has been provided so far. We’re engineers and one gave us generative design. Generative design is going to start from scratch and give us, excuse the term, garbage geometry. An experienced bike designer would want to start with tubular construction. A structural engineer may want to start modeling with I-beams. Not globs. We’re not going to use that.

Hooper: There’ll be some elements that are deterministic and other elements that can be created. The cross sections for steel structure are going to be 100% deterministic. It could be 50 by 50 by 2.5 box section or an action or a W-150 I-beam. Those will be deterministic. Then, again, we’ll have that multimodal input. one might say to the system, here are the different types of steel members that I want to use. Then one might give it a rough line sketch to say I want a structure that is three meters high in this kind of format. It will take the sketch and the list of standard content that one want to use and produce the structure for you.

Engineering.com: That is what I would call Design Assist. It’s going to use shapes and parts I’m comfortable with what I’ve already found to be optimum or standard and start using those things. If I’m making a wall, I don’t want to have to draw the two by fours. If I’m creating a commercial building I don’t want to draw the I-beams. I don’t want to use blobs. Let me use round tubes. AI can help me figure out where the connections between the round tubes should be. What is the optimum configuration of the round tubes for maximum strength and minimal weight?

By the way, no one has taken me up on my bike challenge, designing a bike frame that is better than the standard diamond shape made with tubes. Excuse my impatience, Stephen. I know one guys are trying hard. You’re putting a lot of stuff into the CAD software. This is me saying after one part of the house is redesigned, It looks great but what about the rest of it? Why can’t we do this? Honestly, I love that Autodesk isn’t making me annotated drawings. That’s great.

Hooper: point on levels. I would suggest levels that come after that. The level that comes after that would be multidisciplinary. Now, you’re looking at a 3D model or someone using Cadence is looking at a printed circuit board. There are different AIs and different domain disciplines. An AI that can get into a multidisciplinary model would be ideal. Beyond that, into systems architecture. Now I can generatively produce a systems architecture for a product. Then I’m not going to need to do a detailed design. I’m going to look at the interaction. I’m going to have some black box for the software — some black box for the transmission, the suspension, another black box for the electronics. We can build the systems architecture generatively and then at the next level from systems architecture, then being able to generatively produce the actual details in each of the disciplines. Then I think we’ll get to a generative AI design platform.

Engineering.com: Okay, but don’t give me blobs.

Hooper: I agree — no generative design. Only in the sense of historical generative design, a generative AI platform for design.

Engineering.com: That annotation item and the CNC AI mentioned earlier sound excellent.

Hooper: At Level 1, we have a design check and at level two we eliminate the non-value-added tasks.

Engineering.com: To remove what we don’t want to deal with — because engineers hate to annotate.

Hooper: Level three is the design assist; level four is multidisciplinary; level 5 is systems level and architecture; level 6 is the complete product definition.

Engineering.com: I’ll be taking a stab at establishing those levels. I’ll share them with you. We’ve been hearing companies say they’ve got AI and I think how much? A standard with levels would let everyone see if they are at level one or two.

Hooper: We’re also being secretive, because there may be things that they may be things we’re working on that we don’t want to talk about.

Engineering.com: I thought so but one have told me about Fusion 360 having automatic annotation. Is that public information?

Hooper: The annotations in Fusion will be live in the product soon. That’s public, but there may be other things that we’re working on with Mike Haley that are secret.

The post Part 3: How will Autodesk use AI for product design? appeared first on Engineering.com.

We’ve all been swept up in the AI wave. We have tried all manner of large language models (LLMs, including the media favorite, ChatGPT) and found them all lacking in one way or another. One —Claude — claimed to be good at math but wasn’t. Most seem to be good at answering college-level English literature questions, but incapable of solving freshman physics problems. We have used them for help when researching articles (especially Perplexity). When we gave ChatGPT several engineering-type questions, it always had an answer — but not always correct ones.

Perhaps we should not expect LLMs to be know-it-alls. How could they be? They’re trained on data troves of information scraped from all who would supply it as well as publicly available information such as Wikipedia. While the collective size of the data at their disposal is staggering, the quality, accuracy and depth of their answers is all too often lacking. Plus, the data that may be most valuable to engineers — their own — is off limits to LLMs.

Any organization of a good size and long history will have a tremendous amount of information. Therin is a valuable history with documents, drawings, models, revisions, PLM databases … all those which compromise tribal knowledge.

If this was a class on the failures of LLMs for engineers and the need for AIs to train on their own data, Jeff Platon, head of marketing at Accuris, would be in the front row and emphatically raising his hand.

“We totally got this,” he would say.

Accuris has a product called Goldfire, a semantic search application made to search an organization’s data.

Semantic search is the type of search that can sense context and the user’s meaning, as opposed to keyword search, which looks for exact word matches. By way of example, semantic search will find a different “best football player” in the U.S. than the UK, recognizing the correct sport in each location.

In the U.S., who bigger or with more history than the Navy? Who more to benefit from sematic search through vast repositories of data? Who is less likely to put their data on the cloud for training of civilian LLMs for their training? Instead, the military carefully guards its information from prying eyes, keeping it encrypted on secure servers.

Platon opens with a slide of an aircraft carrier. He comes from a Navy family. With the single biggest weapon system on display, a marvel of technology and an exemplar of operations in the most devastating environment possible — war — he has the attention.

Mining a company’s own data is what Goldfire is all about. Accuris does this by combing through all of it, indexing, linking … in short organizing and hyperlinking in a way to make information findable and usable.

The aircraft carrier was not just for show. The U.S. Navy is an Accuris customer and uses Accuris to help make design decisions.

NASA is another customer. Platon tells a story of helping NASA find information to avoid astronauts returning to Earth, splashing down in the ocean, but their capsule unable to right itself leading to a potentially “very bad outcome.” NASA scoured their information troves for information on inflatable bags used to successfully right reentry capsules for the Apollo mission. They found nothing. They called retired engineers to see if they had kept any information. The search went on for a year until finally they called in Accuris.

“Within 20 minutes of implementing the AI infrastructure, they found 249 documents that solved their problem, say Platon. They were able to fix the engineering problem. NASA saved $2 to $3M dollars and one to two more years of sorting through the data and organizing it.”

“We helped NASA find a solution to astronaut recovery,” says Platon.

Accuris works with text documents and databases. It aims for a wholistic approach to all of an organization’s needs. By analyzing everything, it’s able to connect data islands of CAD and PLM to PLM, ERP, SQL databases as well as manufacturing and upstream operations such as procurement systems. Along with other solutions such as automated BOM reporting from a database of 1.2B parts and Supply Chain Intelligence, the ESDU knowledge base containing engineering design data, Engineering Workbench, an AI platform for standards, codes, and regulations management.

Putting it all together, Accuris should theoretically enable an organization to answer operational questions that span departments and disciplines. For example, if a part fails on a production line, how long it would take for the supply chain to replace it?

This is vital for China Plus One strategies, says Platon.

Accuris’ AI technology is already available for this purpose and in use by 900,000 design engineers, including many large companies, branches of the armed forces and defense contractors.

It may be Accuris’ worst kept secret.

“We don’t make a big deal of it,” says Platon modestly.

Lest we think Accuris is only military, Platon steers us towards green hydrogen, an energy source of such potential that people have stated that it will save the Earth, a technology in which Accuris is engaging.

Accuris may be better known as IHS (its previous owner) and as a publisher of millions of standards of which ANSI, SAE and AS are the best known. The company also has access to seven million technical articles and books and 108 million patents and patent applications.

That, all by itself, would be something for a neural network to feast on.

The post Forget ChatGPT, the answer is in your own data appeared first on Engineering.com.

The first CAD program most designers of the previous generation ever saw was AutoCAD. After its initial release in 1982 by a small band of developers from Marin County (North of San Francisco) created AutoCAD and swept up the CAD market. In 20 years, AutoCAD was established as the lingua franca of design.

The first release of AutoCAD was strictly 2D, but Autodesk added the third dimension in 1985 with AutoCAD 2.1. They must have thought they needed to since they were following on the heels of the 3D MCAD juggernaut, PTC. However, most designers were content to stay in their 2D comfort zone and continued to use CAD as just an electronic version of their drafting tables.

With millions of 2D CAD users able to run on the poor man’s computer (the IBM PC, instead of workstations on minicomputers required by MCAD programs), CAD entered the mainstream. The drafting table was relegated to museum-level obscurity. It was not until SolidWorks 95, 13 years after AutoCAD came onto the scene, that 3D CAD would enter the mainstream.

The first picture show

The first commercially available CAD system was the Picture System by Evans & Sutherland.

Let’s back up. In 1963, Ivan Sutherland of Evans & Sutherland had already created Sketchpad for his PhD dissertation at MIT. The Computer History Museum recognizes Sketchpad as the first interactive 3D CAD program. However, Sketchpad and the workstation it ran on (MIT’s experimental TX-2) would stay in the lab and not enter the market.

After Sutherland graduated from MIT, he teamed up with David Evans to form Evans & Sutherland. It is unclear how much Sketchpad code ended up in Evans & Sutherland’s CAD programs, the first of which was the LDS-1. Their first success came in 1968 with the Picture System for “interactive, dynamic 3-D line drawings” (pictured above).

MIT kept developing Sketchpad after Sutherland left the Lincoln Laboratory building, piling on the 3D features. In an old black and white newsreel, we see the bespectacled Timothy Johnson in a suit and tie earnestly conducting what may have been the first-ever CAD demo. He is seated at what appears to be a nuclear plant control room or perhaps battleship command. Such was the state-of-the-art in computers was at the time, with their dials and switches for inputs, an improvement over punch cards. Sketchpad added a most innovative light pen that seems to react to the screen, a glorified oscilloscope, more or less, with a 7-in. square screen on which lines flicker.

However, the lines were not ordinary lines, such as lines drawn on paper, and Sketchpad no ordinary sketchpad.

Johnson shows a simple house, like one a child would draw. He shows a front, top, and side view, like a drafter would draw. Then the magic happens. He makes the apex of a roof in one view, which immediately shows up in the other views. What may have looked like four independent drawings were not that at all but views of one 3D model. This was the first automatic view creation using a single source of truth, the 3D model.

As with every technological breakthrough, we must consider the impact on humanity. Wouldn’t automatic view creation put those paid to make them out of work?

They briefly pause to consider — and quickly move on. Technology then, as now, is relentless. Automatically and painlessly extracting orthographic views should have been seized upon as Sketchpad’s main selling point — had MIT been selling Sketchpad. The reporter could have filed his story right there had he realized the enormity of what he had witnessed. But we have to wait for it. Johnson has more to tell.

To prove it was truly 3D, not smoke and mirrors, the house, which has F, T, and S on its faces for front, top, and side views, is rotated about the vertical axis. Sure enough, the S is backward when the house is rotated 180°. There is no smoke and mirrors here. It is, however, a setup for what we are about to see next. 

Lawrence Roberts of MIT shows us a wireframe of a simple part. “What if you were to stretch ‘fabric’ across the wires?” The part is rotated about the vertical axis, and the fabric hides the edges or parts of edges that would not be visible the perspective of the user.

The algorithm is taxing for the TX-2. The computer is overcome. It sputters and halts during the rotation. Remember, this is in the 1960s.

“The computer is doing a lot of calculations,” said Roberts.

Perhaps put off by halting display, the reporter does not seem as impressed as he should have been. The ability to remove lines on the screen that would be hidden from view in real life is nothing short of a technological marvel, if not a miracle, 50 years ahead of its time. Anyone seeing this who previously had to integrate a jumble of solid and dashed lines, which 2D orthographic views often are, into a 3D mental image and then into a hidden-line-removed isometric view, a Rubik’s Cube-level exercise, would have been floored.

The reporter asks how big the “paper” behind the “window” is. To his utter astonishment, it is 2 miles wide.

Now we have a story.

The post What came first, 2D or 3D CAD? appeared first on Engineering.com.

There are at least six ways that 3D objects can be modeled with computer software. What makes one of them, surface modeling, so special?

1. Points
2. Lines, arcs, edges, also known as wireframes
3. Faceted surfaces
4. Smooth surfaces
5. Photogrammetry
6. Solids

Here, we focus on surface modeling. First, let’s look at all the advantages surfaces have over other methods.

1. Smooth shapes

When a smooth, organic or curvy shape is to be modeled, a surface modeler comes in quite handy. This makes surface modeling software the preferred tool for industrial design, automotive designers and the streamlined shapes that the aerospace industry demands.

But beware: not all surface modelers deal in smooth shapes. Some surface models are faceted. Just like the facets on a diamond, the shape is composed of a collection of flat faces, either triangular or quadrilateral. For example, there is the TIN (triangular irregular network) that civil engineers use to model terrain by representing it as a mesh. That’s fine for projects measured on a big scale, such as miles. Brilliant is the program that is able to interpret the billions of point from a LiDAR scan, like the stars in the sky, as a patchwork quilt that drapes the Earth, but…

It’s not smooth.

Nor are the 3D prints made using STL files, also faceted models. That may be fine for a 3D print if the facets are small enough, but often, 3D-printed parts require a secondary operation to smooth out the edges a faceted modeler produces.

But to make surfaces that are truly curvaceous and smooth, with one blending into another in a way that is most pleasing to the eye, that requires higher-order mathematics. When an automotive body panel has to blend into another so smoothly that not even a reflection will reveal the blend, that is when we have found the Holy Grail: Class A surfaces (more on that later).

2. Easy to change shape

Most surface modelers used to make smooth shapes rely on polynomial curves called B-splines. Surfaces can be adjusted by manipulating the splines or the surface itself. For example, if a surface patch does not produce a pleasing blend from one area to another, such as on an automotive door panel, a user can adjust the control points of the splines until satisfied.

3. Precision

A surface can be blended and shaped with mathematical precision instead of relying on hit-or-miss visual techniques, such as adding or subtracting solid shapes. There’s also no need to make ultimate smoothness the task of manufacturing, as was done with lofting, when section cuts were drawn at regular intervals along the length of a ship and it was up to the boatmaker to make sure the final shape was smooth. Or the machinist’s file to smooth over the facets made from a STL file.

4. Visualization

The stunning images of automobiles we see on screens or pages are often not made with a camera but with a surface modeler or a hybrid surface-solid modeler. Surface modeling produces Class-A surfaces, the smoothest of all surfaces, which, after they have been textured and rendered, look so realistic that they are difficult to distinguish from a photograph.

Here are the disadvantages of surface modeling.

1. Not the easiest to use

Surface modeling takes more time and effort to master. Advanced surface modeling is no child’s play. Proficiency with surface modeling usually requires training and, almost always, considerable experience.

2. No mass

Surface models are only skin-deep. Only the surface with zero thickness is defined. A surface model can completely and precisely enclose a space (be watertight) and yet not know the volume of the space, like a stomach that doesn’t know it’s full. This limitation is overcome with a hybrid modeler, which combines a surface and solid modeler.

3. No interference

If a surface intersects with another, it can be used like knife to cut it. However, if a surface is completely enclosed by another, it can be of no use. For example, if you would like to make a hollow sphere by subtracting a small sphere from a larger one, you can’t do it with surfaces. The smaller sphere will exist in a vacuum, not in contact with the bigger sphere. However, spheres made with a solid modeler will have their interiors completely defined, as if they were full of matter. You can subtract the smaller solid sphere from the bigger solid sphere and get a hollow sphere (technically still a solid).

Solid models, with their precise internal definition, are therefore better suited to prevent “clashes,” what unintended intersections are called when one pipe “interferes” with another pipe or with the structure of the building, for example. Commercial building and process plant designers find such “clash detection” invaluable in preventing rework and downtime on the site.

The post Why model with surfaces? appeared first on Engineering.com.

Don’t it always seem to go
That you don’t know what you’ve got ‘til it’s gone?
 — Joni Mitchell, Big Yellow Taxi, 1970

There once was a Golden Age for value-added resellers (VARs). In the business of distributing much-sought-after CAD software, all they had to do was wait for it to be sold nearby and they would get their cut. Then, every year afterward, they would collect maintenance fees. With CAD software being hard to learn, they could also cash in on selling training.

Oh, those were the days.

These days, most designers and engineers have CAD. If they need more licenses, they can bypass the VAR and get those licenses directly from the vendor. If they don’t know how to use a feature, they can Google it or watch a YouTube video — again bypassing the VAR. With VARs no longer guaranteed a cut of every nearby sale, it’s a whole new ball game.

“Anybody can sell to anyone now,” says Cameron Carson, SVP of Engineering at Hawk Ridge Systems, at the inaugural one-day Hawk Ridge Systems Partner Summit recently held in San Francisco. Carson has experienced tectonic shifts in the VAR landscape firsthand. Over the years, the disappearance of sales territories has been bad news for some and good news for others. Some VARs will hang on to the seat they have always enjoyed. Others will gladly remove their seat belts and feel free to move about the cabin. Being one of the bigger VARs, and getting bigger with each acquisition, Hawk Ridge Systems is well positioned to benefit from that freedom.

Not your father’s Hawk Ridge

The Hawk Ridge Systems Partner Summit combines the best of SolidWorks’ annual 3DEXPERIENCE World with the intimacy of a local user group. Instead of dozens of user stories, there’s a curated few with ample time between presentations and a format conducive to getting to know other users.

Attendees come from a diverse number of Silicon Valley firms. Some from big firms “my mother would recognize,” and also small firms doing engineering to order or reverse-engineering work.

It’s unusual for VARs to hold partner summits and invite the media, but Hawk Ridge Systems isn’t your usual VAR. They have taken the traditional role of the VAR and expanded upon it, building upon design and manufacturing services to try to understand design engineers’ business issues.

“We’re not the old Hawk Ridge. We’re not just selling seats of SolidWorks or doing simple PDM implementations,” says Carson, leading off the Summit. We’re looking to see how we can adapt and thrive in this environment alongside you. With the solutions provided we can arrive at that single source of truth. It was 20 years ago when I was in industry and hearing model-based information was right around the corner. We would move from paper to PDFs with a giant vault and BOMs. But we’re still not there yet, and today, engineers and designers are being asked to do more. Software is getting more sophisticated and capable while the IT departments are shrinking.”

Increasingly sophisticated and varied software presents a challenge for design and manufacturing teams. Even for companies as expert in software use as Knapheide (maker of truck bodies with their own presentation), some software is either too infrequently used or comes with such a steep learning curve that it’s best left to specialists.

Knapheide may have expertise in all things SolidWorks, but when it came time to program a custom solution for stacking their truck bodies, they called upon Hawk Ridge Systems. Hawk Ridge Systems’ stacking solution is neatly integrated into the SolidWorks interface and other customizations.

“We have expertise in specialties the design engineers don’t,” says Cameron.

Slow to solve  — and why that’s a good thing

Next up is Andrew Parkhurst, who conducts a master class in patience and listening.

“If I had an hour to solve a problem, I would spend 55 minutes studying the problem and 5 minutes solving it,” Parkhurst attributes this quote to Albert Einstein.

Studying a problem is more than just listening, nodding and taking notes. Parkhurst recommends a letter of understanding (LOU) to capture critical business issues.

“If we don’t take the time to delineate the problem and don’t know what we’re trying to solve, how do we know we’ve solved it?”

Hawk Ridge Systems offers the full breadth of the SolidWorks and Dassault Systèmes portfolio as well as third-party solutions, including HCL CAMWORKS, DriveWorks, 3D printing hardware from FormLabs, HP and Markforged and HP, laser scanners from Artec3D and Creaform, and more. Yet, Parkhurst’s discipline is rock solid. He keeps from blurting out an obvious solution to a business issue, as I would have.

For example, DriveWorks’ design engineers have gone wild with customization requests. Surely, Hawk Ridge Systems will offer them DriveWorks.

Instead, he suggests we help each other.

“Sometimes the solutions can be right in the room,” he says and encourages users to share their issues with each other.

And they do. Company A (a household name we can’t divulge) has an Excel infestation. Company G (another Silicon Valley giant) needs help with version control. Company T initially sourced most of its components but is now bringing its manufacturing in-house. All of them have problems moving data between data islands.

The room has been transformed into a lively discussion of a type not seen at big annual user meetings.

Parkhurst stays in listening mode, the Cheshire Cat that knows the secrets of Wonderland and will share them when the time is right. He sympathizes, repeating the engineers’ problems to assure them he understands.

“With CAD, PDM, PLM, CAM, CRM, ERP … engineers are drinking alphabet soup,” he says.

HI not AI

“How many have undergone an ERP implementation?” Cameron asks later. There’s a shudder in the room. It was not one of the city’s earthquakes.

Selling design and manufacturing software across industries should give Hawk Ridge Systems a unique perspective and insight into the broader world of a company’s business. They would’ve been able to observe what works and what doesn’t, especially among divisions of a company, each with its disparate systems and data silos.

Hawk Ridge Systems, privy to a vast pool of customer data, with a history of observing problems and offering solutions, is HI (human intelligence) over AI, but the concept is the same. Resellers have been deep learning on data, the same as AI does with neural networks. But resellers’ HI has one big advantage. They have seen business bottlenecks and blasted through them with software solutions, either off-the-shelf or custom-made, in a way AI still must learn.

Growing too fast for its own good

“How many companies have growth targets?” asks Parkhurst. All nod. After all, isn’t growth good?

Not necessarily, says Parkhurst. Can a company handle the growth? Are its users sufficiently trained? Can manufacturing scale? The more business systems and processes they have, the better off they are, right?

Not necessarily. Systems and processes can be suffocating. Parkhurst gives an example of a company that was acquired and forced to use the big company’s systems. They had a difficult time of it, Parkhurst states. It hindered their creativity and productivity.

An experts’ expert

The changing market for VARs has led to consolidation. Hawk Ridge Systems has been an active acquirer. “We must have had seven acquisitions,” says Cameron. It’s true. Since 2017, Hawk Ridge Systems has acquired:

1. Symmetry Solutions
2. Cimtronics Midwest
3. Parson Technology
4. Quest Integration
5. CAS
6. Design Point
7. Access Manufacturing Systems.

No longer confined to Silicon Valley (where it supports notable tech giants Amazon, Tesla and Google) Hawk Ridge Systems has created a VAR empire that stretches from coast to coast and North to South, with 26 offices in the US and Canada serving 33,000 SolidWorks users.

The post Your business is my business, says SolidWorks reseller appeared first on Engineering.com.

Engineering.com: We discussed how Fusion will help on the design side by automatically generating dimensions. How big of an assist will this be?

Stephen Hooper: On average, engineers can spend a couple of days a week doing technical documentation. It’s a huge savings if we can automate even 50% of that. There’s a huge skills gap in the world of engineering. Any time we can free up a skilled engineer to focus their attention on design, creativity, innovation and engineering work versus technical documentation is a huge benefit for end users. This is what would fall into the category of automation for us. There are several automation projects we’re in the middle of.

Engineering.com: What about the use of AI for manufacturing?

Hooper: We talked about CloudNC at Autodesk University. There’s one example on the manufacturing side that is advanced now. It’ll create soft jaws for tooling. It’ll then create all the machining strategies based on machine tool. So rather than a production engineer wasting time producing pretty generic machining strategies, this add-in will examine the topology and then build specific machining strategies without the user being involved. One thing I would say here, and it’s true for all of the AI applications, is we focus on generating native content. We do this because we don’t believe in cutting the user out of the process. We should do 60 to 70% of the work for the user but still leave the user in complete control of how to edit the output. In this case, you’ll see that it produces standard Fusion machining strategies. It also looks at the cutting parameters. If it’s aluminum and you’re cutting it with a 40 mil end mill with three teeth, one can have it analyze the cutting parameters, speed, feed, and depth of cut. Then, it can investigate where the tool is likely to break, which is in red, and where the optimum cutting strategy is, the maximum speed and feed, and the depth of cut so that one get the best operational efficiency from  hardware.

Engineering.com: Thanks for the examples for drafting and CAM. What about AI helping with design, though?

Hooper: On the creative side, this is what I call augmentation. We’re not going to generate designs for people who don’t have a designer involved. It can become an assistant. As with software development. Microsoft has introduced what they call Copilot. The idea of Copilot is it doesn’t generate code and removes the necessity of an engineer. It sits with the engineer and helps them get rid of some of the laborious work they are doing. One of the things engineers hate doing is commenting code. They’re trying to solve a problem. Commenting code is super important for documentation and revision later on, especially in large teams. Copilot will automate the process of commenting in code. It’s an important function, but one that can be automated and it allows the engineer to focus on the creative side of problem solving.

Engineering.com: Autodesk acquired BlankAI. How will BlankAI’s technology be implemented?

Hooper: BlankAI will come in on the industrial design side. BlankAI takes all historical industrial designs and trains the model on those industrial designs. Then, when you start producing the next industrial design, blank AI will assist by helping incorporate company design language. Suppose you’re at BMW and want to produce a new vehicle architecture and are looking for a new industrial design. In that case, there are aspects of design language to incorporate and ensure that they remain consistent across all portfolios, such as edge design, the kidney grills, and the different aspects of BMW’s aesthetics that carry the brand identity.

In that respect, we’re not trying to replace the designer or generate the surface model for a vehicle. We’re trying to augment the design after they make choices, such as the vehicle’s width. Say you’re working on a coupe and want to incorporate some historical brand identity. The AI will interactively work with the designer. It lets one set up sliders. Say you’re working on the i4 and want to influence the body design of the i4 with the BMW 3 Series, which is what they did. The i4 has the same architecture as the 3 Series and carries many of the same styling cues. You could start with a foundation or surface model at the simplistic stage and then use BlankAI to introduce aspects of the 3 Series design cues into that initial early-stage conceptual design. That is when we start to get into customer-specific IP. BlankAI would be under that model and train it specifically on their data for their use. This is an area that we’re still exploring.

Engineering.com: That is a very specific example. How can AI-assisted design be rolled out to the whole design community?

Hooper: We have yet to figure out how to provide that en masse. It may be that we look at something such as ChatGPT. Everyone uses base-level of ChatGPT, but Autodesk has licensed advanced technology from OpenAI. We have a vertical version. It takes the base ChatGPT and trains it with a differential we call AutodeskGPT. We use it internally with developers and support staff. It takes the base of ChatGPT and trains it with extra data around Autodesk products and methodologies. It’s only available to us because it has the IP. That’s one idea we’re exploring.

Again, it will come down to whether you’re providing horizontal functionality, such as spellcheck, grammar, and dimensioning in a drawing versus vertical capabilities specific to the organization. As a customer, that represents IP. Other vendors are doing the same sort of thing. Adobe is working in that area. They’re injecting AI tools into traditional workflows and combining them. Have you ever used the Photoshop implementation and interacted with the visualization? One can use standard selection set modifiers in combination with generative AI. That is a powerful approach. They still leave the user in control.

To be continued …

The post How will Autodesk use AI for product design? Part 2 appeared first on Engineering.com.

They stab it with their steely knives
But they just can’t kill the beast.
—-Eagles, Hotel California, 1977

3D may have won the war against 2D, but pockets of resistance continue to fight on. It was supposed to have been a quick war. With all the advantages the forces of 3D had on their side, surely 2D would surrender quickly.

I started teaching 3D CAD in 1989 after realizing the enormous potential of 3D. I evaluated and implemented enterprise CAD programs from McDonnell Douglas (which have evolved into NX) and Applicon Bravo (now extinct) in my first full-time position as an engineer. Both programs ran on expensive workstations connected to DEC’s VAX minicomputers. I taught design, engineering fundamentals and CAD at a community college using AutoCAD, a program that was then, as now, known primarily for its 2D. I wrung as much 3D out of AutoCAD as possible, modeling in 3D wireframe and solid modeling. AutoCAD had licensed ACIS and was able to do CSG (constructive solid geometry) and through a combination of primitive shapes, hundreds of blocky, useless parts were born. I was betting on 3D as the imminent future. I hedged my bet by showing how to make front, top and right views. With view creation being push-button easy, I had to devote only a few classes to 2D.

But 2D still hangs in there… literally. I have just hung D-size drawings to the wall for the contractors remodeling my house. They’re familiar with drawings. With CAD, not so much. With 3D, forget about. A year ago when I started planning the project, I dreamt of seeing the remodeled space in 3D, fully rendered and in VR. This was a wake-up call.

Rumors of my death have been greatly exaggerated.
—-Mark Twain, acting the part of 2D

A rush to cut

I have no doubt that the ease with which a pencil can be put to paper to produce drawings is responsible for a perceived speed advantage. A colleague of mine and a contemporary from the drafting table era dismisses CAD, and 3D CAD in particular, continually goads me with “If CAD is so great, how come I can sketch a part on a napkin and take it to the shop before you can start your CAD program?”

Any argument to the contrary falls on deaf ears. I know 2D’s speed to be a false promise. Many a time have I rushed into the shop with such a sketch only to cut parts and regret I haven’t drawn to scale, cut to the wrong dimension, discover detail not detailed or run out of material or parts. I totally understand the hurry to get started, to start cutting.

A little history

2D champions may argue that it a natural form of documentation since we have been drawing in 2D as long as we have been on Earth. They draw support from the cave painting of a pig hunt in the caves on the Indonesian Island of Sulawesi said to be over 50,000 years old. However, 3D predates cave paintings by eras. Archaeologists found a 3D likeness of a person that is older than cave paintings by hundreds of thousands of years. That was the Tan Tan statue, which broke the $100 million barrier for statues of any age in a 2014 auction. It is estimated to be 300,000 to 500,000 years old.

The tail wags the dog

The way to describe an object so it can be manufactured, conveyed to others, or recorded for posterity, is influenced in no small part by the medium at hand. For as old as civilization, the medium has been 2D, such as the wall of caves, stone tablets, papyrus, parchment, scrolls … Maps show the Earth as flat. So pervasive became 2D over millennia that it may have distorted our reality. How else to explain those who saw the Earth as flat long after it was proved otherwise? The world we show as flat becomes the flat world we live in. Ergo, our walls, buildings, roofs … all are flat.

3D is more natural, but it doesn’t matter

Bam! That’s the gavel coming down. 3D is the natural way of describing an object. But no judgment will convince everyone to put down their pencils or switch their CAD to 3D mode.

Millions of 2D practitioners make 2D projections of 3D objects. For them, 2D is not an abstraction of reality but the simplest, most elegant and most efficient way to describe an object. With its set of standard views, it is systematic and complete.

Orthographic projection may not come naturally to all, but what does not come naturally can be taught. Those who can’t get the picture are taught to create the conventional views (front, top and right views for a mechanical part or the plans and elevations of a building) using construction lines. It’s a science, not an art.

Once learned, often painstakingly, 2D drawing and drafting can become second nature, and like every lesson learned after much time and effort, it becomes worthy of retaining and repeating, not casually abandoned. Abandoning 2D would be tantamount to admitting time and effort were wasted learning an unnecessary skill.

Or is it the Stockholm Syndrome, us in love with a master who was once cruel to us, but one we now depend on and is our only hope of survival?

2D will, no doubt, continue while its practitioners still have breath in their bodies and perhaps a generation afterward. There’s just so much of it in the system. And so we are bound to ask for a while longer, “Why will 2D CAD not go away?”

The post Why are we still using 2D CAD? appeared first on Engineering.com.

You’ve probably driven behind vehicles with a Knapheide logo. The logo is big and bold and on the back of many a hard-working pickup truck. Knapheide doesn’t make trucks, just what’s bolted onto chassis in the back of trucks to secure tools and equipment. There are many truck bodies from which to choose, each purpose-built for a given trade or need. One completely encloses items; another holds fuel; another holds a crane. The Knapheide logo is on them all.

History of Knapheide

The idea that a vehicle could be adapted to haul equipment or goods came to Harold Knapheide in 1910 after he strapped the body of a horse-drawn wagon to the back of a Ford Model T.

The company had been making horse-drawn wagons since 1848 when Herman Heinrich Knapheide, a German immigrant, settled in Quincy, Illinois. It was the year before the California Gold Rush, and Knapheide’s wagons were used by many for the westward voyage. Knapheide prospered. But then it was one existential crisis after another — the Civil War, the obsolescence of horse-drawn carriages, the Great Depression and two devastating floods.

Through thick and thin, the company managed to survive. Today, it thrives with Harold “Bo” Knapheide at the helm, the sixth generation of family leadership. The company has more than 2,500 employees who design and manufacture equipment mounted to trucks, including heavy-duty platforms and enclosures, road-clearing devices such as snowplows and cranes.

Hotbed of SolidWorks experts

We learned about Knapheide at the inaugural Hawk Ridge Summit held recently in San Francisco from Martin Ohnemus, the PLM/CAD Administrator at Knapheide.

Knapheide is proud of keeping all its manufacturing in the U.S. and ensuring its design and manufacturing software is American, all of it from SolidWorks. Neither company makes a big show of patriotism or waves flags. They don’t need to.

Ohnemus makes no flashy entrance. He doesn’t need to. His work at Knapheide is impressive enough.

SolidWorks has the most rigorous program for certifying its users. The most coveted certificate is the CSWE, or the Certified SolidWorks Expert designation, which comes after passing a grueling four-hour hands-on exam. Most companies are fortunate to have one CSWE among all its users. Knapheide has 28.

“SolidWorks says we have more CSWEs than any company in the world,” says Ohnemus.

Ohnemus oversees 350 seats of SolidWorks and PDM. Knapheide also uses Simulation, SolidWorks Simulation Visualize, CAM, eDrawings, DraftSight, Inspection, and Composer, as well as third-party applications such as Flatter Files and DriveWorks. By day, Ohnemus manages all the company’s CAD, CAM, CAE, and PDM training. By night, he leads the local SolidWorks user group.

Ohnemus counts 450 SolidWorks certifications at Knapheide, including the less demanding CSWP (Certified SolidWorks Professional). Awards are also issued for certifications in specialties, such as sheet metal and surfaces. The SolidWorks site lists 22 possible certifications.

Ohnemus encourages users to get certified in as many areas as possible. Each time a user receives certification, they are issued a sticker they can put on a framed certificate that shows all possible certifications.

Current and former CEOs of SolidWorks Manish Kumar and Gian Paolo Bassi were so impressed with Knapheide’s certifications that they flew from Boston to Quincy (where the company is still headquartered) and visited for three days.

Train, train, train

Instrumental in keeping SolidWorks users trained is SolidProfessor, which Ohnemus refers to as “the ultimate CAD training platform.” It’s a library of lessons on every aspect of SolidWorks and of great use in getting ready for certification exams.

“The nice thing about SolidProfessor is companies can add their standards and documentation,” says Ohnemus.

Every SolidWorks app

SolidWorks is the CAD and CAM software vendor of choice at Knapheide since the manufacturer has used virtually everything SolidWorks offers. Starting with the core SolidWorks mechanical CAD program, they added SolidWorks Electrical for the wire harnesses after first drawing wires as lines without electrical or physical meaning. Ohnemus recalls many mistakes were made.

“SolidWorks Electrical includes routing, but rather than train the EEs in using MCAD, we trained ourselves to use Electrical,” he says.

Knapheide uses SolidWorks Inspection, too. A 90-minute video produced by Ohnemus was enough to get nine engineers up and running on inspection in three weeks.

Knapheide decided to make their videos after watching users get frustrated by having to sort through too many videos of poor quality or irrelevant to their work.

“There are a lot of YouTube videos, and you can almost always find one, but it takes time to sort through them,” says Ohnemus.

Users would also ask each other questions if they didn’t know something, but the other user would often need to learn the answer.

“So, the two of you must go to a third person. Now, there are three of you trying to find an answer and not getting any work done. We estimated it took 45 minutes of lost time every time someone asked a SolidWorks-related question.”

They use Visualize to show design engineers how truck bodies are in pristine conditions rather than use photographs of actual equipment.

“Our stuff is out there in the field and getting banged up,” says Ohnemus. “But in the renderings, it looks perfect.”

Knapheide has found SolidWorks Composer helpful in making shop manuals, user manuals, and assembly instructions.

“We used to give them a stack of drawings and expected the artists to figure out what the parts looked like,” says Ohnemus. “Now we make the illustrations ourselves with Composer.”

With Knapheide in business for so long, they have seen several competitors come and go. He says using SolidWorks’ more specialized modules to keep up with new competition.

“When a new company jumps in, they will be jumping in with all the latest, greatest software,” he says. “We can’t afford to fall behind.”

But even with all of SolidWorks’ and third-party tools at their disposal, there are still times when off-the-shelf software is insufficient, and it’s time to develop  own customization. One example is Knapheide needing to show how truck bodies could be nested to get the maximum number on a truck trailer.

“There are programs made to stack containers, but all assume flat, rectangular shapes. Knapheide’s truck bodies are anything but flat and rectangular.”

This is where they had to call in the cavalry, also known as the reseller.

“We got together with Hawk Ridge and explained the need for stacking shapes. They wrote us a program to do it,” says Ohnemus.

Solid models, databases, graphics, and video make for incredible amount of data.

“We were using 40 terabytes of storage when I left [to speak here] and we’re adding 20 gigs daily.”

Ohnemus is a strong supporter of user groups, and with so many users at Knapheide, he has no trouble with membership.

“Our local user group is all Knapheide employees,” he says.

He encourages taking advantage of the SolidWorks user group program, now headed by Dan Wagner, which will provide $400 per meeting.

“That’s enough for food and venue expenses,” he says.

Ohemus encourages users to join or even start their own user group. And don’t be surprised if Manish and GP show up.

The post How did Knapheide create the most SolidWorks experts? appeared first on Engineering.com.

Earlier this year, we met with Stephen Hooper, vice president of Software Development, Design & Manufacturing at Autodesk, to discuss how Autodesk will use AI for mechanical design. It is part of an ongoing project to determine how all the major CAD and CAE companies will employ AI — if at all. So far, everyone has said they would use AI in some form. More than anyone else at Autodesk, Hooper oversees the road map for Autodesk’s product design software, such as Fusion and Inventor.

This follows our interview with Mike Haley, Autodesk’s Senior VP of Research, who is laying the groundwork for AI across all of Autodesk’s industries (design and manufacturing, AEC, and entertainment). See part 1 of our conversation with Mike Haley here.

The following article is mainly based on our conversation.

We find Hooper eager to talk about AI and its implementation in future versions of its products, which is unusual since most software companies have been reluctant to discuss features of products yet to be released. We attribute this to the desire to show that they are answering the loud call for AI after the hysteria over ChatGPT.

Engineering.com: Would you clarify your role at Autodesk and how you work with Mike Haley?

Stephen Hooper: Mike is working on definitions of foundational large language definitions for the whole company. He’s not responsible for product creativity in the different industry verticals. My job is to work with Mike and figure out how he might build generalized services and help him understand market opportunities from a manufacturing perspective and define them so that we can prototype, test, and then subsequently build production-ready code based on his generalized services in Fusion. Fusion is our target tool of choice for manufacturing obviously (for reasons I can explain in a moment). Then Mike worked on building a generalized learning model, and my team worked on implementing it on the client side so that it would show up as functionality in the product.

Engineering.com: So, when we see the next release of Fusion, will we see any AI?

Hooper: We’re introducing what we call drawing automation. It’s a drawings automation service. It will take any assembly, break it down into its parts, then take each one of those parts, create the layouts, scales… It does all the view orientations and annotates all the drawings for you. It’s probably easier to show you than it is to describe. I can give you a rapid overview of it. Here is a base plate from a machine. [Refer to the image above.] Fusion will pick out all the individual parts and create drawings. The one part at an angle [part 4 above] is correctly oriented in the drawing. It does all of this on the cloud, not on the client side. You choose a couple of critical dimensions, and AI will make all the rest. It has scaled and annotated the drawing for you. There are still some ways to go with this, such as geometric tolerancing and dimensioning. But already, it will count the parts [for a parts list]. But rather than jump ahead and do all the dimensions, because, as you know, you can interpret a drawing in different ways, it gives you options. On the side, you see the optional ways it can be dimensioned. You can have some basic dimensioning to fill in the rest yourself. Maybe you like a particular dimensioning style, but you want to reduce the dimensions’ density. You can interact with it. It’s not perfect yet. It does an excellent job of producing all the drawings and laying them out. It’ll do balloons, bills of materials, annotations, baseline dimensioning, ordinate dimensioning… It won’t do things like drawing notes or geometric tolerancing and dimensioning. Those are sophisticated areas. We’re working on them.

One of our next steps is to use a large language model for the notes in the drawing rather than train a specific model on things like annotations.

The other thing we’ve done is to instrument the product. We instrumented because even if you train a large model on vast data, you still don’t get things perfect. It would help if you refined the model iteratively. In this example, one of the leader lines crosses another leader line, so where it says two by six by 20 holes [see circled area in the image above], it’s crossing the leader line with the text annotation [marked in red in the image above]. The user would manually drag that dimension after we generate it. Because we’ve instrumented the product, we will see that the user has made that amendment, and then we will retrain the model so the user no longer has to drag the dimension. That will happen over some time. This way, the product will get more and more refined as more and more people use it.

Engineering.com: Can you talk more about using customer data? Is that not proprietary?

Hooper: That’s an important aspect to consider. First of all, you’ve got to have access to a lot of unique data. Otherwise, you might as well license something. In the case of drawing notes, there’s no point in building a large language model. We might as well license OpenAI and build on top of it with specific training on technical documentation. But if there’s something specific to your industry, like the actual drawing layouts themselves, then Autodesk is in a perfect position. We have vast internal data on drawing specifications, standard styles, layouts, etc. Plus, we can instrument our product to learn from users’ interactions. That is not proprietary data we are generating. We’re not learning from somebody’s design, which I call horizontal AI. This is vertical AI that helps augment a user’s design, creativity, and innovation. We are very sensitive about data protection.

Let’s use Grammarly as an example. That’s what I call horizontal functionality. Grammarly is not AI-based but updates its algorithms based on your use. No one would care that Grammarly learns from your grammar, spelling, or accuracy. You don’t care about that. Especially as a writer, you would care if Grammarly started producing articles for other people based on your IP and the creative content you’ve created. The same thing is valid here. No one cares if we train a model based on drawing standards because everyone works on drawing standards. It just has productivity benefits without IP issues. This is an excellent area for us to apply machine learning.

Continued in Part 2.

The post How will Autodesk use AI for product design? appeared first on Engineering.com.