AI review in a CAD workflow: how one engineer keeps GEHA fan projects moving

For AIdataNewsGlobal’s AIApplications column, the practical question is not whether a model can chat, but where AI can sit inside a design process. One engineer at GEHA uses AI review tools inside a CAD workflow, keeping Designcenter Solid Edge as the stable base while AI supports checking, documentation review, macro conversion and report routing. The company is in Nieuw-Amsterdam, the Netherlands, and has built industrial fans since 1957. Customers for most of its history included refineries, chemical plants and the heat-transfer suppliers that serve them; fan housings run up to 7.3 meters in diameter, while data-center cooling work is now the newest entry in the order book.

Application context: high-volume fan work with one engineer

For a global manufacturer of datacenter cooling systems serving the North American market, GEHA is producing fans of more than 2.1 meters (7 ft) in diameter in high volume. A single datacenter campus can require several hundred of these large fans.

Jan Bos, GEHA’s manager of engineering and R&D, says, “We make the biggest and most silent fans for data centers,” and adds, “That’s our specialty.”

GEHA supplies industrial air-cooling solutions. (Image: GEHA.)

What makes the application unusual is staffing. The engineering department at GEHA once had four people; when the oil market turned in 2017, it shrank to two. Today, Bos is alone with Designcenter Solid Edge and a growing kit of AI tools.

He says, “I’m on my own and running the whole show without problems,” and adds, “And I’m enjoying my work.”

Rendered model of a complete cooling system designed in Designcenter Solid Edge. (Image: GEHA.)

Background: from 2D CAD to a Siemens community role

In 1997, on version 3.5, Bos first saw Solid Edge when a colleague’s son modeled a part for him in an afternoon and placed views, a detail and a cross-section onto a drawing in minutes. He had been working in 2D CAD and was unimpressed by it, but after the introduction to 3D CAD he was hooked. In 2003, he brought Solid Edge to GEHA to move the company’s knowledge “out of the brains, out of the paper, into the model.”

He helped found the Benelux user group since then, became one of Siemens’ Community Catalysts, and presented at Realize LIVE on everything from Visual Basic automation to a refinery retrofit that GEHA engineered from a point cloud. This year in Detroit, his talk was titled “Practical AI as an Everyday Solid Edge User.”

His introductory slide reads: “Not a programmer. Just an engineer who got curious.”

Jan Bos, lead engineer and designer at GEHA, presents his AI design hacks to the community at Realize LIVE Detroit 2026. (Image: GEHA.)

Where AI enters first: a drawing review workflow for big projects

The old ritual lived on Sunday mornings, not in Sunday morning, two minutes, because a 150-drawing project took about two to three hours. On large projects, Bos had a ritual of printing the drawings and checking them at home with a red pen, a highlighter and a cup of coffee. He says, “It only goes wrong when it’s a big project,” and adds, “So, if it’s a big project with a lot of parts, you have to check it very well.”

In spring 2025, a full set was exported to PDF and loaded into Google AI Studio running Gemini 2.5 Pro; Bos told the model to act as the kind of old-fashioned drawing checker engineering departments used to employ, with close attention to the bill of materials (BOM).

The tool found uneven hole spacing on a perforated plate and asked whether it was intentional. It wasn’t. It flagged a transport wheel listed in the BOM with no diameter and suggested 150 mm, a value pulled from the detail drawing. “The AI did what a real drawing checker would do,” Bos says.

Typical fan in the GEHA product line: machine-produced fan rings are up to a diameter of 7.30 meters, with components such as drives, bearing blocks, louvre dampers and plenums. (Image: GEHA.)

Step path: turning a one-off check into a reusable review skill

AI review is now standard on every major GEHA project. Bos has since turned the checker into a reusable skill and taught it about the individuals that work in the factory. The people responsible for fan construction, assembly, work preparation and laser cutting are all in his system, so it frames feedback around them.

1. Export the drawing package as PDF.
2. Load it into Google AI Studio running Gemini 2.5 Pro.
3. Prompt the model as a drawing checker focused on the bill of materials (BOM).
4. Ask it to flag inconsistent hole patterns, missing dimensions, and manufacturability issues.
5. Review each result against the model and the shop-floor context, then keep the judgement with the engineer.

Given a PDF or a STEP file, the model will also flag a hole placed too close to a bend, where forming will distort it, providing design-for-manufacturability feedback that once depended on a shop-floor veteran catching it. Bos has used the same approach to check GEHA’s technical construction files against updated machine-safety regulations.

Bos says, “[AI] tells me, ‘Jan, you made a heavy construction, but there is no hoisting lug on this one, and your colleague Eric will not be very pleased,'” and adds, “It holds up a mirror.”

Trust boundary: what the AI is allowed to check

Does he trust it? He says, “Yes,” and calls the results “very solid.” Part of that trust is scope: the model checks his drawings against his standards, rather than roaming the open internet. The other part is that the AI doesn’t make the call. As one of his Detroit slides puts it, “AI amplifies expertise. It doesn’t replace it. You still need engineering judgement.”

The second front: using AI to convert old CAD macros and reports

The second front is automation. Like many long-time users, Bos had a library of VB6 and VB.NET macros that “sort of worked,” built by adapting examples. Using the open-source coding tool OpenCode with a paid Claude model, he turned an old VB6 utility called PM-Tabelle into Fit Table, a C# macro that connects directly to Designcenter Solid Edge without opening Visual Studio. Every GEHA macro is now C#. When the AI needs to understand Designcenter Solid Edge’s programming interface, it reads the programming documentation available online; as a starting point, Bos can hand it one of the roughly two dozen example programs he’s collected over the years.

He tested the limits with a property add-in that launches on his second monitor, something he’d wanted for years. After a few failed attempts, the AI asked whether he had other add-ins it could study, found one in a folder and produced a working version. Bos refined the layout by marking up screenshots with red arrows in Microsoft Paint. The total time to develop the concept was about an hour and a half. He ultimately preferred his macro and shelved the add-in, but the experiment once would have cost five or six days of a contract programmer’s time.

The same approach has cleared bottlenecks downstream of the model. A colleague used to spend half an hour on each of roughly 20 orders a week splitting Designcenter Solid Edge BOM reports into lists for purchasing, laser cutting and other departments. Bos gave Claude Code an emptied-out report as a template, and it built a Python tool that does the job with a right-click.

Some newer work skips CAD entirely. With Claude connected to SketchUp, Bos can photograph a dimensioned hand sketch of a bracket and get back a STEP file ready to import into Solid Edge or send to a 3D printer. In the other direction, he’s handed the model a STEP file of a GEHA bearing block and gotten back a bill of materials, plus a conversation about bearing selection and where the part could be made cheaper.

Bos went from whiteboard to STEP 3D model without opening a CAD tool. (Image: Jan Bos | GEHA.)

The base layer: why the CAD model still matters

For all of this, Bos is clear about what holds it together: AI is the layer on top, and Designcenter Solid Edge is the base.

He says, “Solid Edge is a very, very stable building block,” and adds, “Everything we communicate to suppliers, it’s always spot on.”

What he wants now is for the software to absorb the work he’s been bolting on. He says, “The nice thing about being an engineer is the creating part, solving the puzzle,” and adds, “We don’t like making a drawing, checking a drawing, printing drawings, doing conversions. All that stuff should be handled by Solid Edge.”

His Detroit presentation wish-list starts with native batch drawing management:

• gather every drawing,
• verify it is current,
• update hundreds of drawings at once,
• report what changed.

That would build toward AI that drafts and checks drawings alongside the engineer, ideally on a local model so data never leaves the building. Siemens is working toward AI-driven drawing creation and checking, and Bos sees Designcenter Solid Edge 2026 already heading that way. He’d just like it faster. “We are getting spoiled,” he says. “Why not?”

Jan Bos, third from the left, joins fellow Community Catalysts and Siemens staff at Realize LIVE Americas 2026. (Image: GEHA.)

Common misconceptions about this kind of AI adoption

• The AI now replaces the engineer. The case shows the opposite: AI amplifies expertise, while engineering judgement stays with Bos.
• The checker can roam freely. The useful trust boundary is that the model checks drawings against his standards, rather than roaming the open internet.
• You must be a programmer. Bos describes himself as an engineer who got curious, using documentation, old example programs and interface files rather than deep coding.
• The model should make final decisions. As the case puts it, it holds up a mirror instead of calling the shot.

Action suggestions for teams starting with a CAD review pilot

• Begin with a large drawing package, not a single sheet, because the old problem appears when there are many parts.
• Keep Designcenter Solid Edge as the source of record and put AI at review, macro conversion, and report routing.
• Use existing artifacts as context: PDF drawings, STEP files, BOM reports, old VB6 or VB.NET macros, and example programs.
• Define the reviewer role narrowly, including the bill of materials, manufacturability risks, and the factory people affected.
• Reserve regular time for tooling work; Bos keeps Friday afternoons free for sharpening the tools.

For AIdataNewsGlobal’s AIApplications view, the practical takeaway is that a single engineer can make a small department work by keeping the CAD base stable and adding AI where checking, macro conversion, and report routing used to consume time. Bos keeps Friday afternoons free for “sharpening the tools.” The feeling, he says, takes him back to that afternoon in 1997, when a cross-section appeared from a 3D model in seconds. He says, “It’s that magic feeling I had in ’97,” and adds, “It’s a multiple of that magic at the moment.”