Industrial Design & CAD

Kaern Schools — Professional Certificate Track

Built and taught on cycleCAD, the in-house browser-native B-Rep CAD platform from the cycleX project family.

Welcome to the course

Who this is for. This course is for aspiring and early-career industrial designers, mechanical product designers, makers, and engineers who want to move fluently from a vague brief to a manufacturable, documented product. You should be comfortable with a computer and basic geometry (angles, circles, distances). You do not need prior CAD experience — we build everything from zero. You do need patience, a willingness to sketch badly before you sketch well, and the discipline to name your features.

What you'll be able to do by the end.

• Translate a written design brief into resolved 3D form using sketches, references, and cycleCAD's Text-to-CAD and Photo-to-CAD assists.
• Build robust, editable parametric models that survive late changes without exploding.
• Assemble multi-part products with proper mates, motion, and interference checks.
• Reason about a human hand, body, and eye — ergonomics and aesthetics as engineering inputs, not decoration.
• Apply tolerances and GD&T so your parts actually fit when they're made.
• Produce manufacturing drawings and a clean handoff package a real shop can build from.

The tool: cycleCAD. cycleCAD is a browser-native, B-Rep (boundary representation — solid, watertight geometry) modeler. It runs in your tab, no install. Its distinctive features show up throughout the course: Text-to-CAD (describe a feature, get editable geometry), Photo-to-CAD (reconstruct form from reference images), an integrated FEA/DFM panel (stress, deflection, and manufacturability checks live), and a library of 150+ PBR materials (physically based render materials that double as physical-property sources — density, finish, cost). Everything you model is parametric and history-based, so we lean hard on that.

Your tutor: Cad. Cad has spent fifteen years between the studio and the factory floor — consumer electronics, medical handhelds, mobility products. Cad teaches the way a good shop foreman mentors: show the move, explain why the move, then put the tool in your hand and watch you do it. Expect analogies, war stories, and a relentless focus on intent over clicks.

Format. Six modules. Each has learning objectives, taught lessons (with the actual teaching scripts), a worked example, a hands-on build, the three mistakes everyone makes, and a check for understanding. The course ends with a capstone and a grading rubric.

Module 1 — From Brief to Form

Learning objectives

• Interpret a written brief and extract design intent, constraints, and success criteria.
• Establish scale, proportion, and primary axes before touching a single solid.
• Use reference imagery and cycleCAD's Photo-to-CAD and Text-to-CAD to bootstrap form quickly.
• Build a "form study" — a low-commitment massing model — and critique it against the brief.
• Distinguish exploration geometry from production geometry, and know when to throw work away.

Lessons

Lesson 1.1 — Reading a brief like a designer

"Let's start with a confession: most beginners read a brief like a to-do list, and that's exactly why their first model is wrong. A brief is not instructions — it's a negotiation between a user, a business, and physics. So when I read 'a compact, premium-feeling handheld scanner for warehouse staff,' I don't see one sentence; I see four hidden requirements. 'Compact' means it lives in a pocket or a holster — that's a size envelope. 'Premium-feeling' is about mass distribution and surface finish, not price. 'Handheld' commits us to a grip, a thumb reach, a center of gravity. 'Warehouse staff' means gloves, drops onto concrete, twelve-hour shifts. Notice I haven't drawn anything — I've turned prose into constraints I can measure. The discipline here is to write a one-page 'design intent' before modeling: who, what envelope, what must-nots. Think of it like a recipe versus a dish — the brief is the dish someone's hungry for; your intent doc is the recipe that gets you there reliably. When you skip this, you end up beautifully modeling the wrong thing. So before our next session, I want you asking: what is this brief actually afraid of?"

Lesson 1.2 — Scale, proportion, and the first massing model

"Here's something the software won't tell you: a model with no sense of scale is just a pretty lie. Before we make anything precise, we make something roughly right — a massing model, the clay rough-out. In cycleCAD I always drop a human-reference primitive first: a 78mm-wide hand block, or a credit-card-sized 86×54mm plate, so every shape I pull is judged against something real. Then I block out the product as simple boxes and cylinders — no fillets, no detail, just volumes and their relationships. Why bother? Because proportion is decided in the first ten minutes and defended for the rest of the project. Architects call it parti — the big organizing idea. If the proportion's wrong now, no amount of gorgeous filleting saves it later. I treat massing like framing a photo before I worry about focus. And cycleCAD makes this cheap: I can Text-to-CAD 'a 120 by 60 by 25 rounded slab' and have a block to react to in seconds. React, adjust, react again. The model is a conversation, not a monument. So as you build your first massing study — what is the one proportion that, if I changed it, changes the whole character of the object?"

Lesson 1.3 — Bootstrapping form with Text-to-CAD and Photo-to-CAD

"Now the fun part — and a warning. cycleCAD's Text-to-CAD and Photo-to-CAD are accelerators, not authors. Text-to-CAD is like a talented junior who hands you a solid block of marble roughly the right shape: brilliant for getting unstuck, useless if you treat it as finished. I'll type 'ergonomic pistol-grip housing, 130mm tall, trigger recess' and it gives me editable B-Rep geometry — real features in the tree, not a frozen mesh. That last part matters enormously: because it's parametric, I can grab the grip diameter and tune it. Photo-to-CAD is the same spirit for reference: I drop in three photos of a competitor's grip, and it reconstructs the surface so I can measure and learn from it — not copy it. The skill isn't prompting; it's judging. You must look at what comes back and ask 'is this serving my intent, or is it just impressive?' I've seen students fall in love with a generated shape that fought the brief the whole way. Use these tools to skip the boring 20% and spend your energy on the 80% that's actually design. When the AI hands you a form — what will you keep, and what will you have the courage to delete?"

Worked example — Warehouse handheld scanner massing

Cad works live: writes a 6-line intent doc (envelope 130×70×35mm, one-handed, gloved grip Ø34–40mm, must survive 1.5m drop). Drops an 86×54mm reference plate and a 78mm hand block. Text-to-CAD prompt: "rounded slab body 130×70×30mm with an angled pistol grip 130mm tall." The generated body lands in the feature tree as editable solids. Cad overlays the hand block, sees the grip is too fat at Ø46mm, and dials it to Ø37mm via the grip parameter. Photo-to-CAD imports two reference grips to compare thumb-reach. Result: a critiqued massing model with named features — Body_Mass, Grip_Mass, Trigger_Recess — ready to refine, not a finished part.

Hands-on exercise

Pick one brief: (a) a reusable coffee cup for cyclists, (b) a desk-mounted webcam, or (c) a child's flashlight. In cycleCAD: write a 5-line intent doc as a text note; place at least one human-reference primitive; produce a massing model from primitives and one Text-to-CAD pass; rename every feature meaningfully. Deliver one annotated screenshot explaining your key proportion decision. Time-box: 45 minutes. Do not fillet anything.

Common mistakes

1. Detailing too early — filleting and texturing a massing model, locking in proportions you haven't validated.
2. Treating Text-to-CAD output as final — keeping generated geometry you don't understand and can't edit.
3. No human reference — modeling in a vacuum so the object is secretly the wrong size.

Check for understanding

1. Name two hidden constraints buried in the phrase "premium-feeling handheld for gloved workers."
2. Why do we deliberately avoid fillets and detail in a massing model?
3. What is the one question you should ask of any Text-to-CAD or Photo-to-CAD result before keeping it?