Development Is The Gap Between A Sketch And A Build
Your chosen idea is a concept, and a concept is missing most of the information needed to make anything.
Development is where the remaining decisions get made and recorded, meaning exact sizes, joints, fixings, clearances, finishes and assembly order.
It sits between your justification and your planning drawings, and it is where most of the real design work happens.
The output is not one perfect drawing but a run of pages showing version 1 turning into version 4 with reasons attached.
Concept Versus Buildable
A concept says folding tray with a hinged leg, while a buildable design says two 3 mm plywood leaves joined by a 40 mm brass butt hinge with M3 screws.
The test is whether you could write a cutting list and a step by step making plan from what is on the page.
Any step that would still need a decision is your next development job.
Resolve The Details A Sketch Let You Ignore
Joints and fixings means which joint, which glue, which screw size, and whether it can ever be taken apart again.
Tolerances and clearances matter because a 3 mm sheet will not go into a 3 mm slot, so cut the slot at 3.2 mm and write that down.
Kerf is the material a cut removes, roughly 0.2 mm on a laser cutter, and it changes every slotted joint you draw.
Ergonomics means real body numbers, such as a 30 to 40 mm grip diameter for a comfortable hand grip, plus control heights and viewing angles.
Electronics and code need the same resolution, covering component choice, pin assignments, power budget and the flow of the program.
Finish and edges means sanded to 240 grit and oiled, or acrylic edges flame polished, and which edges need rounding for safety.
Assembly order matters because some jobs become impossible once the case is glued shut, so decide the sequence now.
Example
Version 1 of the tray used a 6 mm dowel hinge, which split the 3 mm plywood on the first test fold.
Version 2 moved to a 40 mm brass hinge screwed into a 9 mm plywood spine, adding 60 g but surviving 200 folds.
The mass row in the matrix dropped from a 5 to a 3, and that change is recorded with the test result beside it.
Model It Before You Cut Anything Expensive
A model is a quick version made to answer one question, not a small copy of the finished product.
Card and hot glue answer questions about size, proportion and whether a hand actually fits, in about 20 minutes.
CAD such as Tinkercad, Fusion or Onshape answers questions about fit and clearance, and gives you flat patterns ready to cut.
A 3D print at 0.2 mm layers answers whether a clip, thread or snap fit works before you commit to it.
For digital products, a clickable wireframe prototype answers whether users can find the booking button at all.
Photograph every model, failures included, with a caption saying which question it answered.
Tip
Model at full size whenever the answer depends on the human body.
A half scale card model tells you nothing about whether the handle is comfortable to hold.
Test Components, Not Just The Whole Thing
Testing one part is faster, cheaper and more informative than testing an assembled product.
Make three versions of the joint, load each one until it fails, and record the load each time.
Test the material rather than the product, so soak the plywood offcut, bend the printed PLA tab, and run the LED at full current for an hour.
Test the mechanism alone, so the hinge got folded 200 times before the tray around it existed.
Record numbers rather than impressions, because "seems strong" is not a result you can use later.
Every component test is direct evidence for one of the success criteria you wrote earlier in criterion B.
Activity
Pick the part of your design most likely to fail and build only that part this week.
Load it, fold it, drop it or run it until something changes.
Write down what changed and at what point it happened.
Show The Design Changing And Say Why
Development pages have to show movement, so version 1, what went wrong, version 2, what improved.
Every change needs a trigger written beside it, whether that is a test result, a client comment, a material limit or a specification point.
Keep the earlier versions on the page rather than replacing them, since a folder holding only the final design hides all your thinking.
Annotate the change itself, such as "base widened from 90 mm to 130 mm because the loaded lamp tipped at 8 degrees".
Three or four recorded iterations is normal for an MYP project, and one is not development at all.
Common Mistake
Redrawing the same design more neatly is presentation, not development.
If nothing differs between two pages except line quality, the second page earns nothing.
You Are Finished When Nothing Is Left To Decide
Development ends when every part has a material, a size, a joint and a place in the assembly order.
Check by trying to write the cutting list, since any part you cannot give a size to is an unfinished decision.
Check the specification again, because development changes things and a point your concept met may now be failed.
Update the affected matrix row or add a dated note wherever development has changed the answer.
What comes out of development goes straight into your planning drawings, where someone else has to be able to build it.
Active recall
What is the difference between a concept and a buildable design?
Why cut a slot at 3.2 mm for a 3 mm sheet?
Give two questions a card model can answer in 20 minutes.
Why test a single component rather than the assembled product?
What has to appear beside every change you make to the design?