Skip to content
Curiotide

Invent a Gadget That Solves a Real Family Problem

  • Ages 5-7
  • Ages 8-9
  • Ages 10-12
Illustration of a gadget with a dial and antenna drafted on a blueprint sheet, on warm orange waves

Invention projects usually start with “invent something,” which is why they usually stop there. Start instead with what makes your family shout in the morning, and the invention arrives on its own.

Ages 6 to 12

Time An hour to design, an afternoon to build

You'll need Paper, cardboard, tape, and a chatbot on your account

How it goes

  1. Spend a day noticing. Everyone writes down one thing that went wrong: lost socks, the bin bag that splits, nobody can find the scissors.
  2. Your child picks the problem and states it in one sentence. Naming a problem properly is most of engineering.
  3. Sketch three fixes before asking anything. Bad ones count. Three is the number that forces the second and third to be different.
  4. Now use the AI as a critic: Here is my design for a thing that stops the bin bag splitting. Ask me three questions about how it works. Do not redesign it. Questions expose the bit your child has not thought about.
  5. Fix the design and draw a final version with labels and measurements.
  6. Build it out of cardboard and try it on the actual problem.

Who does what

AI asks awkward engineering questions, and lists materials your child might not have considered.

Your child finds the problem, sketches, revises, builds, and tests on the real bin.

You supply cardboard, keep hold of the good scissors, and let it fail once before helping.

Make it easier or harder

For a five-year-old: skip the sketching round and go straight to cardboard. Young children design by building, and asking for three drawings first stops the project before it starts.

For a twelve-year-old: require a cost. Give them a budget of five household items and make them justify each one, then ask what the gadget would need to be to be worth actually keeping. Most invented gadgets fail that test, and working out why is more useful than the build.

Take it off the screen

The build is the point. A prototype that half works and gets iterated beats a beautiful drawing every time.

Hold a family demo. Your child presents the problem, the design, what failed, and what they changed. That structure is a real engineering report and they will not notice they are writing one.

When it goes sideways

The prototype will fail on first contact with the real bin. That is the intended outcome and worth saying out loud beforehand, so it reads as data rather than failure. Ask which part failed and whether it bent, tore, or slipped, then rebuild that part only.

If the AI redesigns despite being told to ask questions, ignore the redesign and pull out only the questions it buried inside it. There is usually one good one, and one good question is all this step needs.

One thing to watch

Chatbots will suggest components you do not have, and occasionally something involving mains electricity or heat. Read suggestions before your child acts on them, and keep the build to cardboard, tape, string, and rubber bands. Our guide to AI for science and STEM at home covers this in more detail.

Related: improve a cardboard prototype is the next step, and invent a toy and write its manual uses the same design loop.

Keep reading