The first idea hasn't won
An idea can seem obvious simply because it came first. ChatGPT's role here is to force exploration: produce several genuinely different families of solutions, then put them through the same criteria.
Propose six fundamentally different architectures for a toolbox that turns into a workbench. Do not change only the shape. Vary the deployment principle, the layout of the modules, the mobility and the way the top is created. For each concept, state the main advantage, the main risk and the question to test on a prototype.
Set the criteria before scoring
The criteria come from the specifications. A possible list:
- ease of transport;
- stability in use;
- deployment time;
- access to the tools;
- estimated weight;
- likely cost;
- difficulty of making;
- maintenance and repairability.
A first qualitative comparison
| Concept | Strength | Main risk | Priority test |
|---|---|---|---|
| Lid-top | Few parts and simple opening. | Insufficient working height. | Full-height mock-up. |
| Folding legs | A true workbench position. | Locking, pinching and stability. | Prototype of one leg and its link. |
| Telescopic frame | Compact, adjustable deployment. | Weight, cost and mechanical play. | Mechanism test under careful load. |
| Stackable modules | Adaptable storage. | Assembly time and separate parts. | Timed scenario. |
| Two boxes + top | Large surface and good distribution. | Carrying several items. | Loading into the vehicle. |
Don't manufacture false precision
A score of "8.4 out of 10" looks scientific, but it means nothing if the data are unknown. First use simple ratings: favourable, uncertain, unfavourable. Replace them with measurements after the tests.
Give more weight to certain criteria
Not all criteria are equal. For a mobile technician, transport may be the priority. For a fixed workshop, stability and surface dominate. The weightings must be decided with the user, before the winning concept is known.
Play the opposition
The currently preferred concept is described below. Find five realistic reasons why it could fail with the user. For each risk, suggest a simple test to confirm or rule out the problem. Do not defend the concept.
The decision sheet
Concepts compared: list and versions.
Criteria used: taken from the specifications.
Available data: measurements and sources.
Uncertainties: points still estimated.
Concept chosen for prototype: a reversible decision.
Reason: three arguments at most.
Mandatory tests: what could still invalidate the choice.