Define success before choosing a design
Engineering begins by identifying a problem. Criteria describe what a successful solution should do, while constraints set limits such as time, materials, size, safety, or cost. Measurable criteria make later tests useful because a team can compare a result with a target instead of deciding only whether a prototype looks good.
Engineers usually brainstorm more than one possible solution, compare the choices, and select a design to model or prototype. A prototype is built to answer questions and expose problems; it does not need to look like a finished product. Choosing one design also does not prove it is the only design that could work.
Use testing as a loop
A useful test matches the stated criteria, controls conditions when possible, and records observations or measurements. When a prototype misses a target, the location and type of failure can point to a change. The team can then rebuild, retest, and compare the new result with the earlier one.
Iteration means repeating this build–test–evaluate–improve cycle. Changing one important feature at a time can make cause and effect easier to judge, although some connected features must be redesigned together. Passing one test supports a limited claim about those test conditions; it does not prove a design is perfect, safe in every situation, or better on every criterion.
Remember these ideas
Key ideas
- Criteria define desired performance, while constraints limit the possible solutions.
- A prototype helps a team test an idea before committing to a final product.
- Measurements and observations should connect directly to the criteria being evaluated.
- Engineers iterate by using test evidence to make a reasoned change and then testing again.
Make the reasoning visible
Improve a paper bridge with test evidence
- 1
Define the challenge: span a 20-centimeter gap using no more than two sheets of paper and 30 centimeters of tape. Success means holding 40 coins for 10 seconds.
- 2
Brainstorm a flat sheet, a folded beam, and a corrugated shape. Select the folded beam, build a prototype, and keep the material limits the same.
- 3
Add coins in the same place and at the same rate. The bridge collapses at 28 coins, and the recorded observation shows the center folds before the supports move.
- 4
Deepen the center folds, rebuild, and repeat the test. The revised bridge holds 44 coins for 10 seconds, so it meets the criterion in this test; more tests would be needed before making a broader reliability claim.
Try the next step
Choose an evidence-based redesign
A model container must keep an ice cube from melting for 20 minutes while using no more than 50 grams of insulation. The first prototype uses 45 grams, but the ice melts after 14 minutes and the lid feels warmer than the sides. What should the team change and measure next?
What could the team conclude if the revision improves the time to 18 minutes but still misses the criterion?