
compare the features of three sorter designs with data and combine the best features into one improved design.
Phase 4: Build
Do Now. Imagine a bike with the brakes of one bike, the tires of another and the frame of a third. When could mixing parts make a better bike? When could it go wrong?
The best final design always comes from just one of the first designs, unchanged.
Sorting engineers, every design taught us something. Today you will look past “which design won?” and ask “which feature made each design strong?” Then you will combine those features into Version 2.
Step 1: list features, not designs. A feature is one part or choice, such as the number of turns, the shape of the core or how the sorter is held. Step 2: link each feature to data. Step 3: check that features work together and stay inside the constraints.
Worked example 1: a feature table. Priya’s team used their trial data (sample data).
| Feature | Found in | Evidence | Keep? |
|---|---|---|---|
| 40 turns | B, C | Sorted 8–10 caps vs. 5 with 20 turns | Yes |
| Two cells | C | Fastest, but 1 cap stuck after switch-off and the coil got warm | Test again |
| Flat bolt-head end | B | Picked caps evenly; no foil rode along | Yes |
Worked example 2: a claim about the combination. Their claim: “Version 2 will use 40 turns from B and C, the flat bolt-head end from B, and one cell instead of two. The data shows 40 turns raised the rate. The flat end avoided foil. One cell kept release clean and the coil cooler.” They also noted a trade-off: one cell may sort a little slower, so they will add a wider sweep to win back rate.
A trade-off is giving up some of one thing to gain another. Naming trade-offs honestly is part of a strong engineering argument.
Fill in your own feature table as you work through the blocks.
Plan Version 2.
Make a claim: which feature from EACH of your three designs best meets the criteria and constraints? Cite one number from your data for each.
Sketch Version 2. Label which design each feature came from, and name one trade-off you accepted.
Version 2 is on paper, sorting engineers. Next class we see how real recycling centers sort metals.
Exit ticket: name the single feature you are most sure about keeping, and the data that makes you sure.
How confident are you that you can I can use data to pick the best features from several designs and combine them.
Materials: team data tables and graphs from 2.16, feature table handout, colored pencils.
Setup: 12 minutes for the feature table, 10 for the claim, the rest for sketching. Post one strong claim as a model.
Listen for: “we’ll just use Design C” with no look at its weaknesses; features chosen without data; combinations that break the low-voltage constraint (no more than two cells).
Support / stretch: give a feature table with the first row filled in. Stretch: borrow one feature from ANOTHER team and justify it with their data.