# Urban Planning — steal this assignment
_EL3vate 2026 · Day 8 · Part 07 of 15 · build 3dfcaff_

## Try it Tuesday (90 minutes)
90 minutes, one contested Honolulu parcel with public documentation. Minutes 0–20: students pull the actual constraints from the record — zoning, height limit, existing use. Minutes 20–55: they open the embedded tradeoffs demo (or build one with the prompt) and, for three different height and density settings, write the plain-language consequence a resident would feel — shadow, traffic, who gets priced out, what is demolished. Minutes 55–90: a mock testimony round — half the room plays residents and reacts to each scenario; planners revise the wording of any consequence the residents did not understand. The deliverable is the revised consequence sheet, graded on whether a non-planner could act on it.

## The assignment
**Make it (Laser cut · layered).** A layered site model of a real Honolulu parcel: topography, existing structures, proposed massing, each on its own removable layer. Residents can lift the proposal off and see what is underneath, which is exactly the conversation that renderings prevent.

**Build it (AI chat · no code).** An interactive tradeoff explainer for the same parcel: move a slider for height or density and see the consequences a resident actually cares about, in plain language. Built in an afternoon, used in a real meeting.

## 4-week plan
- **Wk 1.** Pick one contested parcel with public documentation. Students extract the actual constraints from the record — and the topography, existing structures and proposed massing the layered model will render.
- **Wk 2.** Submit the layered-model cut file at the start of the week so fabrication runs in parallel. Build the interactive explainer; test it on someone outside the major and rebuild the parts they did not understand.
- **Wk 3.** While the model is cut, students prepare the meeting facilitation, not the presentation.
- **Wk 4.** Run a mock community meeting with the model and the tool on the table. Invite the neighborhood board if the timing allows.

## What this replaces
- **Replaces:** The studio site-analysis report with rendered massing diagrams.
- **What is lost:** The comprehensive drawn analysis and the rendering craft.
- **What is gained:** Students translate zoning into consequences residents can weigh, and learn that a rendering hides the tradeoff a slider exposes; the grade rewards communication with the affected public over presentation graphics.

## Where AI is bad at this
Ask a model for planning consequences and it produces authoritative-sounding figures — trip-generation counts, shadow lengths, affordability thresholds — with no siting data and no acknowledgment of local code, presenting rules of thumb as parcel-specific findings. It also strips the politics, describing a tradeoff in neutral technocratic language that hides who wins and who is displaced — exactly the information a community meeting needs surfaced.

## Rubric
| Criterion | Weight | What it assesses |
|---|---|---|
| Fidelity to the actual constraints | 25% | Consequences are tied to the parcel's real zoning and documented conditions, not generic assumptions. |
| Plain-language accessibility | 30% | A resident with no planning background can understand and act on each stated consequence; jargon and acronyms are removed. |
| Distributional honesty | 25% | Names who benefits and who is displaced under each scenario rather than describing tradeoffs neutrally. |
| Revision from feedback | 20% | The consequence sheet is measurably clearer after the mock-testimony reactions. |

## Starter prompt
> Build me a single self-contained HTML page with sliders that shows planning tradeoffs for a site. The variables are: [building height, unit count, parking ratio, open space, whatever applies]. When a slider moves, update plain-language consequences that a resident would care about: shadow, traffic, who can afford to live there, what gets demolished. No jargon, no acronyms. Assume the reader has never been to a planning meeting.

## Budget & logistics
- **Instructor prep:** 2 hours
- **Class time:** 90 minutes
- **Per-student cost:** $0 for the Tuesday version; roughly $15–30 per team in laser stock **[unverified]** for a layered site model in the four-week build.
- **Fabrication file due:** First day of Week 2 — submit the layered site-model cut file once the parcel’s record is extracted in week 1. PACE quotes 7–10 business days (up to 14 calendar days), so it clears before the week-4 mock meeting.
- **Calendar dependency:** The interactive tool needs no lead time. The layered model (topography, existing structures, proposed massing) comes straight from the public record extracted in week 1, so its cut file goes in at the start of week 2 and the up-to-14-calendar-day turnaround clears before the week-4 meeting.

## Three sizes
- **One session:** The 90-minute constraint-to-consequence session with a mock testimony round; no fabrication.
- **4 weeks:** The seeded four-week plan: extract constraints, build and test the explainer on non-planners, submit the cut file, then run a mock community meeting with the model and tool on the table.
- **One semester:** A full charrette on a real parcel — constraint extraction, an iterated public-facing tool tested on non-planners, a layered fabricated model, and a facilitated neighborhood-board session.

---
_PACE · Shidler College of Business · University of Hawaiʻi at Mānoa · pace.shidler.hawaii.edu/maker_
_All fifteen assignments, the demos and every handout: https://el3vate.vercel.app_