Multi-Objective Optimization: The Design Move · Earth & Env Science

Home › Unit 6: Biosphere & Land Use › Block 4: Multi-Objective Optimization

AI Partners available for today's work

The Urban Planner The Equity & Resilience Partner

Remember: every AI use gets documented. → AI Documentation Template

Story

What today is about

The central move of the unit: allocate the land, and watch every objective respond at once.

Today you make the move the whole unit is built on. You take your locked area, picture it as a grid, and allocate that grid across the 8 urban land uses (residential, mixed-use, commercial and markets, civic and services, green space, blue space, streets and mobility, and the existing fabric including informal settlements). The pieces have to add up to the whole area, so every square you give to one use is a square you took from another. That is what makes this a multi-objective optimization: you are not solving for one goal, you are balancing seven at once, and no allocation wins on all of them.

The planned Land Allocation Optimizer is the tool for seeing this. You redesign the grid and watch seven objective scores move in real time (15-minute access, housing capacity, green-blue cover, heat comfort, flood resilience, equity and displacement, and mobility). When two objectives cannot both go up, the tool shows you the Pareto frontier: the line past which gaining on one objective always costs you on another. Hitting that line is not a failure. It is the moment the real design problem starts.

Two AI partners enter here, and they are built to disagree. The Urban Planner pushes density, housing supply, and economic vitality: build more, build up, intensify. The Equity & Resilience Partner pushes green-blue space, flood and heat protection, and the rights of people who could be displaced: keep land open, keep it permeable, do not clear the people already here. Your job is not to average them and it is not to pick a side. Your job is to synthesize their disagreement into an allocation you can defend, knowing exactly what you gave up to get it.

Concepts

Before you start the work

Start with the Interact: your first scored allocation. The Watch sets up the trade-off logic behind it, and the partner work pressure-tests what you chose.

Placeholder

Watch (~6 min)

TBD: what multi-objective optimization means, and why a single best answer rarely exists. The Pareto frontier explained on a simple two-objective example (housing versus green space) before you face all seven.

Whole-class, projected.

Placeholder

Interact (~25 min)

TBD: the Land Allocation Optimizer (planned tool). Redesign your area on the grid and watch the seven objective scores respond. Find one place where pushing an objective up pushes another down, and note it. While the tool is in development, this runs as a paper grid with a scoring table.

Solo · the day's anchor.

Placeholder

Work the partners (~10 min)

TBD: take one allocation question to each partner. Ask the Urban Planner where to add density, and the Equity & Resilience Partner what that density would cost. Log both in your AI Documentation, including one thing a partner got wrong.

Solo. The signature partners land today.

Today's work: choose one path

All four paths end with a first scored allocation plus a note on the objective it sacrificed most. The path is about how far you push the design.

Learning Intention: I have produced a first allocation of my area across the 8 urban land uses, can read its objective scores, and can name one place where two objectives trade against each other.

First allocation and read the scores (default)

Allocate your grid across the 8 land uses so they cover the whole area. Record the seven objective scores. Write one sentence naming which objective your allocation sacrificed most, and why you accepted that.

Solo · Default

Push one objective to its limit (stretch)

Make a second allocation that pushes one objective (say flood resilience, or housing) as high as it will go. Record what breaks: which other scores collapsed to buy it? You are mapping your own Pareto frontier by hand, which is exactly the reasoning the trade-off analysis in Block 5 needs.

Solo · Stretch

Partner synthesis (choice)

Run the same allocation question past both partners and write the allocation that holds the most of each push without abandoning either. Document the disagreement and how you resolved it. This is the synthesis move the rubric rewards.

Solo · Choice

Peer-help: what a Pareto trade-off is (practice)

If "you cannot raise both at once" is fuzzy, come to the peer-help table. Work one two-objective example (housing versus green space) on paper with a partner until the frontier is clear, then return to your full grid.

Pairs · Practice

Open your Class Notebook and type today's entry header as Heading 2:

A: Apr 8 | U6 B4 | Multi-Objective Optimization: The Design MoveB: Apr 9 | U6 B4 | Multi-Objective Optimization: The Design Move

Show

What you're submitting today

A first scored allocation plus the sacrifice it carries. This becomes the spine of the trade-off analysis in Block 5.

First allocation + biggest sacrifice

Two things: (1) your first allocation across the 8 land uses with the seven objective scores recorded, and (2) one sentence naming the objective your allocation sacrificed most and why you accepted it. If you ran the partners, attach the AI Documentation, including the one thing a partner got wrong.

Submit via Google Classroom →

Rubric link: producing an allocation and reading its scores is the design move (Thinking and Transfer) on the Sustainable Neighborhood Plan. Naming the sacrifice sets up the Block 5 gate. See the rubric.

Reflect

One question before you leave

Three to five minutes. Your answer is saved to your reflection journal, where you can read back everything you have written this year.

The disagreement is the point.

Today's reflection is in Google Classroom, under Reflection Journal.

Link posted in August