Class 12 is the final stage. Students carry out real mathematical research and modelling: decision science, simulation, mathematical economics, engineering applications and the mathematics behind artificial intelligence. They leave with a portfolio that shows how they think, not only what they remember.
1Mathematical domains this year
- Real-world calculus: differential equations in growth, cooling and motion (introductory)
- Optimisation and decision science: linear programming and decision trees
- Mathematical simulation: Monte Carlo methods and agent models
- Statistical research: hypothesis testing and confidence
- Data modelling with real Bangladeshi data
- Mathematical economics: supply, demand and elasticity
- Engineering applications: structures, signals and energy
- Computational mathematics and numerical methods
- AI and mathematical thinking: vectors, probability and learning from data
2Learning objectives
- Carry out a complete mathematical research project.
- Use simulation and numerical methods when exact solutions are hard.
- Apply optimisation to real decisions.
- Understand the mathematical ideas behind modern AI.
- Communicate mathematics clearly to experts and non-experts.
3Expected learning outcomes
By the end of Class 12, a student can:
- write a research report with a clear question, method, analysis, limitations and references;
- estimate a probability or area with a Monte Carlo simulation;
- set up and solve a small linear programming problem for a real decision;
- explain in simple words how a computer can 'learn' a straight line from data;
- present and defend their work in front of an audience.
4Mathematical thinking skills
- Research thinking: from question to defensible conclusion
- Computational thinking at scale
- Decision science: trade-offs and constraints
- Critical thinking about models and AI
- Mathematical communication
5Creative learning activities
Our Class 12 story world: Dreamer Future Lab
In Class 12, the final year of the programme, the students become the founding team of the Dreamer Future Lab, a pretend lab that asks one big question: how can mathematics help build a better Bangladesh? Through the year they play strategy games, run a classroom market, build a straw bridge and a solar cooker, simulate traffic and insurance, teach a "machine" to learn, design a flood warning rule and a scam filter, and finally present their own research at the Future Lab Expo. The Lab has five desks that match the five project categories: the Strategy Arena (play), the Economy Desk (daily life), the Engineering Works (making), the Simulation Centre (experiments) and the Future Council (problem solving). Each student keeps a research journal and earns a desk badge for each desk. The year closes with a graduation letter to their future self.
Joy rituals in every session
- Joy check (continued from Class 3): thumbs for enjoyment and confidence, noted privately, never marked.
- Mistake of the week: one interesting mistake goes on the "Mistakes that Helped Us" wall with what it taught.
- Model of the day (5 minutes): one real situation (a queue, a price, a rumour, a rainstorm) and the question "what would a simple model of this look like?" Students name the variables and one assumption.
- Claim check: one number or claim from the news or an advertisement. The class asks: what is the source, what is being compared, and could it be chance?
- One-minute researcher: each project ends with every team stating its finding in one sentence, how confident it is, and one thing the model leaves out.
Creative activities through the year
- Designing the Future Lab logo and a badge for each desk.
- A "Mathematics for a better Bangladesh" wall: ideas from the projects that could help a family, a school or a community.
- Mentoring: each Class 12 team runs one game or experiment for a younger class (for example Class 6 or 7).
- A Future Lab Open Day at the end of each term, where families play the games and try the simulations.
Note: Class 12 students prepare for the Higher Secondary Certificate examinations. The programme is enrichment and should be lighter or paused in the busiest exam months; nothing here is a test. Many projects are short and can be used as a joyful break from revision.
6Practical applications
- Understanding prices, taxes, inflation and the real value of savings
- Choosing mobile packages, insurance and other everyday products wisely
- Designing strong, light structures and using solar energy
- Understanding how computers learn from data and why AI can be wrong
- Recognising scam messages and misleading claims
- Planning production and resources for small businesses
- Designing early warnings and fair decision rules for communities
7Real-life problem solving
- How many cakes and biscuits should a small bakery make to earn the most? (21 The Best Plan for a Small Bakery)
- When should a community raise a flood warning? (22 Raise the Alarm?)
- Can a few words reveal a scam message? (23 Spot the Scam)
- Which voting method is fairest for a class decision? (24 Who Really Won?)
- Is insurance worth buying if it costs more than the expected loss? (10 Why Does Insurance Work?)
- Why do traffic jams appear with no accident? (20 Jams from Nowhere)
Strategies taught explicitly this year: write the constraints first, check the corners, compare with chance before believing a result, simulate when exact maths is hard, separate training data from test data, ask which mistake is worse, and state what the model leaves out.
8Mathematical games
- Trust or Betray? (game theory and repeated games)
- Rain of Rice (estimating π with Monte Carlo)
- Take the Money or Open the Box? (decision trees and expected value)
- Can You Really Predict the Coin? (hypothesis testing)
- Teach the Machine (how a computer learns a line)
- Model of the day warm-up
9Experiments
- How fast does tea cool? (16 Newton's Law of Cooling: a first differential equation)
- How many packets to collect all six stickers? (17 Collect All Six!: simulation and expected value)
- Is one paper plane really better? (18 A Fair Statistical Test: permutation tests)
- What sets the rhythm of a pendulum? (19 The Pendulum Clock: a square-root law)
- Why do jams appear from nowhere? (20 A Traffic Simulation: agent models)
Every experiment follows the same cycle, printed on the class wall: Wonder → Predict → Measure → Record → Model → Check the error → Wonder again. New this year: every report must say how the result could be wrong (chance, bias, a missing variable) and what further test would help.
10Construction activities
- A straw truss bridge tested to failure (11)
- A one-string instrument (monochord) with a marked scale (12)
- A parabolic foil solar cooker (13)
- A spreadsheet root-finding machine (14)
- A matrix flip-book animation (15)
- A seconds pendulum (19)
11Individual and group projects
| Individual projects | Group projects | Either |
|---|---|---|
| 7 Should the Tea Stall Raise Its Price?, 8 What Will Tk 1,000 Buy?, 14 Square Root Machine | 1 Trust or Betray?, 3 Take the Money or Open the Box?, 5 Teach the Machine, 6 Market Day, 10 Why Does Insurance Work?, 11 The Straw Bridge Challenge, 13 Cooking with the Sun, 16 How Fast Does Tea Cool?, 18 Which Paper Plane Flies Further?, 19 The Pendulum Clock, 20 Jams from Nowhere, 21 The Best Plan for a Small Bakery, 22 Raise the Alarm?, 23 Spot the Scam, 24 Who Really Won?, 25 Future Lab Expo (individual or small team) | 2 Rain of Rice, 4 Can You Really Predict the Coin?, 9 Which Data Package?, 12 Music You Can Graph, 15 The Matrix Animator, 17 Collect All Six! |
In group projects every student has a lab role that rotates: project lead, data manager, modeller, sceptic and presenter. New this year: the sceptic must find the strongest reason the team's conclusion could be wrong, and the team must answer it in the report.
12Innovation challenges
- Business Mission: find the best production plan for a small business with linear programming (21).
- Safety Mission: design and evaluate a flood early-warning rule (22).
- Digital Safety Mission: build and test a simple scam-message classifier (23).
- Fairness Mission: compare voting methods and write a class constitution (24).
- Research Mission: complete and present an independent mathematical research project (25).
- Future Lab Expo (December): students present their research to families, teachers, younger students and invited guests, then celebrate the end of the programme and write a letter to their future selves.
13End-of-level portfolio
Each student completes the Mathematics Discovery Portfolio, now kept as a research journal. By the end of Class 12 it contains:
- At least 20 journal pages, each with: the question, the model or method, my data, my result, how it could be wrong, and what I would test next.
- The desk badges earned in each of the five desks.
- Physical makes (or photos): straw bridge, monochord, solar cooker, flip-book, pendulum.
- A "Models I can build" page: one differential equation, one simulation, one optimisation and one statistical test, each with a worked example.
- A "Claims I checked" page: three claims from news or advertising and what the mathematics showed.
- The Future Lab research report and Expo poster (project 25).
- A 'Strategies I can use' page, gathered across the whole programme.
- The Joy and confidence tracker, now covering all the years.
- A graduation letter: 'Dear future me'.
The programme ends with the Future Lab Expo and a celebration for families. No marks, no ranking: the portfolio shows how each student thinks, not only what they remember.
1412-month learning calendar
One weekly 60-minute session; about two projects a month. Months follow the Bangladeshi school year (January to December). Projects are numbered as on this page. Higher Secondary examination dates vary from year to year, so teachers should move or pause months around the exam period; the short games (1–5, 17) work well as one-off sessions between revision.
| Month | Future Lab theme | Projects | Portfolio evidence |
|---|---|---|---|
| January | Opening the Lab: strategy and chance | 1 Trust or Betray?, 2 Rain of Rice | Tournament table, π estimate graph |
| February | Language Day: decisions and evidence | 3 Take the Money or Open the Box?, 4 Can You Really Predict the Coin? | Decision trees, binomial table |
| March | Economy Desk: markets and prices | 6 Market Day, 7 Should the Tea Stall Raise Its Price? | Supply-demand graph, elasticity note |
| April | Pohela Boishakh: money over time | 8 What Will Tk 1,000 Buy?, 9 Which Data Package? | Family basket index, package decision rule |
| May | Simulation Centre: rates and randomness | 16 How Fast Does Tea Cool?, 17 Collect All Six! | ln graph and k, sticker simulation |
| June | Monsoon: risk and warning | 10 Why Does Insurance Work?, 22 Raise the Alarm? | Insurance club results, confusion tables |
| July | Engineering Works: structures and sound | 11 The Straw Bridge Challenge, 12 Music You Can Graph | Bridge test report, monochord |
| August | Engineering Works: sun and numbers | 13 Cooking with the Sun, 14 Square Root Machine | Solar cooker data, spreadsheet machine |
| September | Machines that learn | 5 Teach the Machine, 15 The Matrix Animator, 23 Spot the Scam | Loss graph, flip-book, classifier results |
| October | Fair tests and fair systems | 18 Which Paper Plane Flies Further?, 19 The Pendulum Clock, 20 Jams from Nowhere | Permutation test, g estimate, flow graph |
| November | Future Council: decisions for communities | 21 The Best Plan for a Small Bakery, 24 Who Really Won?; 25 continues | LP graph, class constitution, research draft |
| December | Future Lab Expo (Victory Day) | 25 Future Lab Expo (begun in October) | Research report, poster, 'Dear future me' letter |
Weekly option (60 minutes): as above. Fortnightly option (120 minutes every two weeks): one project per session; use the multi-week projects as single long sessions with home tasks. Holiday camp option: five days of 3 hours cover about 10 projects (one Future Lab desk per day).
25 projects in five categories
Mathematics Through Play5 projects
Trust or Betray? A Game Theory Tournament
Pairs of students play a simple two-choice game many times: each secretly chooses Cooperate or Defect, and points come from a practice payoff table. Defecting looks clever in a single round, but over many rounds friendly, forgiving strategies such as "copy what the other player did last time" often score best. The class runs a round-robin tournament of written strategies and discovers the mathematics of trust, the central puzzle of game theory.
2Rain of Rice: Estimating π with Monte Carlo
A quarter circle is drawn inside a square. Students scatter grains of rice (or drop random points with a spreadsheet) and count how many land inside the quarter circle. Because the fraction inside should be close to π ÷ 4, the count gives an estimate of π. Teams pool their results, watch the estimate improve as the number of points grows, and learn Monte Carlo simulation, a method used everywhere from weather forecasting to finance.
3Take the Money or Open the Box? Decision Trees
A classroom game show offers a safe prize or a gamble with boxes of different practice values. Students draw decision trees, calculate expected values, and compare the "best on average" choice with what people actually choose. They then build a two-stage tree with a second chance, and discover why risk, not only averages, matters in real decisions about farming, business and health.
4Can You Really Predict the Coin? Testing a Claim
A student claims to have a special talent for predicting coin tosses. How many correct guesses out of ten would convince the class? Students calculate how likely each score is by pure luck, set a fair rule before testing, run the test, and meet the key ideas of hypothesis testing: the null hypothesis, the p-value, significance and the two kinds of error. They then see why scientists insist on deciding the rule in advance.
5Teach the Machine: How a Computer Learns a Line
Students play the part of a learning machine. Given a few data points, a team starts with a guessed line, measures how wrong it is with a "loss" (the sum of squared errors), and then makes small adjustments to the slope and intercept, keeping only the changes that reduce the loss. Round by round the line improves. Students discover the core idea behind much of modern AI, gradient descent, and compare their learned line with the exact least-squares answer.
Mathematics in Daily Life5 projects
Market Day: Where Supply Meets Demand
The class runs a pretend market for a single product, such as pitha at a winter mela. Half the class are buyers with secret maximum prices and half are sellers with secret minimum prices. After several rounds of trading, prices settle close to one value. Students then draw demand and supply lines, find the equilibrium by solving simultaneous equations, and explore what happens with a tax, a subsidy or a price ceiling.
7Should the Tea Stall Raise Its Price? Elasticity
If a tea stall raises the price of a cup, it sells fewer cups, but does it earn more or less? Students meet price elasticity of demand, the percentage change in quantity divided by the percentage change in price. They calculate it for practice data, see when a price rise increases revenue and when it backfires, and then interview a local shopkeeper (with permission) about how they decide prices.
8What Will Tk 1,000 Buy in Five Years? Inflation and Real Value
Prices of everyday things rise over time, so the same note buys less. Students build a small "family basket" price index from prices they record at home or in the market, learn how inflation compounds like interest, calculate the real value of savings, and compare a savings rate with an inflation rate to see whether money is really growing.
9Which Data Package Is Really Cheapest? Piecewise Functions
Mobile operators offer many internet packages with different prices, sizes and validity periods. Students record a week of their own data use, write each package as a cost function of monthly use, draw the functions on one graph, find the break-even points and build a simple decision rule: "if you use less than this, choose that". They then check the advertising claims critically.
10Why Does Insurance Work? Sharing Risk with Probability
Students run a classroom "insurance club" for a practice risk, such as losing a phone or a crop being damaged by flood. Using dice to decide who suffers a loss each "year", they compare going alone with pooling money in a shared fund. They calculate expected loss, see why a premium must be a little above the expected loss, and discover how pooling many people makes a frightening risk manageable, the mathematics behind insurance and microinsurance.
Mathematics Through Making and Construction5 projects
The Straw Bridge Challenge: Forces in Triangles
Teams design and build a truss bridge from drinking straws (or paper rolls) and tape to span a 40 cm gap, and test how much load it holds. Before building, they use vectors and trigonometry to calculate the forces in the members of a simple triangular truss, find which parts are squeezed (compression) and which are stretched (tension), and use this to place their strongest members wisely. The winning bridge carries the most load per gram of material.
12Music You Can Graph: Sine Waves and a String Instrument
Students build a simple one-string instrument (a monochord) from a board, a wire or fishing line and two bridges, and discover that halving the string length raises the note by an octave. Using a free tone-generator and a sound-wave viewer app on a phone, they see that pure notes are sine waves y = A sin(2πft), measure frequency and amplitude, and explore how the mathematics of ratios builds musical scales.
13Cooking with the Sun: Building a Parabolic Reflector
A parabola has a special point, the focus, where all rays parallel to its axis meet after reflection. Teams use the equation y = x² ÷ (4f) to draw an accurate parabola, cut cardboard ribs, cover a trough or dish with aluminium foil, and test whether it can heat water. They measure the temperature rise, compare different focal lengths and connect the result to satellite dishes, torches and solar power plants.
14Build Your Own Square Root Machine: Newton's Method
How does a calculator find √2? Students build their own root-finding "machine" in a spreadsheet using Newton's method, which improves a guess by sliding down the tangent line. They watch the number of correct digits roughly double with each step, use the machine to solve equations that have no simple formula, and discover when the method fails.
15The Matrix Animator: How Computers Move Pictures
Every time a phone rotates a photo or a game turns a character, a computer multiplies coordinates by a matrix. Students draw a simple picture (a boat, a kite or a letter) as a set of points, then use 2 × 2 matrices to rotate, reflect, stretch and shear it, by hand and in a spreadsheet. They discover that the determinant tells how much the area changes, combine transformations by multiplying matrices, and make a short flip-book animation.
Mathematics Through Experiments and Discovery5 projects
How Fast Does Tea Cool? Newton's Law of Cooling
Teams record the temperature of a cup of hot tea (or hot water) every two minutes as it cools. They find that it cools fast at first and then slowly, test the model "the rate of cooling is proportional to the difference from room temperature", solve this simple differential equation to get an exponential, and use a logarithm graph to find the cooling constant. They then predict when the tea will reach a comfortable drinking temperature and test design changes such as a lid.
17Collect All Six! How Many Packets Will It Take?
Imagine a snack packet that contains one of six different cricket stickers, each equally likely. How many packets must you buy, on average, to collect all six? Students simulate the problem with a die, record how many rolls each full collection takes, draw the distribution, and then derive the exact expected value using a clever splitting argument. They discover why the last sticker is always the hardest to find.
18Which Paper Plane Flies Further? A Fair Statistical Test
Two plane designs, two sets of flights, and a lot of variation. Is one design really better, or did it just get lucky? Teams fly each design many times in a fair, randomised way, summarise the data with means and standard deviations, and use a permutation test (shuffling the labels many times) to find how often a difference as big as theirs would happen by chance. They learn how real experiments decide whether a difference is meaningful.
19The Pendulum Clock: Discovering a Square-Root Law
Teams make pendulums from string and a small weight, and time 10 swings for different string lengths, masses and starting angles. They discover that only the length really matters, that four times the length gives twice the period, and test the model T = 2π√(L ÷ g) by plotting T² against L. From the gradient they estimate the strength of gravity, and then design a pendulum that ticks once per second.
20Jams from Nowhere: A Traffic Simulation
Why do traffic jams sometimes appear with no accident or signal? Students play the role of cars on a circular "road" of squares drawn on the floor or on paper, following a few simple rules each turn: speed up if there is space, slow down to avoid the car ahead, and sometimes brake randomly. Jams appear by themselves and even move backwards. Students then run the same model in a spreadsheet, measure how traffic flow depends on the number of cars, and discuss what could reduce jams.
Mathematics Through Innovation and Problem Solving5 projects
The Best Plan for a Small Bakery: Linear Programming
A small home bakery makes cakes and biscuits but has limited flour and oven time. How many of each should it make to earn the most? Students write the limits as inequalities, draw the feasible region, and find the best corner, which is linear programming, the mathematics of making the best decision under constraints. They then interview a local small business (or use a teacher-provided case) and build an LP model for a real decision.
22Raise the Alarm? Designing a Flood Early Warning Rule
An early warning must be raised in time, but too many false alarms make people stop listening. Teams use a practice record of daily water levels (or water levels they measure in a local drain, pond or school rain gauge) to design a warning rule based on the level and its rate of rise. They count hits, misses and false alarms in a confusion table, explore the trade-off as the threshold changes, and present a rule with a clear explanation of its strengths and limits.
23Spot the Scam: Building a Simple Message Classifier
Scam SMS and social media messages trick many people. Teams build a tiny "AI" that decides whether a message is a scam by counting suspicious words. Using a set of practice messages they write themselves (half normal, half scam-style), they calculate how often each word appears in each group, turn these into scores using probability, test the classifier on messages it has not seen, and discuss where it fails and why human judgement still matters.
24Who Really Won? The Mathematics of Fair Voting
The class chooses a picnic spot by ranking three options. Then they count the same ballots in three different ways, plurality, instant run-off and the Borda count, and are surprised to find three different winners. Students compare the methods using pairwise contests, learn about fairness criteria and Arrow's famous result that no ranked voting method can satisfy every fairness rule, and design a voting method for a real class decision.
25Future Lab Expo: My Mathematical Research Project
The final capstone of the whole programme. Each student (or a small team) completes an independent research project that uses mathematics to understand or improve something real: a model, a simulation, a data study, an optimisation or an AI-style classifier. They write a short research report with a clear question, method, analysis, limitations and references, and present it at the Future Lab Expo to families, teachers, younger students and, where possible, a guest from a university or local organisation.