How to use this page
As in Lesson 1, keep a notebook beside you. At each Pause and answer box, write your answer before you open the fold. The fold is one line of reasoning, not the official answer.
This lesson uses a few words from TRIZ, the theory of inventive problem solving: contradiction, Ideal Final Result, resources, separation, and the 40 inventive principles. They are explained in the TRIZ refresher; read it first or keep it open alongside.
What we assume you already know
- From Lesson 1: every job has a floor, set by conservation of mass and energy and by measured properties, that no design can go below. The gap above it is set by process, scale, and habit. The question to carry is: is this limit nature’s, or ours?
- Warming water takes about 4.18 kJ per kilogram per °C.
- Evaporating water takes heat. At room temperature, turning 1 kg of liquid water into vapour takes about 2.4 MJ, drawn from whatever the water was touching. That is why sweat cools you.
- The TRIZ basics in the refresher: a technical contradiction (improve one thing, another gets worse) and a physical contradiction (one thing must be both A and not-A).
- Machines can now produce fluent, plausible lists of ideas, quickly and cheaply, for anyone who asks.
If ideas are cheap, what is still worth being good at?
- When ideas cost almost nothing, what becomes scarce — and how do I decide which problem is worth solving?
- Can the way I state a problem hide its best answers?
- Once a problem is well framed, can inventing be done on purpose, instead of waiting for a flash?
- How do I judge a pile of ideas — mine, a colleague’s, or a machine’s?
- Write down one problem you would like solved, exactly as you would type it into a chatbot. Keep it: you will reframe it before the end.
Ice for Calcutta
In May 1833 the brig Tuscany sailed from Boston carrying about 180 tons of ice, cut in winter from New England ponds. Four months later it reached Calcutta with about 100 tons still frozen. The Boston merchant Frederic Tudor built the trade that followed, with ice houses in Calcutta, Bombay, and Madras. The Madras ice house, built in 1842, still stands near the Marina; it is now the Vivekanandar Illam.
The trade kept getting better at ice. Horse-drawn cutters developed by Tudor’s supplier Nathaniel Wyeth cut uniform blocks that packed tightly and melted less; sawdust insulated ships and stores. Its customers were mostly the wealthy: British officials and well-off Indian households.
By the 1880s the trade to India had collapsed. Ice made by machine, close to where it was used, had replaced it.
- Write the problem as an ice merchant would have stated it. Then write it as a customer in a hot Calcutta house would.
- Which of your two statements names a solution, and which names a function — something to be done, written as a verb and an object?
- Melting on the voyage: floor or gap? And what is the floor of the customer’s real problem?
- Who did the problem, as framed, serve? Who did it leave out?
After you have written your answers — compare your reasoning
The merchant’s problem: “How do we get more New England ice to Calcutta, with less melting?” The customer’s: “How do I keep this drink, this food cool?” The first has the solution built in — ice, from far away. The second names a function: remove heat from a drink.
Melting en route was a gap: better blocks, insulation, and faster ships all shrank it. But no improvement to shipping could compete with a machine that removes heat on the spot, because that machine needs no voyage at all. The customer’s floor is set by the drink: a kilogram must lose 4.18 kJ for each degree it cools. Where that heat goes, and what energy moves it, is left open. Ice was one answer; machines another; the clay pots households already used, a third (Observation 4).
Who matters too. The trade served those who could pay for imported ice. Asking “for whom?” shows who a solution leaves out — sometimes a much larger problem.
State a problem as a function — a verb and an object, such as “remove heat from water” — not as the solution you happen to use today, such as “ship more ice”. A frame that names a solution can only improve that solution. Add who it is for, what must happen, and why it matters to them.
The plane built to crash
In 1959 the British industrialist Henry Kremer offered a prize of £50,000 for the first human-powered aircraft to fly a figure-of-eight course around two pylons half a mile apart, clearing a 10-foot (3 m) height at the start and finish. Earlier attempts had largely adapted conventional aircraft designs, often in wood, and proved too heavy. The prize stood unclaimed for 18 years.
In 1976 Paul MacCready and Peter Lissaman began the Gossamer Condor: thin aluminium tubes covered in Mylar plastic, braced with steel wire, with leading edges of corrugated cardboard and foam. Its wings spanned 29 m; it weighed 32 kg. The Smithsonian’s National Air and Space Museum, which now displays it, notes its key advantage: after a crash it could be flying again within 24 hours, so the team could test and change it again and again.
On 23 August 1977 Bryan Allen, a championship cyclist, pedalled it around the course in a flight of 7 minutes 27.5 seconds, at about 16–18 km/h, producing about a third of a horsepower — roughly 250 watts.
- Find the floor. From Lesson 1, an ordinary adult can keep up about 75 W all day; a trained cyclist, a few hundred watts for some minutes. What does an engine of a few hundred watts demand of the aircraft?
- Compare two framings: “Build a better human-powered aeroplane” and “Find a design that works, by testing many designs quickly”. Where does each lead?
After you have written your answers — compare your reasoning
The engine is a person: a few hundred watts at best. At the low speeds of human flight, the power needed to stay up rises with weight and falls as the wing gets longer. Physics points to an aircraft that is very light, very long-winged, and slow — a jogging pace. Sturdy structure and compact wings are habits from aircraft with engines; for this job they were the gap.
“Build a better aeroplane” assumes you already know what the aeroplane should be. Nobody did. The real problem was learning fast enough, so the team framed it as a problem of cheap, quick experiments: a plane that is easy to crash and quick to repair. MacCready’s team treated human power as fixed — nature’s limit — and treated almost everything else as negotiable.
A good frame has three parts: the function (fly a figure-of-eight on human power), the floor (a few hundred watts), and what you still need to learn. When you do not yet know the answer, frame the problem so that you can learn quickly and cheaply.
A river open and shut
In 1953 a North Sea storm surge killed over 300 people in eastern England and pushed floodwater into London’s East End. London needed a wall against surges from the sea. But the Thames is a working river: ships must pass, and the river must flow out to sea.
The answer, operational from 1982 and officially opened in 1984, is the Thames Barrier near Woolwich: ten steel gates across 520 m of river. Each main gate is a curved segment that, when open, lies in a concrete sill on the river bed, so ships pass over it. To close, the gates rotate up until they block the river. The gates are hollow: they fill with water as they sink and empty as they rise. The engineer Charles Draper, who conceived the rotating gate, based it on the taps of his gas cooker.
The Environment Agency reports 221 flood-defence closures between 1982 and November 2025.
- State the technical contradiction. Then state the physical contradiction underneath it.
- Which of the four separations — time, space, condition, scale — did the design use?
- Find at least two of the 40 inventive principles in the gates. Which resource, already present, do the hollow gates use?
- Draper borrowed from a gas-cooker tap. In Lesson 1, analogy could suggest but not decide. Is that still true here?
After you have written your answers — compare your reasoning
Technical contradiction: raise protection, and navigation and river flow get worse. Physical contradiction: the barrier must be there (during a surge) and not there (the rest of the time).
Separation in time: closed only when a surge comes. Separation in space: when open, the gate is not removed, only moved — down into the river bed, out of the ships’ way.
Principles: Spheroidality (curvature) — a curved gate that rotates rather than lifts; Dynamics — a structure that changes position to suit the moment; Another dimension — the gate moves out of the horizontal plane ships use. The hollow gates use the river itself as ballast: a resource already there. Your labels may differ; the directions matter more than the numbers.
Draper’s tap is an analogy from an unrelated field — the cross-field transfer Altshuller found in patents. It suggested a shape; engineering, with models built and tested before construction, decided it.
Turn a trade-off into a physical contradiction — “this one thing must be A and not-A” — then ask when, where, under what condition, and at what scale it must be each. Separation gave London both protection and a working river — not a compromise.
The pot that cools itself
Across India, drinking water has long been kept in unglazed clay pots: the matka, the ghara, the surahi. Water seeps through the pores and evaporates from the outer surface, taking heat with it.
In 1939 three physicists in India — H. L. Gupta, A. C. Jain, and N. N. Khanna, working on a problem suggested by D. S. Kothari — measured earthen jugs and compared them with theory. The water settled close to the wet-bulb temperature: the reading of a thermometer wrapped in a wet cloth. That is the lowest temperature that evaporation into the surrounding air can reach. It depends on humidity. In hot, dry air the wet-bulb temperature is far below the air temperature; in humid air it is close to it.
Two days, worked out with a standard formula: air at 40 °C and 20% humidity has a wet-bulb temperature of about 23 °C. Air at 32 °C and 80% humidity, a monsoon afternoon, has one of about 29 °C.
- Write an Ideal Final Result for cool drinking water. How close does the matka come? List the resources it uses.
- Estimate: to cool 10 kg of water by 8 °C, how much heat must leave it? How much water must evaporate to carry that heat away?
- Which of the 40 principles does the pot use?
- On a humid monsoon day the pot barely cools. Is that a gap or a floor? Would a better pot fix it?
After you have written your answers — compare your reasoning
IFR: “The water cools itself, with no device and no bought energy.” The matka comes close. Its resources: the water itself, the dryness of the air, and the heat evaporation carries away.
Heat: 10 × 4.18 × 8 ≈ 334 kJ. Water evaporated: 334 kJ ÷ 2,400 kJ per kg ≈ 0.14 kg — a small glassful, to cool a bucketful by eight degrees.
Principles: Porous materials; arguably Self-service.
The monsoon limit is a floor. On the dry 40 °C day the pot can at best approach 23 °C, a fall of about 17 degrees; on the humid day, only 29 °C, about 3 degrees. A more porous pot or a breeze gets closer to the wet-bulb temperature faster — that is the gap — but evaporation into that air cannot go below it. To beat it you must change the job: pay energy to a refrigerator, or dry the air first.
The Ideal Final Result and the resources question lead to solutions with almost nothing added. But every such solution still sits on a floor — here, the wet-bulb temperature. Know the floor before you fall in love with the idea.
Occam: judging a pile of ideas
Suppose you ask a machine for “ways to keep drinking water cool for a family in a hot town, in a place with frequent power cuts”. In seconds it offers these five, among others:
- A. A solar-powered refrigerator with a battery.
- B. A matka in the shade, wrapped in a wet cloth, where a breeze reaches it.
- C. Daily ice delivered from the nearest town.
- D. A “smart clay pot” that cools water by evaporation alone to 10 °C on any day, including monsoon afternoons.
- E. Two pots: a matka for dry months, plus a small refrigerator run when power is available in humid months.
- Check each idea against the floor. Does any of them claim to beat it?
- For the rest, count what each adds: parts, energy, money, upkeep, people, ways to fail.
- Which would you choose — and what did you need to know about the family, the town, and the season to choose?
D fails first, on physics. Evaporation alone cannot take water below the wet-bulb temperature, which on a monsoon afternoon may be near 29 °C. A fluent description does not change that.
Of the rest, prefer the idea that delivers the function with the fewest added parts and assumptions — TRIZ’s ideality and Occam’s razor agree. B adds almost nothing but fails in humid weather. A works in all weather but adds cost and upkeep. C repeats the ice trade. E is a separation in time: the pot when the air is dry, bought energy only when the floor demands it.
To decide, you needed the climate, the budget, who maintains it, how often power fails. Those come from the frame, not the list. A machine can produce the list; judging it takes the floor and the people.
A working model: frame, then invent
Before opening ours, write your own version of the steps in five or six lines.
- What does a good frame contain?
- In what order would you use the TRIZ tools, and why?
- Where in the process does the floor from Lesson 1 belong — and how many times?
After you have written your summary — compare it with ours
| Step | What it means | The question to ask |
|---|---|---|
| Choose | Decide which problem deserves your effort. | “Who suffers from this, how much — and is it mine to solve?” |
| Function | State what must be done as a verb and an object, not today’s solution. | “What must happen, whatever does it?” |
| Who, what, why | Name the people, the situation, and what success means to them. | “For whom, where, and how will we know it worked?” |
| Floor | Use Lesson 1 to find what physics charges. | “Is the limit nature’s, or ours?” |
| Contradiction | Find the trade-off, then the A-and-not-A beneath it. | “What must be both, and when, where, and at what scale?” |
| Ideal & resources | Aim at the function with no added system; list what is already there. | “What if it did itself? What is free and nearby?” |
| Invent | Apply separation principles; try a few of the 40 principles. | “Which directions have worked for this kind of conflict?” |
| Judge | Check the floor again; prefer high ideality; run the cheapest test. | “Does it beat physics? What does it add? How would I find out fast?” |
The floor appears twice: to frame the problem and to judge the answers. The TRIZ tools sit in between; they widen your options inside the gap and never lower the floor. This, too, is a model, not a proven theorem: an order of questions to change when a problem shows you a better one.
Exercise: frame and invent
Return to the problem you typed at the start. Work through it on one page of your notebook.
- Function: rewrite your problem as a verb and an object. Strike out any solution hidden in it.
- Who, what, why: one line each.
- Floor: one rough sum, in the units physics uses. If you cannot find a physical floor, write down what limits the problem instead, and whether that limit is nature’s or ours.
- Contradiction: the trade-off, then the A-and-not-A.
- Ideal Final Result in one sentence, and a list of at least five resources already present.
- Separations: try all four. Write one idea for each, even a bad one.
- Principles: pick three from the list of 40 and force each onto your problem.
- Judge: check each idea against your floor, count what it adds, and name the cheapest test that could prove it wrong.
An example to start from, with no numbers filled in: “dry clothes indoors in the monsoon”. Function: remove water from fabric. Floor: weigh a load wet, then dry; every kilogram of water you evaporate costs about 2.4 MJ, whoever pays it. Squeezing water out as liquid does not pay that price — a resource worth noticing.
- Compare your new frame with the sentence you first typed. What did the old wording rule out?
- Now give your new frame to a machine, or to a friend, and ask for ideas. Do the answers improve? Which ones fail your floor check?
- Which step was hardest? Why do you think that is?
After you have done the exercise — compare your reasoning
Most first statements contain a solution: “an app for…”, “a cheaper machine that…”. Rewriting as a function opens answers the first wording ruled out, as “ship ice” ruled out “make cold on the spot”.
A sharp frame usually gets sharper ideas, from people and machines alike. You still judge the answers. The hardest steps are usually choosing and judging, not generating — the point of this lesson.
Pushing on the model
- A machine can be asked to run every step in the table, including the TRIZ ones. Which steps can you safely hand over, and which should stay with the person who will live with the result?
- Was Tudor wrong to ship ice in 1833?
- Sometimes no separation works, and every idea is a compromise. Has the method failed?
- Who decides which problem is “worth” solving? Can first principles answer that?
After you have written your answers — one way to refine the model
Generating can be shared; accountability cannot. A machine can list functions, contradictions, and principles faster than you. It cannot know what you have not told it about the people and the place, and it can produce a fluent idea that breaks a floor. Choosing the problem, measuring the floor in the real world, and judging acceptable risk stay with whoever answers for the result.
Frames expire. In 1833 there was no practical way to make ice by machine in Calcutta, so shipping it was a strong answer to “keep things cold”. When the means changed, the function stayed and the best solution moved. Re-frame when the means change.
A compromise can be honest. TRIZ is heuristic, not law; it suggests where to look and guarantees nothing (see the refresher’s honest limits). If no separation works, a well-chosen compromise is an answer, not a failure.
Worth is a question of values. First principles show what is possible, not whose problem matters most. That judgement is made with other people — which is why it stays human.
Why this matters now
When capable machines can produce ideas, designs, and plans on request, the bottleneck moves. It is no longer “can anyone think of something?” It is: is this the right problem, stated as a function, for the right people, with the floor in view? And afterwards: which of these answers survives physics, and which is worth building?
When wages stop being the centre of daily life, those questions become most of the work that is left: deciding what to build, for whom, and judging well when ideas arrive faster than anyone can read them.
Review & discussion questions
Answer these by explaining, in full sentences, as if teaching someone who missed the lesson. Try them alone first, then discuss with a friend or study group.
- Using the ice trade, explain the difference between stating a problem as a solution and stating it as a function. What did each frame make visible?
- Explain why the Gossamer Condor team framed their problem around learning quickly, and how the human power floor shaped the aircraft.
- State the Thames Barrier’s physical contradiction, and explain how separation in time and in space resolved it.
- Without notes, estimate how much water must evaporate to cool a bucket of water by a few degrees. Then explain why no clay pot can beat the wet-bulb temperature.
- Explain why idea D in the Occam section fails, and why choosing among the rest required knowing the frame.
- Which parts of solving a problem can be handed to a machine, and which should stay with people? Defend your answer with an example.
…you can rewrite a problem stated as a solution into a function with who, what, why, and a floor; turn its trade-off into a physical contradiction; generate options with the Ideal Final Result, resources, separations, and inventive principles; and judge any options — yours or a machine’s — against the floor.
Go further The TRIZ refresher lists all 40 inventive principles, the separation principles, and the honest limits of the method.
Where the facts come from. Rounded values are used throughout; every case is a documented historical or engineering record.
- Water, 4.18 kJ per kg per °C; latent heat of vaporisation ≈ 2.44 MJ/kg at 25 °C (rounded to 2.4): standard reference values, e.g. CRC Handbook of Chemistry and Physics; NIST / IAPWS steam tables.
- Ice trade — Tuscany, May–September 1833, about 180 tons shipped and 100 tons delivered; ice houses in Calcutta, Bombay, and Madras; Wyeth’s ice cutters; decline by the 1880s: D. G. Dickason, “The Nineteenth-Century Indo-American Ice Trade: An Hyperborean Epic”, Modern Asian Studies 25 (1), 1991, pp. 53–89; G. Weightman, The Frozen-Water Trade (2003). Madras ice house, 1842, now Vivekanandar Illam: Dickason (1991); historical exhibits at the Vivekanandar Illam (Ramakrishna Math, Chennai).
- Kremer prize (1959, £50,000, figure-of-eight around pylons half a mile apart, 10-foot height); Gossamer Condor materials, 29.25 m span, 31.75 kg, repairable within 24 hours; flight of 23 August 1977 by Bryan Allen, 7 min 27.5 s, 10–11 mph, about one-third horsepower: Smithsonian National Air and Space Museum, object record for the MacCready Gossamer Condor; Royal Aeronautical Society, Human Powered Aircraft Group.
- Thames Barrier — 1953 surge; ten gates across 520 m; rotating gates lying on the river bed when open; operational 1982, opened 1984; 221 closures to November 2025: UK Environment Agency, “The Thames Barrier” (GOV.UK). Hollow gates that fill and empty; Charles Draper’s gas-cooker-tap inspiration: Institution of Civil Engineers, “Thames Barrier”; Science Museum, London.
- Earthen jugs approach the wet-bulb temperature, the lowest reachable by free evaporation: H. L. Gupta, A. C. Jain and N. N. Khanna, “Evaporation from earthen jugs” (1939), Indian Association for the Cultivation of Science repository. Wet-bulb examples (about 23 °C at 40 °C and 20% humidity; about 29 °C at 32 °C and 80%): calculated with R. Stull, “Wet-Bulb Temperature from Relative Humidity and Air Temperature”, Journal of Applied Meteorology and Climatology 50, 2011; consistent with standard psychrometric charts.
- TRIZ terms and the 40 inventive principles: see the TRIZ refresher’s sources (Altshuller’s books and standard references).
Attribution. Lesson structure — Foundation, questions, observations and deductions, refined models — adapted from John S. Hutchinson, Concept Development Studies in Chemistry (Connexions / Rice University), licensed under Creative Commons Attribution 2.0 (CC BY 2.0). The topic, examples, and text of this lesson are original to learn.curiosta.com.