Image Image Image Image Image Image Image Image Image Image
Scroll to top


Polysystem Approach to TRIZ

Polysystem Approach to TRIZ

| On 05, Sep 1997

By Kalevi Rantanen
Brahenk. 9 E 18
phone/fax +358 2 251 1623


Traditionally in engineering design one system, or monosystem, is selected for further development. One system, one problem, one idea. In the paper will be presented the thesis that if we study simultaneously two systems, or bisystem, instead of a single one, the result improves. And if we analyse many systems, or polysystem, instead of bisystem, the result will be yet better.

“Traditional” TRIZ, too, has usually worked according to the line “one system – one problem – one idea”. But inside TRIZ has been developed some important new concepts already related to the polysystem: the trend mono-bi-poly (Altshuller), alternative system (Gerasimov and Litvin) and feature transfer (Gerasimov and Litvin). A contradiction appears: some tools have polysystem approach, many others continue to consider the monosystem.

The contradiction can be easily resolved, when the core concepts of TRIZ are looked from the polysystem point of view. The presentation of TRIZ in terms of polysystem and feature transfer permits to connect concepts and tools in a more simple, straightforward, and logical way than formerly. Despite the growth of the theory and methodology, the innovative technology of design can be studied and used more easily and effectively than ever before.


The first sources of inspiration have been Finnish engineers who have taken part in different seminars and courses in 1985-1997. Valuable meetings with Vladimir Gerasimov in 1996 and in the beginning of 1997 have helped me to clarify many concepts and ideas. Of course I carry the whole responsibility of any drawbacks in the paper.

Some necessary definitions

System: The common definition is that the system is a set of parts that is more than the sum of these parts. We’ll add the requirement that the system should resolve contradictions in the parts. Further, the system can be the set of all sorts of objects. It can be machinery, a process, a set of features, a combination of problems, ideas, marketing measures, and so on.

Bisystem. System evolved from two objects. An important case is the bisystem evolved from an original system and an alternative one.

Polysystem. System evolved from many objects.

Competing system. The system having the same function as an original system.

Original system. The system we want to improve.

Alternative system. The system which have pluses and minuses “opposite” to the properties of original system, coupled as follows: In the original system feature A improves, and feature B worsens, in the alternative system feature A worsens and feature B improves. The definition means that both systems in the pair are alternative to each other.

Polysystem of problems

I would like in the beginning to demonstrate the new concept and the new approach by one case. The bicycle is a very good example. All typical features and problems of design and the development of technology can be seen in the bicycle. Let’s list some actual problems in the most common, upright bicycle:

  • No weather protection
  • Limited speed
  • Limited amount of luggage
  • Not very safe in collisions
  • High aerodynamic resistance or drag
  • Etc.

The important and new feature in the list is that we try to collect all possible problems. Usually one or few bottleneck problems are selected. However, strange as it sounds, a set of problems can be solved more easily than a single “mono-problem”.

Polysystem of competing designs

The next step is to list most important systems competing with the common bike. The common bike is an original. The function of the bike is “to move a person and luggage”. The competing systems are all systems having the same function. Examples of competing systems:

  • Conventional, upright bicycle (original system)
  • Recumbent. A word recumbent refers to the seated position. The cyclist sits in the seat a little bit like in the car.
  • Hybrids: human powered, electrically assisted bikes
  • Electric scooter (“electric bicycle”)
  • Car
  • Trice
  • Wheelchairs
  • Spacecraft
  • Flying carpet
  • Etc.

The list has five important features:

  1. Systems have the same function
  2. List contains existing technologies or ideas
  3. List should contain all existing ideas
  4. Function is only limitation
  5. List already gives clues

First, we collect systems in the limits of function, not in the limits of the branch of industry. Functional approach helps to think out of the box. It also helps to activate existing knowledge.

Second, we list only existing technologies and ideas. We don’t try to guess new solutions. Facts should be collected before creating new concepts.

Third, all existing designs should be listed. No one shouldn’t be forgotten. Training experience shows that sometimes people hide a solution they think as good and then try to “fit” the methodology to the old idea. All old things should be pronounced, written down and listed.

Fourth, the function is the only limitation. Systems can be exotic (spacecraft), fantastic (flying carpet), or even crazy (perpetuum mobile). We are not selecting any of competing systems. We are interested in features. A crazy idea can contain some useful feature. Perpetuum mobile is not possible, but a very low-energy bicycle is.

Fifth, the list already speaks to the user. He/she can see immediately some new unusual problems and ideas. From any list of old systems or ideas we can always get new combinations. However, combinations should not be random sets of things, but pairs of alternative systems.

Pairs of alternative systems

From the list of competing systems we can select different pairs of alternative systems. Let’s look at some examples of pairs and properties. Only a few of possible combinations and only some of important features are named.

Upright bicycle VS. recumbent bicycle. Properties of an upright bicycle:

  • Plus: maneuverability, old and feasible technology
  • Minus: high aerodynamic resistance or drag

Properties of a recumbent bicycle:

  • Plus: low aerodynamic resistance or drag
  • Minus: less maneuverable than an upright

“Open” bicycle VS. faired bicycle
Properties of an open bicycle:

  • Plus: light
  • Minus: no weather protection

Properties of a faired bicycle:

  • Plus: weather protection
  • Minus: heavy

Bicycle powered by a cyclist only, VS. electrically assisted bicycle
Properties of an human powered bicycle:

  • Plus: light
  • Minus: little power

Properties of an electrically assisted bicycle:

  • Plus: powerful machine
  • Minus: heavy

Bicycle VS. Car
Properties of an bicycle:

  • Plus: high energy efficiency
  • Minus: low speed and low carrying capacity

Properties of a car:

  • Plus: high speed and high carrying capacity
  • Minus: low energy efficiency

Bike VS. Tricycle
Properties of an upright bicycle:

  • Plus: maneuverability
  • Minus: low carrying capacity

Properties of a three-wheel bike of tricycle:

  • Plus: more luggage
  • Minus: less maneuverable than a two-wheel bike

Pairwise comparison of competing systems, with opposite pluses and minuses, allows one to get quickly interesting and non-trivial problem statements.

The ideal final result and feature transfer

We describe ideal final result first simply picking up the pluses:

The ideal bicycle.
Properties of an upright bicycle:

  • Plus: maneuverability, old and feasible technology

Properties of a recumbent:

  • Plus: low aerodynamic resistance or drag

The ideal bicycle
Properties of an upright bicycle:

  • Plus: light

Properties of a faired bicycle:

  • Plus: weather protection

The ideal bicycle
Properties of an human powered bicycle:

  • Plus: light

Properties of an electrically assisted bicycle:

  • Plus: efficiency

The ideal bicycle Properties of an bicycle:

  • Plus: high energy efficiency

Properties of a car:

  • Plus: high speed and high carrying capacity

The ideal bicycle
Properties of an upright bicycle:

  • Plus: maneuverability

Properties of a three-wheel bike of tricycle:

  • Plus: more luggage

We will combine the pairs of systems so that a good feature is transferred from the one system to another. For the combination many different TRIZ tools are used.

For example: we want to combine good features of the human powered bicycle and electrically assisted one. Let’s try to formulate the physical contradiction. A bicycle should be LIGHT, for easy riding, and a bicycle should be HEAVY, for… For what? When do we need a heavy bike? Obviously when we want use the bike for fitness purposes. The ideal final result: The bike should be heavy enough for exercise, and light enough for avoiding too big loads harmful for the cyclist. An electrical motor helps to maintain an optimal load, depending, for example, on heart rate. The essential point is to coordinate the “rhythms” of the machine and the cyclist. We can refer also to the principle: “Turn lemons into lemonade”. If we have heavy batteries, why not to get some benefit from them? When pedalling is too light, extra weight helps.

The first pair in the list is an upright bike and recumbent bike. One physical contradiction: The bike should be low, to decrease resistance, and the bike should be high, to be visible on the road. One suggestion is to use the copying principle: to add for example a flag, which is visible but doesn’t cause aerodynamic resistance.

Or the last contradiction: two wheels – three wheels. Maybe use the dynamization trend? If all three wheels will draw and steer, we can get the maneuverability of a two-wheel bike in the three-wheel bike.

There is no strict rule saying, how to find the proper physical contradiction, trend or principle. And there should not be; the model should give clues, work as a map and a compass. It should show the direction and general “landscape”, and at the same time leave enough room for imagination.

The presentation of two systems with their opposite pluses and minuses can be called the “two pluses” formula. One can also make a table, called the “two pluses” matrix. The formula or matrix is simple, nearly childish, but it works! The user can quickly get some new viewpoints and ideas. He/she gets fast success, which is so important, especially in the beginning, when learning new skills. At the same time the formula prepares the problem for more exact analysis, for example by ARIZ.

Polysystem of solutions

We see that there are many systems, many features, many contradictions and many ideal final results. Actually we have always had the polysystem of problems and solutions. We consider the model of two systems and two features because the human problem solver cannot see much more at once. It’s easier, and takes less time to make many simple models than a single complex one.

So the future bike will be the combination of many features from existing technologies and designs:

  • maneuverable as an upright bicycle
  • convenient and energy-efficient as a recumbent
  • powerful as an electrical bicycle
  • stabile as an three-wheel bicycle
  • etc.

Not only the engineering system, but the design process, too, is a polysystem. The process is combined from many steps of improvement starting from the first concepts. Instead of “one step design” we need “many step design”. The concept should be criticized and improved, criticized and improved, many times.

Smashing psychological barriers

The early mountain bike of Charlie Kelly and Gary Fisher was an ugly duckling. The inventors, of course, saw the bright future. Mountain bike WILL give:

  • possibility of off-road riding
  • easy balance
  • certain stopping power
  • low gears
  • few jolts

Distinguished bike manufacturers practiced negative thinking. They saw many problems:

  • weight: too heavy
  • fat frame
  • fat and heavy tires
  • heavy hub brakes
  • off-road riding: who in the hell does want ride on mountains and in forests?

The suspicions were strong. The new concept turned the philosophy of bike design on its head:

  • beefed-up frame instead of slender one
  • low gears instead of high ones
  • fat tires instead of lean ones
  • heavy-duty brakes instead of nice, small and sophisticated brakes

Both sides were right their own way: the ugly duckling became a success, but only after numerous improvements:

  • less weight
  • better brakes
  • better tires
  • better gears

“Think positively”, we are taught. Actually we should think BOTH positively AND negatively, properly combine both thinking modes. So we need the bisystem of positive and negative thinking. By the way, there is a polysystem, too. De Bono suggests not two, but six thinking modes (thinking hats).

Bisystem, two pluses matrix, ideal final result and other models help to switch on, and switch off, negative and positive thinking, when needed. Models help to weaken the psychological barrier. The barrier is very strong. In everyday life we need a psychological defence mechanism. We need to think positively, at least of ourselves and of our own ideas. But just this defence mechanism disturbs, when we should improve the ideas. To improve the idea is necessary first to find drawbacks, but positive thinking requires: don’t seek drawbacks.

That’s why is necessary temporarily “switch on” negative thinking. My friend and colleague Vladimir Gerasimov conducted 1995-1996 some consultation projects for a U.S. company. One commercially very successful project was carried out in less than 3 months. During these three months the conceptualization cycle was repeated about 40 (forty) times.

“But yet we are doing this all the time!”

The most important objection to the polysystem, many step approach is: “Here is nothing new. This is trivial. We are designing complex systems all the time.” OK. Let’s consider successes and failures in the history of technology to see, whether Polysystem are trivial. Here are two points.

First, the polysystem is the system which solves contradictions, not every device or machinery containing many parts.

Famous successes in the history of a bicycle are, for example, a safety bicycle, introduced in 1884-1890, and a mountain bike, which appeared in 1970s. Famous failures in the bicycle design are, for example, Ikera in 1982 and Sinclair´s C5 in 1985.

The safety bicycle is the most common bike we are riding now. In the beginning it was called Safety, because of the low machine with a chain drive to the back wheel was safer than the high bicycle with a big wheel. The safety bicycle combined beautifully already existing technologies. The most important were the following two:

  1. Chain transmission, which was already widely used in bicycles. The invention was to place chain transmission just from the pedals to the back wheel.
  2. Pneumatic tires, which were invented already in 1845 and used on heavy horse-drawn vehicles. The invention of Dunlop in 1888 actually wasn’t the air tire itself, but the idea to use it in the bicycle.

No totally new components, and no totally new designs of PARTS were used in the safety bicycle. The idea was in the system as whole. The NEW, superior system was build from OLD sub-systems.

The mountain bike evolved as the combination of old and even bad sub-systems: beefed-up frame, low gears, fat tires, heavy-duty breaks. These components were not very good in city bikes, but worked well in mountain bicycles. A GOOD system was combined from BAD sub-systems. Of course old components were gradually improved. The BREAKTHROUGH was combined from many INCREMENTAL sub-steps.

Let’s now consider failures. In 1982 was introduced the bicycle Itera, made from plastic. Itera with plastic frame, forks, wheels and handlebars was light and strong and would not rust. But it didn’t sell.

In 1985 Sir Clive Sinclair tried to introduce a recumbent bicycle with an electric motor, called C5. The bike was environmentally friendly and cost-effective. In spite of many good features C5 became first laughing stock in London and then a commercial failure.

What is difference between success and failure? Successful engineering systems are developed according to the “contradiction formula”: old and new, bad and good, incremental improvements and a revolutionary breakthrough, etc. Unfortunate systems have a linear formula: new components – new quality, only incremental changes, or only one single breakthrough. Itera was based on one idea: use of plastic in a bicycle. New material – new quality. No contradictions. Actually the material became more important than customer needs. Sinclair´s C5 was an attempt to make a breakthrough in one step, without necessary cycles of design. The mountain bike is here a good comparison. Frame, tires, breaks, gears all were improved after the first concept.

A fresh success story in vehicle design is the Sojourner rover of NASA, landed in Mars in July 4,1997. It is no surprise that here, too, we can easily see the contradiction which was solved. Newsweek described the rover as follows: “More Than the Sum of Its Parts. Sojourner was fitted with many items that there purchased off the shelf…” (Newsweek, July 21, 1997, p. 45). There were standard power converters, modems, cameras and motors. This like solutions enabled the superior performance at low cost: 171 million dollars compared to the 3 billion budget (in today’s dollars) of Viking in 1976. At last, the work was done in short time. As whole: better, cheaper, faster.

Questions: Which is more trivial, typical and common in industry, successful “polysystem design”, or average trade-off? Do we really have nothing to learn from designs like safety bicycle and NASA`s Mars rover?

It is important, too, to pay attention to the second point: time. Some good polysystems have evolved in the history without conscious “polysystem design”. But they are too rare and the “wild” evolution takes too much time. Radial tires are now taken for granted in cars and many people don’t know, or have forgotten, that in past there were cross-ply tires. However, the breakthrough of radial tires took 55 years, from year 1914 to 1970s. Can we really afford such time losses inevitable in the work by trial and error?

From past to future

History is interesting when it is connected with future. Let’s consider yet two “real” problems, that is problems not yet solved: air bags and urban planning.

Very clear example is the air bag problem, considered earlier in TRIZ Journal, see papers:

  • J. Kowalick. “No-compromise” design solutions to the air bag fatalities problem, TRIZ Journal, April 1997
  • E. Domb. Contradictions: Air Bag Applications, TRIZ Journal, July 1997

Let’s imagine that the air bag problem itself is solved. We have an airbag which protects an average adult best possible way and never causes injuries or fatalities to small persons. We’ll have all advantages of the airbag, without any drawbacks. Can we be satisfied with this like solution? Obviously, not, because of there is, for example, a psychological problem. Kowalick refers to the study conducted by The Nation Center for Policy Analysis in the U.S. Some conclusions of the study: drivers tend to be more aggressive in vehicles equipped with air bags, and this aggressiveness offsets any safety gains. The best air bags and other technical improvements in the car cannot help much, if drivers continue to “compensate” improvements by increased speed and more hazardous driving habits.

We get a new contradiction: some drivers want to drive at high speed and hazardously, and they should drive slowly and carefully. To solve the contradiction, we can use separation in place and time, and taking out principle. One possible solution: to encourage automobile sport, to draw hazardous driving from public roads to more safe places.

There are “pure” engineering problems, too. One problem is that a steering column and pedals are not necessarily the best possible “interface”. Mercedes-Benz, for instance, is studying the car controlled by joysticks. Removing the steering column removes one of safety problems, too.

I am nearly sure, that the air bag problem will not be solved until a set of problems in the “environment” of the bag will get their solution, too.

Another example: urban planning. I would like to plan the city of future. We know well many problems of a modern city: congestion, slow traffic, air pollution, noise, accidents, aggressive people. We know old solutions well, too: overpasses, multilevel highways, parking facilities, traffic control, smaller cars, public transport. Of course there are plans of totally new cities, but it is difficult to reconstruct a whole city. Maybe try to find contradictions and solve them? To get the ideal city using in first order old “sub-systems”: common materials, technologies, plans and designs? Say, multiplying overpasses and multilevel streets, roads and highways. By “addition” of old constructions we can get a totally new quality: a two-level or multilevel city.

Getting customer oriented quality

World class quality requires that the product meets the needs of the customer. But how to find needs? When the safety bicycle was introduced in 1880s, people considered it ugly. But soon they fell love with it, and to our days most bicycles have been “safeties”. The need existed beneath surface.

When Itera bicycle was introduced in 1980s, people again said how ugly it was. But this time they didn’t begin to buy. Although market research had convinced that it should be more than 100 000 buyers in Sweden alone.

The verbatims of potential buyers cannot alone forecast customer demands. Customers often cannot know exactly what they’ll need in future. Polysystem approach help to see hidden, non-spoken needs.

Boosting of TRIZ

Many key concepts of TRIZ can be more easily understood and used, if we present them in terms of the bisystem and polysystem. The new presentation will complete old definitions and make them more exact. It will not cancel them.

Engineering contradiction is the contradiction in a bisystem. If we have an engineering contradiction, we always have an alternative system, too. And we have alternative engineering contradiction.

Physical contradiction. If we have the physical contradiction, we have a bisystem. Alternative systems have opposite physical properties.

Ideal Final Result. Ideal final result is a polysystem of features got as the result of successive combination of alternative systems.

ARIZ. The two pluses formula contains in a hidden form the key concepts of ARIZ. The formula helps to learn and use ARIZ.

Polysystem approach and computer aided innovation

Computer aided innovation (CAI) is a new concept developed as a result of progress in innovative software. When I speak of software I mean products of Invention Machine Corporation which I know, although I have heard there are other TRIZ based SW packages, too. More exactly, I´ll consider TechOptimizer Professional Edition.

Software Package TechOptimizer Professional Edition contains five modules: TechOptimizer, Effects, Principles, Prediction and Feature Transfer. The TechOptimizer module analyzes components, functions and links in the system and proposes ways to simplify the system. This module produces problem statements. The Effects module provides the user with access to more than 520 scientific effects. The Principles module solves engineering contradictions by introducing innovative principles (Altshuller´s matrix). The Prediction module helps define the future direction of innovation. The Feature Transfer module allows to transfer desirable features from one engineering system to another. More detailed information you can get from the web site of Invention Machine Corporation

The Feature Transfer module in TechOptimizer is a direct example of the polysystem approach… If you want to improve an object, you find a number of objects that perform the same function. Then you analyze these objects, investigate the best features of each one and try to transfer these features to one object.” (TechOptimizer, manual, p. 150)

In the function model and trimming stages a single engineering system is analyzed. One can say we have monosystem approach here. If we want to analyze the bicycle, we should select one of competing systems: common bike, recumbent, electrically assisted bike, faired bike, or some else. Feature transfer and trimming are complementary, and, obviously, give the best result when used together.

Prediction, Principles and Effects modules permit in short time finding many structural solutions or predictions, many principles and many effects. The software allows one to develop a polysystem of ideas. So the polysystem approach and TechOptimizer support each other.

At last, TechOptimizer as whole is a good polysystem, much more than the sum of five modules.

There is an interesting similarity between the polysystem approach and the IM modules Prediction, Principles and Effects. We have lists of competing systems and bisystems (pairs of alternative systems). IM contains lists of predictions, principles, functions and effects. It seems that people prefer lists and very simple tables or matrixes to long algorithms.

Trend mono-bi-poly

TechOptimizer gives in the Prediction module four examples of the trend mono-bi-poly: evolution of pencil, propeller, stitching head and light source. They are good examples of hard systems. I would like to widen the definition of the trend, and add “soft” systems, in first order following lines of evolution:

  • one problem, two problems, many problems
  • one idea, two ideas, many ideas
  • one step, two steps, many steps


Altshuller G. S. 1986. Naiti ideju (in Russian). Novosibirsk. Nauka

Litvin S. S., Gerassimov V. M: Development of alternative technical systems by incorporating them into supersystem. .Proc. ICED 91 Zurich, Vol. 1, 1991, pp 42-45

Software: TechOptimizer Professional Edition, Invention Machine Corporation,