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


Levels of Solutions

Levels of Solutions

| On 16, Dec 1997

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


Classification of engineering solutions by an inventive level is a handy tool and gives many practical benefits. Further, the classification is the basis of scientific methodology for innovative design. The five levels and their applications are described. The use of classification is illustrated by the example of an airbag.

The Five Levels

Five levels of solutions were first introduced by G. S. Altshuller. The levels have been then considered in many publications (Altshuller 1984, pp. 16-26; Altshuller 1986, pp. 44-49; Altshuller 1987, pp. 22-29; Altshuller et al. 1989, pp. 14-17; Ivanov 1987, pp. 22-25; Petrovich-Tsurikov 1986, pp.70-72; Rantanen 1985, pp. 28-30; Salamatov 1990, pp. 82-88, Savransky 1996).

We can describe the five levels as follows:

1 The Standard Solution. Typical trade-off, quantitative changes without new quality. Choosing between a few obvious variants. Knowledge within one profession is used. Altshuller studied patents from 14 different classes, published 1965-1969, and found that 32 % of patented solutions represent the first level (Altshuller 1987, pp. 22-29).

Examples: Thick (but not too thick) insulation to decrease heat losses. More heavy trucks to improve cost-efficiency of transport. Better resolution of display (say, 800×600 instead of 640×480).

2 Change of a System.The object is changed qualitatively, but not substantially. Choosing between tens of possible variants. Knowledge within one industry is used. About 45 per cent of patented solutions are at Level 2.

Examples: A fire extinguisher added to a welding device. The hollow handle of an ax (giving less weight and a more optimal center of gravity). A map turned upside down (comfortable when you are driving from North to South).

3 Solution Across Industries.The object is radically changed. Choosing between hundreds of possible variants. Knowledge outside the specific industry is used. About 19 % of solutions are in this category.

Examples: Radial tire. Ball point pen. Mountain bike. Mouse in a computer. Model-T Ford.

4 Solution Across Sciences.The new object is created. Choosing between thousands or tens of thousands of possible variants. New scientific knowledge, rather than technological information, is used. About 4 % of solutions are at Level 4.

Examples: Use of memory metal in a coupling. Internal combustion engine. Integrated circuit. Personal computer. Pneumatic tire. Virtual reality. Mining metals by using bacteria. Pilkington process in glass manufacturing. Use of electrohydraulic effect.

5 Discovery.Solution based on a scientific discovery. Choosing between hundreds of thousands or millions of possible variants. New discoveries are made first and then applied. About 0.3 % of solutions are Level 5.

Examples: Aeroplane. Transistor. Computer. Photography. Penicillin. Bicycle. Memory metals. Polymers. Steam engine and thermodynamics.

Table shows the level of invention, the character of change, the number of variants, the character of knowledge and the distribution of inventions by level:

The level

The change in the system

The number of variants

The knowledge used

The share of all inventions

1 The standard solution

trade-off, quantitative changes

a few

one profession

32 %

2 Change of a system




one industry

45 %

3 Solution across industries





many industries

19 %

4 Solution across sciences

new system created

thousands, tens of thousands

many sciences

4 %

5 Discovery

new discovery

hundreds of thousands, millions

new knowledge created

0.3 %

Table: The levels of inventions

The evaluation of solutions by level is, obviously, subjective to some extent. For example, some people think that the integrated circuit is the most important invention in the 20th century, that is, more important than a transistor. However, when a great number of solutions are classified, a small group of best inventions can be separated rather unanimously. All agree that the transistor and the integrated circuit both are very high level inventions. Subjectiveness is not a big problem. There are ways to get consensus – for example Delphi technique. The only condition is that all evaluators should know the classification criteria of TRIZ.

The criteria of inventiveness are, in addition, changing in time. The idea to use memory metal in couplings was an invention on the fourth level in the beginning of 1960s. Today the idea to introduce memory metals, which already are rather well-known, into some new branch may be an invention on the third level.

One open question is: does the distribution of inventions by level change with time? Altshuller studied in 1982 patents (published in the former USSR), from three patent classes. The criteria were the same as in the study of patents from years 1965-1969. The results: first level – 39 %, second level – 55 %, and third level – 6 %. Inventions of fourth and fifth level weren´t found at all (Altshuller 1986, pp. 51-52). The change, perhaps, can be explained by the statistical variation between samples. Or maybe the distribution really changes. More research is needed to answer to this question.

The Use of the Classification

Evaluation of solutions. Classify solutions, products and ideas by an inventive level.

Improving products. Knowing the current state of products and designs we can build more conscious strategy for product improvement. There are usually much designs on the first level. Often they are profitable today, but cannot maintain the competitiveness in long run. Maybe we should to increase the share of second, third and fourth level innovations?

Implementation of TRIZ. TRIZ and tools of TRIZ are based on the selection and study of high level inventions. The evaluation and classification of solutions make it easier to learn and use innovative technology of design. You will understand where the trends, principles, effects and predictions come from, and how to use them most effectively. Different tools work on different problems. Principles produce second level solutions easily. To get solutions of third and fourth level we need effects and inventive standards or predictions, too.

Improving core competence. Don´t improve only products. Improve the capability to develop better products. It is interesting to locate the levels of well-known design technologies. Narrow professional knowledge and CAD mainly help to find optimal quantitative parameters. That is, they produce first level solutions. So, TRIZ and Computer Aided Innovation, CAI, is really needed for generating solutions of high levels.

Knowledge management. The classification helps to detect innovative ideas in the first beginning and select valuable information. We know that high level innovations use scientific effects outside the field where they are developed. A new type of ship´s propeller, invented by analyzing a fish´s tail, may be more interesting than small improvements in a traditional screw propeller.

Airbag as an Example

Backround information of airbag problems in cars you can find in the papers of Kowalick (“No-compromise” Design…, and Domb (Contradictions: Airbag…, (See The TRIZ Journal Archive for these articles, which originally appeared in April and July, 1997). The problems and ideas concerning an airbag, and the supersystem “around” the bag, can be beautifully described by the five levels:

1 The Standard Solution. Quantitative features of an airbag are sought. The deployment threshold (the speed of the car that is required for the airbag to fire) is made lower or higher. The inflation speed of the bag is made lower or higher. The result has been an unbearable trade-off between lives saved by airbags, and fatalities caused by the airbag itself. So called smart airbag with different sensors and a computer is – at best – the compromise between device complexity and safety. The supersystem: speed limits. Limiting the use of private cars. – A supersystem is a system that includes the engineering system as one of its components.

2 Change of a System. The airbag is changed qualitatively. For example, segmentation of an airbag, or local quality. There are ideas to use multiple crushable layers, which cushion the blow from the bag itself, but don´t prevent the original purpose (to protect from striking the vehicle interior). See “E. Domb. Contradictions: Airbag Applications” ( In the supersystem: improved, more safe dashboard and steering wheel.

3 Solution Across Industries. Use of materials and technologies from other industries. For example ferromagnetic and electric fluids. “Mechatronic” airbag. Components of the supersystem: principally new safety belts, seats, the improved body of a car, etc.

4 Solution Across Sciences. Use of new, “smart” materials in airbags. The supersystem: The whole transport system is changed so that collisions are excluded, without limiting individual transport.Warning system for preventing collisions. Navigation systems for safer routes.

5 Discovery. ??? An airbag based on nanotechnology? Maybe someone discovers or invents a smart substance that itself (without sensors, computers and actuators) detects the occupant and properly regulates itself. The supersystem: New physical principles of transport? Totally automatized, maximally individual transport??

The classification helps to state the problem correctly. Discussion of airbags seems to be concentrated on to the problems and ideas of first and second level. But it is possible that the problem cannot be solved on lowest levels. We should consider the solutions on the third and fourth level, too. Further, propably there are not only one single solution, but the complex of solutions on different levels. So, look at many levels.

So Simple, So Genial

Think: classification by level is very common procedure. In schools, in universities, in sport, in music etc. the performance and results are evaluated by level. There are criteria for quality awards. Hotels and restaurants are ranked: from one to five stars. Mushrooms get stars by gastronomic value. And so on.

Altshuller transferred the idea of ranking, old and trivial on many other fields, into the world of inventions, ideas and innovations. It was the beginning of TRIZ and, later, Computer Aided Innovation.


G. Altshuller. Creativity as an Exact Science. NY. Gordon & Breach Science Publishers, 1984

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

G. Altshuller. Derzkie formuly tvorchestva. In: Derzkie formuly tvorchestva. Petrozavodsk. Karelija, 1987 (in Russian)

G. Altshuller et al. Poisk novyh idei: ot ozarenija k tehnologii. Kishinev. Karta Moldovenske, 1989 (in Russian)

G. Ivanov. …I nachinaite izobretatj! Vostochno-Sibirskoe knizhnoe izdateljsvo. Irkutsk 1987 (in Russian)

N. Petrovich, B. Tsurikov. Putj k izobreteniju. Moskva, Molodaja gvardija, 1986 (in Russian)

K. Rantanen. Teknisen luovuuden kehittaminen, Helsinki, Orienta-Konsultit, 1985 (Development of Engineering Creativity, in Finnish)

J. Salamatov. Kak statj izobretatelem. Moskva. Prosveshtshenie, 1990 (in Russian)

S. Savransky. The Methodology of Inventive Problem Solving, 1996, http://www.jps/triz/Tech1.Rev.htm