Keywords: TRIZ, contradiction, problems, technical systems, elementary structures

Introduction

Usually the routine technical problems are solved by trading off one thing for something else, and the inventive problems are solved through resolution of the contradictions. TRIZ [1,2] states even that the most effective inventive solution of a problem is usually one that overcomes all contradictions. G.S. Altshuller [1] distinguished 3 types of contradictions named Administrative, Technical and Physical. The whole picture is more complicated and contradiction ontology can be built [3,4]. Nevertheless for many engineering problems technical and/or physical contradictions plays the critical role [1-4], hence we will discuss these types of contradictions in the problem structure in this paper.

Contradictions in the problems structure

The traditional contradictions for TRIZ [1,2] are the following:

1. The administrative contradictions have a provisional form:

Anything is required to make, to receive some result AND/OR to avoid the undesirable phenomenon, but it is not known how to achieve this result. It is reminiscent the inventive situation [1,2]. The administrative contradiction itself has not heuristic value, it does not show a direction of search of the answer.

2. The technical contradiction have the form:

An useful action causes simultaneously and harmful; the introduction (or amplification) of the useful action or the recession (or easing) of the harmful effect leads to a deterioration of some parts of a system or all system as a whole, e.g., an inadmissible increase of control subsystem.

The technical contradiction represents the conflict of TWO PARTS of a system. An inventive situation is inherent usually with some group of the technical contradictions in the system. The choice of the one contradiction from the group means transition from an inventive situation to a starting phase of the problem solution.

3. The physical contradictions have the form:

The given zone (element) should have the property A, to execute necessary function, and should have the property non-A /anti-A/, that satisfy to conditions of a problem.

The physical contradiction presents the inconsistent requirements to a physical condition OF THE SAME ONE (SINGLE) PART. Successful formulation of the physical contradiction shows usually the problem's nucleus. The contradiction is exacerbated maximally, that often makes the problem's solution straightforward.

Current TRIZ provides very powerful tools for resolution of problems with the technical and physical contradictions on the base of the hierarchical structure-function-behavior description of the technical systems [2]. We can define these contradictions from this point of view as:

Technical Contradiction - an interconnection between two sub-systems (or independent functions) when improvement of one of the sub-system (function) causes deterioration of the other one and vice versa;

Physical Contradiction - a demand to the elementary sub-system (function) of technical systems to have essentially contrary properties (parameters) in order to satisfy different requirements.

Because of diversity of various technical systems, problems that can occurs in these systems have different native structure. It is important to recognize the problem's structure because different TRIZ problem solving tools work better or worse for different problems [2]. Clarification of a structure of problem should simplify also manipulations with the knowledge databases of TRIZ [1,2]. For simplicity we will operate with sub-systems of a technical system in this paper, although the proposed ideas are also correct for the technical functions. The main structures can be described briefly as:

Point problems have the physical or natural contradictions or NO contradictions at all. Point problems with the physical contradiction are usually hidden, but they are often the root cause of all other problems. Point problems with the natural contradiction are usually the subject of research that has to be performed in natural sciences. Often the natural contradictions are constraints in the system development. Note the non-monotone dependence for degree of problem's difficulty from the number of constraints [3]. In non-contradiction case the point problem are required to produce practically unconstrained different ideas, so creative methods of idea generation such as brainstorming, synectics, lateral thinking, etc. are quite useful for this as well as for administrative contradictions.

Pair problems have the single technical contradiction. They usually appear after reduction of inventive situation to a set of independent technical contradictions.

Linear problems have chains of contradictions. They usually can be presented as a sequence of independent technical and physical contradictions in the same system.

Triangle problems have three dependent technical or physical contradictions. A triangle problem is the simplest case of a loop problem that has a set ( with no less than 3 entries) of dependent contradictions. The heuristic method for loop problem rupture have been developed by the author. It will be reported elsewhere.

Star problems have set of independent technical contradictions. They usually appear when one characteristic (or sub-system) k of a system gets better, but other l, m, n, … get worse. If contradictions A with B, C and D are independent, the Star problems can be reduced to the Pair problems. Note that a star problem often has a hidden physical contradiction.

Network problems have set of dependent contradictions. They appear when l, m, and n sub-systems depend on each other and/or can be dependent on k sub-system, hence we have several technical and/or physical contradictions that are linked. Network problems become one of the structure simplification targets in TRIZ if the idea of mathematical contradictions is serviceable to resolve problems. In the particular case when a network /star problem has (or can be separated by) triples, such problem can be often resolved by the Su-Fields Analysis [1,2].

Hierarchical problems have organized set of mostly human-like and technique-like contradictions, while Puzzle problems have disordered loops between previous structures. Hierarchical and Puzzle problems cannot be solved "as is". The algorithm for their simplification have been developed by the author. It will be reported elsewhere.

It seems beneficial to present simple structures of problems in terms of useful (UF) and harmful (HF) functions of a system [2]:

POINT with the physical contradiction : Find a way to enhance, provide UF_n that eliminates, reduces, prevents or does not cause HF_n in the SAME subsystem , i.e.

UF_n causes HF_n.

PAIR: Find a way to resolve a SINGLE technical contradiction by which (UF) should eliminate [HF] and should not cause preliminary [HF] in ANOTHER subsystem or/and (UF) should provide preliminary (UF) and should not cause [HF] in ANOTHER subsystem, i.e.

UF_n causes or/and require HF_k (k=/=n).

The pair problem is defined by two sub-systems and at least one link (causes or require).

TRIANGLE: Find a way to solve the simplest loop of DEPENDENT contradictions, i.e.

STAR: Find a way to resolve a FEW technical contradictions by which (UF) should eliminate a few INDEPENDENT [HF], i.e.

UF_n causes HF_k, HF_l, HF_m …. (k=/=l =/=m=/=n).

LINEAR: Find a way to benefit, eliminate, reduce or prevent HF under the SEQUENCE condition of UF or/and to perform the SEQUENCE steps of a system improvement, i.e.

UF_n-1 is required for UF_n and UF_n+l is required for UF_n or

UF_n is introduced to eliminate HF_k , while UF_n-1 is introduced to eliminate HF_k-1(k=/=n).

The understanding of the problem's structure gives a strategy and helps choose appropriate information needed for solving the inventive problems. It seems rational [3] to label the so-called physical contradictions as ANDNOT operator and present the so-called technical contradictions through XOR operators.

Let illustrate the reduction of a real problem in TRIZ style:

The administrative contradiction:

It is necessary to detect the number of small <300 particles in a liquid with very high optical purity. Particles reflect light very bad even if we use laser. What to do ?

The technical contradiction

If the particles are very small the liquid stays optical pure, but the particles are invisible. XOR if the particles are big they are detectable, but the liquid is not pure.

The physical contradiction:

The particles has to increase their sizes to be viewed ANDNOT to increase the sizes to keep the optical purity of the liquid.

After such transition the search of solution is pretty straightforward.

Conclusion

Designers deal with more complicated situations to compare with the simple AND/OR graphs that are studied, for example, in Artificial Intelligence. The simple logical operators AND together with OR can be more complicated in design reality, e.g., when the output behavior occurs only if all (or part of all) input functions take place in some usually predetermined order. The configuration of problems in terms of contradictions and sub-systems and functions of a system helps to resolve the technical problems.

References

1. G.S. Altshuller "To Find an Idea", Nauka Novosibirsk, 1991, 226 p.

2. S. D. Savransky, "TRIZ", (To be published, 1999, 454 p.).

3. S. D. Savransky, Attributes of the Inventive Problems , AAAI Spring Symposium on Search Techniques for Problem Solving under Uncertainty and Incomplete Information, Stanford University, March 1999.

4. S. D. Savransky, "Role of contradictions in problem solving", AAAI-99 Sixteenth National Conference on Artificial Intelligence, Orlando, July 1999, Abstract ID: A376