METHODS FOR THE SEARCH FOR SUBSTITUTE PROBLEMS AND TASKS IN THE INVENTIVE PROGRAM "NOVATOR 4.0"

 

V.N.Glazunov, Russia 

 

 

The notion of problem situation (PS), substitute problem and substitute task. Transition from problem situation to the task (-s). Modeling PS by means of ternary language of description. Extension of PS model using the methods of ascending and descending cause/effect analysis. Distribution of values characterizing "demand" and search for problems on the PS model. Object and parametrical contradictions on PS model. Identification of PS elements. Identification of the order of solving source problem and substitute problems. Search for substitute tasks. Examples of search for substitute problems and tasks. Demonstration of search for substitute tasks using the Novator 4.01 software program.

 

Introduction

 

Proposed methods of search for substitute problems and tasks have been developed within the scope of the "Novator 4.0" project. The goal of this project is to create an inventive software program of the second generation. The result of operation of the programs of this class takes the form of finished variants of a concept for a device or technology to be developed, which are prepared as a standardized scientific research report. Work with such programs does not require from a user any in-advance training and mastering of methods for inventive problem solving.

 

The indicated requirements could be met only under the condition that the operation of the inventive software program is based on formal methods for problem solving. In this connection the proposed methods of search for substitute problems and tasks are strictly formal.

 

 

1. The notion of problem situation, substitute problem and substitute task.

 

TRIZ methods or software programs for TRIZ methods are used by developers (users) for solving engineering problems. Usually the problem is understood by a human in the form of a certain problem situation (PS).

 

The problem part and the situational part in the problem situation are interconnected and are inseparable from one another. Problem is what does not comply with the requirements set forth within the scope of a situation. Situation is a set of conditions and circumstances, which are to be taken into account when solving a problem under consideration.

 

Verbal structures of the following type are used to fix a problem in the description of the problem situation:

"action - object", - for example, "to generate a magnetic field",

"action - process - object", - for example, "to eliminate the vibration of case",

"action - parameter - object", - for example, "to measure the temperature of liquid"

and so forth.

 

In its turn, the situation is a structure that is formed by logical and specific relationships established between particular objects, features and relations. Examples of situation descriptions are given below in Section 3.1. 

 

A problem that was initially comprehended by a user could be called a source problem. Usually the source problem is associated with other problems, which are explicitly or non-explicitly indicated in the problem situation. These problems could be the reasons for the emergence of the source problem or could be the consequences of the latter. Such kinds of problems could be called substitute problems because solving of these problems enables to eliminate the source problem without solving the latter directly.

 

Source problem is associated with substitute problems through cause/effect relationships. The set of these relationships determines the ways of transition from a source problem to a substitute problem.

 

As practice shows, in some cases it is easier to eliminate the substitute problem than the source problem.  For this reason, the identification of explicit and non-explicit substitute problems in a problem situation is of great practical interest.

 

As a rule, the description of a source problem situation is poorly structured. It has no clear subdivision into problem part and situation part. Singling the problem out from the context of the problem situation transforms the former into a task. Subsequently this problem could be solved both taking the constraints indicated in the situation part into account and ignoring these constraints. 

 

When singling the source problem and substitute problem out from the problem situation the cause/effect relationships existing between them are rejected. As a result, these problems are perceived as separate ones that are not mutually interconnected. Therefore, a transition between such problems could be called contextually dependent (see an example in section 3, item 9, 3rd list of problems).

 

On the other hand, the research conducted during the development of the Novator 4.0 program showed that there are at least 11 non-symmetrical relationships of transition that are not contextually dependent. This fact enables to establish the relationships of transition between the problems that are implemented in any situation. In other words, a transition between the problems in compliance with the indicated relationships is contextually independent. A list of these relationships is given in Table 1.

 

Relationships of transition between the problems

Table1

No.	Transition relationship
1	Transition to an identical problem based on term
2	Transition to a general problem based on term
3	Transition to a specific problem based on term
4	Transition between specific problems based on term
5	Transition to an identical problem
6	Transition to a general problem
7	Transition to a specific problem
8	Transition between specific problems
9	Transition between sub- problems
10	Transition to a "simpler" problem
11	Transition to a "similar" problem

 

Some non-symmetrical transition relationships are determined by an expert based on his knowledge, while the remaining relationships are calculated during compilation of the of Novator 4.0 database.

 

A set of problems and non-symmetrical relationships of transition form a structure that is a part of ontology of problems pertaining to a particular subject area [1, 2].

 

Since the transition relationships between the problems are asymmetrical, and the number of problems in a subject area is restricted, there exists a finite number of problems associated with it. If some problem is looked upon as a source one, then all problems associated with it through relationships of transition could be called substitute ones.

 

Fig. 1. Fragment of problems ontology.

 

The presence of sequence of transitions between the source problem and any substitute problem guarantees that all solutions of a substitute task are also solutions of the source task. 

 

The transition from the source problem situation to substitute problems, and then to substitute tasks in the Novator 4.0 program is implemented in modules "Situation analysis" and "Concept development".

 

 

 

 

 

 

Fig. 2. Functional block diagram of Novator 4.0 program.

 

 

 

2. Short description of steps for the search for substitute problems and tasks

 

2.1. Modeling a problem situation

 

A user records the initial description of problem situation in the "Task Statement" module. The initial description of problem situation is usually poorly structured. It consists of parts having different formats - namely, fragments of text, drawings, diagrams, mathematical formulas, etc. Such information cannot serve as source data for a formal method (see section "Introduction"). Therefore, first the developer should create the model of the problem situation based on its initial description.

 

As far as the author knows, at present only binary languages having two categories are currently used for modeling problem situations. Two types of binary languages for description are possible: object-relational and attribute-relational.

 

The object-relational languages are based on the categories of object and relationship, while the attribute-relational languages are based on categories of feature and relationship. To type 1 relate, for example, the language of Su-Field analysis [3] and the "object - action" language, which is used in the TechOptimizer software (product of IMCorp). To type 2 relates the language, which is used in the parametrical method for analyzing contradictions [4] and Novator 3.01 program (product of "Metod" company).

 

The languages of both types are limited because the model of the situation can concurrently contain:

·       objects that form the situation, including engineering systems, their subsystems and components;

·       features of objects;

·       relationships between objects;

·       relationships between features - for example, between the parameter values characterizing the engineering system and the features of its components.

 

Therefore, in the most general case a language is necessary, which uses three categories: object. Feature and relationship. Such a group of languages can be called ternary languages [5, 6].

 

A variant of ternary specification language has been developed for the program Novator, which is based on formal ternary ontology  (product of Method company), which, in its turn, is implemented as a specialized object graphic editor. Using this editor, the developer can create a model of this situation as a functional diagram.

 

In order to fix the problem in the situation model, the modality "demand" is used with such meanings as: "indefinite", "desirable", "undesirable", "desirable and undesirable". If the meaning of demand characterizing the situation element (object, change of feature or relationship) is equal to "desirable", it means that the essence of the problem is the absence of the required element in the system. In the opposite case the problem is reduced to the presence of undesirable element in the situation. If the meaning of the demand characterizing the element is "desirable and undesirable", there is a contradiction in the situation, i.e. antinomy - problem [4, 7].

 

The alphabet of the analyzed variant of ternary language is presented in table 2.

 

 

 

 

The alphabet of ternary specification language

Table 2

Group	Symbol name	Graphic form of the symbol
Objects	Material object
	Process
	Dynamic object
	3-d object
	System	The system is determined by the presence of an element or subsystem
	Subsystem or its element	The subsystem is determined by its being inscribed into the system
Features	Parameters
	Indicators
Relationships	Leads to
	Spatial relationships	symmetric asymmetric
	Time relationships
Actions	Action upon
	Action from

 

 

Continuation of Table 2

Actions	Action between
	Function from	Similar to "Action upon"
	Function upon	Similar to "Action from"
	Function between	Similar to "Action between"
Changes	Change
	Increase
	Decrease
	Presence
Demand	Indefinite	Is marked with black color of the situation element.
	Desirable	Is marked with green color of the situation element.
	Undesirable	Is marked with red color of the situation element.
	Desirable and undesirable	Is marked with lilac color of the situation element.

 

The rights for creating the language formulae (admissible combinations of source symbols) are determined by the accessibility of selecting this or that symbol in the menu of object graphic editor during the process of creating a problem situation model.

 

2.2. Model extension

The model of problem situation fixes source idea of the problem, which the user has. However, it is often possible to find the best way of eliminating the problem situation by solving not the source problem, but a certain problem associated with it (substitute problem).

 

In order to identify what problems are associated with the source problem,  Novator uses the methods for descending and ascending cause/effect analysis. The first of these methods determines the causes of the source problem, while the second - its consequences. In order to identify the reasons of the source problem, Novator asks the User a sequence of questions  "Why does… happen?", why in order to identify the consequences of the same problem it asks the User a sequence of questions of the type "What does… lead to?"… Based on user's answers, Novator creates two interconnected cause/effect chains: descending and ascending. An abstract example of such chains is given in Fig. 3.

 

Fig. 3. Ascending and descending cause-effect chains.

 

Further on, Novator proposes that the User should extend the source situation by including new objects, relationships between them as well as the features of these objects into it.

 

In the course of developing the model of problem situation the User can several times go back to the procedure of model extension.

 

 

2.3. Search for problems

If the meaning of the demand of at least one element is assigned on the model of the situation, Novator can find the meanings of demand characterizing its other elements. For this purpose Novator uses the entire set of relationships, which are indicated in the model of the situation.

 

When the meanings of demand are determined, there may be cases, when opposite meanings are ascribed to one and the same element of the model: "element – desirable" and "element  – undesirable". In this case Novator ascribes the third meaning to the demand of the element: "desirable and non-desirable". For the elements of model with such meaning Novator formulates antinomies - problems in the form of engineering or physical contradictions. [3].

 

Due to the use of ternary language of situation description, Novator can find not only parametrical [4],  but also objective physical contradictions. The latter type of contradictions consists in the demand for presence or absence of object or process (See example in section 3.1).

 

After the distribution of the meaning of demand, Novator formulates a list of all problems, which are characteristic of the situation.

 

 

 

 

 

 

2.4. Determination of the model elements

When the model of the situation is being developed, the User applies a set of terms, which are habitual for him.   On the other hand, the knowledge database of Novator contains generalized engineering solutions (concepts), i.e., solutions, which have the highest degree of similarity. A fixed set of terms exists, which is used for describing them.  Therefore, it is highly probable that the description of identified problem will not formally correspond to the description of its solution in the knowledge database of Novator.  In this case Novator will not be able to find the solution of identified problem. In order to eliminate this terminological non-coincidence, Novator proposes the user that the relationships should be established between the terms, which are used in the name of the problem, and the terms of the dictionary.

 

 

2.5. Determining the sequence of problem-solving

As a result of analyzing the extended model of problem situation Novator can formulate a very big list of problems.  A necessity to determine the priority order of solving them appears here. For this purpose the user should indicate the problem in the received list, which is the most important one, in his opinion. After that the Novator determines the priority order for solving all other identified problems. In calculating the priority order for solving this or that problem Novator takes into account its "distance" from the first, the most important problem.  Novator also takes into account the potentiality of problem solving – a complex indicator, which is calculated, when a knowledge database is compiled.

 

Novator selects the source problem by default, as a most important problem. Thus, if the user did not indicate the most important problem, Novator will arrange all identified problems according to their importance (priority order of solving them).

 

When the priority order for problem-solving is being identified, Novator refers the antinomies - problems to the end of the list. It is associated with the fact that Novator is intended for unprepared users, for whom it is always difficult to correctly interpret some abstract methods for contradictions resolving.

 

 

3. Examples of search for substitute problems and tasks

 

3.1. Example of transition from source problem to substitute problems

A scenario of analysis of test problem situation using the program Novator Version 4.02. Some actions of the main menu are mentioned in scenario: "Study the problem", "Find problems", "Identify elements" and "Determine priority order". The majority of these actions exactly corresponds to steps, which are indicated on the functional diagram (See Fig. 2).

 

 

Scenario of situation analysis

 1  The user creates a source model of the situation and indicates a non-desirable element in it - process "destruction".

 

 

 

 

 

2  The user selects in the menu of the program the action "Find problems" and after that Novator compiles the 1st list of problems.

 

List of problems No. Problems Put aside Number of solutions 1 Eliminate - destruction  - surface layer - wing   2 Statistics Total number of problems - 1, number of fully identified problems - 1.  Number of solutions: 2 -

 

 3   The user selects in the menu of the program the action "study problem".

 

1) Novator asks the User a number of questions of type 1 (see section 2.2).

 

Study of the problem   Why destruction - surface layer - wing? Because: ·        cavitation bubbles collapse on the surface of the wing.   Why do cavitation bubbles collapse on the surface of the wing? Because: ·        the wing moves in water at a high speed.   Why does the wing move in water at high speed? Because: ·        it is necessary to raise the body of the ship above the water.   Why is it necessary to raise the body of the ship above the water? Because: ·        it is necessary to decrease hydrodynamic resistance of the ship.   Why is it necessary to decrease the hydrodynamic resistance of the ship?  Because:

Note.

Here and below the questions of Novator are marked blue, while the answers of the User are marked black.

 

2) After the last question is displayed the User presses the key [Final cause]. Novator creates a descending tree and displays it in a corresponding window.

 

 

 

 

 

 

 

 

3) Novator asks the User a number of questions of Type 2 (see Section 2.2).

 

Study of the problem What does destruction – wing lead to? Destruction – wing leads to:

 

4) After this question is displayed the User presses the key [Inadmissible consequence]. Novator crates an ascending tree with one peak, which is not active.

 

 4  The User fulfills the recommendation of Novator and extends the model, using the results of study.

 

 

5  The User selects in the menu of the program the action "Identify the problem" and after that Novator forms the 2nd list of problems.

 

List of problems № Problems Put aside Number of solutions 1 Eliminate - destruction  - surface layer - wing   2 2 eliminate - collapse - cavitation bubble   259 3 create  - lifting force - wing   2 4 create - lifting - vessel body   16 5 Retain - eliminate - motion - wing   0 Statistics   Total number of problems  - 5, number of fully determined problems (out of these 5) – 3. Total number of solutions: 279.

Note.

Names of problems, which contain indefinite terms, are grey-marked here. The problems, which contain indefinite terms, have a significantly larger number of solutions than similar problems, which are fully determined. 

 

 6  The user selects the action "Determine elements" in the program menu. Novator displays a corresponding window. The user establishes coordination between the names of indefinite elements and the terms of Dictionary.

 

Elements determination Name of the element model Term from the dictionary Collapse Some process Cavitation bubble Some material object Vessel body Solid body -

Note.

The user did not find the terms in the Dictionary, the meaning of which is similar to that of model elements "Cavitational bubble" and "Collapse".

 

 7  Novator forms the recommendation.

 

 

 

 8  The User changes extended model of the problem situation.

 

 

 9  The User selects in the program menu the action "Identify the problems" and after that Novator forms the 3rd list of problems.

 

List of problems No. Problems Put aside Number of solutions 1 eliminate  - destruction - surface layer - wing   2 2 eliminate - cavitation - water   5 3 create - lifting force - wing   2 4 create - lifting - vessel body b 2 5 create-eliminate - movement - wing b 0 Statistics: Total number of problems - 5, Number of fully determined problems(out of these 5) - 5 Total number of solutions: 11 -

 

10 The user switches over to the mode "Develop a concept". Novator shows the solutions of the problem (task), which is the 1st in the list: "eliminate  - destruction - surface layer - wing".

 

After the database is compiled, the ontology contained in it would include a number of substitute problems (goals), which are associated with the problem "eliminate - destruction - surface layer -  wing". These are the problems:

·        "eliminate  - destruction - surface layer - solid body"

·        "protect from - destruction - surface layer - wing"

·        "protect from - destruction - surface layer - solid body".

 

Novator database currently contains the concepts for solving only the first problem. The wording of these concepts is given below:

·        Protective coating eliminates the destruction of surface layer of solid body;

·        Reconstruction of surface layer of solid body compensates for its destruction.

 

 

3.2. Switch-over from source problem to substitute problems

As it is seen from the functional diagram (See fig. 2), Novator can use not only the description of problem situation, but also the condition of the problem as source data. In this case such a problem could be called source problem.

 

Switch-over from source problem to substitute problems is performed automatically by Novator. The user only selects in the list one of the tasks (goals) and afterwards Novator finds its direct and substitute solutions. The result of analysis of test task by the program Novator Version 4.01 is given below.

 

Task: Find the concepts of methods for softening water.

 

To solve the stated problem (to attain the goal), Novator found the following direct and substitute concepts.

 

 

For each substitute concept Novator forms an explanation showing why the use of this concept enables to solve source problem (attain the stated goal).

 

Explanation of switch-over from source task to substitute concepts

Table 3

Substitute concept	Explanation of switch-over to substitute concept
Porous body adsorbs dissolved substance	This concept enables to attain the stated goal, because:  "dissolved salt is a dissolved substance" "remove - dissolved substance" is possible if "absorb - dissolved substance" (transfer to absorber) "absorb - dissolved substance" is possible, if "adsorb - dissolved substance"
Non-miscible fluid extracts the substance from the solution	This concept enables to attain the stated goal, because:  "dissolved salt is a dissolved substance" "remove - dissolved substance" is possible if "absorb - dissolved substance" (transfer to absorber) "remove - dissolved substance" is possible if "extract - dissolved substance"
Chemical sedimentation decreases the concentration of the solution	This concept enables to attain the stated goal, because:  "dissolved salt is a dissolved substance" Instead of "remove - dissolved substance" it is possible to "decrease - concentration - solution"
Addition of additional amount of solvent decreases the concentration of the solution	This concept enables to attain the stated goal, because:  "dissolved salt is a dissolved substance" Instead of "remove - dissolved substance" it is possible to "decrease - concentration - solution"

 

 

 

Reference

1.     Gavrilova T.A., Khoroshevsky V.F. Knowledge Databases of Intellectual Systems. Saint Petersburg, Piter Publishers, 2000.

2.     Guarino N., Giaretta P. Ontologies and Knowledge Bases. Amsterdam, IOS Press, 1995.

3.     Altshuller G.S. Creation as an Exact Science. Moscow, Soviet Radio, 1979.

4.     Glazunov V.N. Parametrical Method for Resolving Contradictions in Engineering. Moscow, "Rechnoy Transport Publishers",1990.

5.     Ujomov A.I. Things, Features and Relationships. Moscow, Publishing House of the USSR Academy of Science, 1963.

6.     Ujomov A.I. System Approach and General Theory of Systems. Moscow, Mysl Publishers, 1978.

7.     Gorskaya T.G. Typology of Dialectic Contradictions in Cognitive Activities.  "Voprosy Filosofii" Journal, 1981, No. 11.