CC BY
2DOI: 10.24412/cl-37100-2023-12-56-64
M.S. Rubin, A.V. Kulakov, A.V. Trantin
Analysis of density problem as a tool for ranking a complex of problems
SUMMARY
Relevance and practical demand for methods for decision-taking is preconditioned by the restriction of resources irrespective of the kind of activity. The presence of a simple and objective method for selecting one option out of several ones enables to concentrate the effort in required direction without any quality losses. The present article is focused at disclosure of the essence of the method developed by the authors and consisting in simple and objective ranking of problems and tasks, based on the new notion of problem situation density. In order or the method to be practically used by other users the authors developed an algorithm for ranking problems and tasks, which was many times verified in the process of practical activity. This algorithm and some of the cases offered by the authors are also quoted in this article. The review of the existing methods for ranking concepts, problems and for decision-taking has been made for the purpose of identifying disadvantages, hindering the practical use of these methods, by the broad circle of users.
INTRODUCTION
The overwhelming majority of modern production facilities in this or that way include a number of different interrelations, which exist between machines and mechanisms, tools and materials, human being and equipment.
The authors noticed in the course of their practical activity that at such variety of system elements it is difficult to select a correct and vital problem. However, after selecting the problem, one cannot be always sure that the solving of this problem will lead to attainment of the goals, set before the enterprise. However, it is not possible and not feasible to get scattered between all problems.
The authors singled out at least three reasons:
• Restriction of resources of TRIZ-specialists and participants of project teams at industrial enterprises;
• Degree, rate and complicacy of implementing the solutions, obtained during the process of project performance at the functioning batch and serial productions;
• Loss of reputation of project team participants.
"Selection of the problem largely preconditions the fate of the invention and the inventor".[1]
In problem ranking such problems should be certainly selected first, which more than other problems influence the hindrance of enterprise development and the solving of which could bring the highest economical effect. Such logics could lead to erroneous selection of priority directions in project development.
In order to enhance the objectiveness of this process, parametric landmarks are introduced -consumption factor, time parameters, economic criteria. However, such logics could fail in selecting the problem. For example, based on their experience of project activity the authors noted that if the overall amount of consumed electric energy dominates over other kinds of energy, however, at the same time the number of different consumers of this energy is very large, it will be necessary to solve many problems associated with each consumer, not one problem. Therefore, according to the opinion of the authors of this article, the issue of methodology for selecting correct problems is also vital and open as of today.
In such situations interviewing the owner of the problem or the problem-setter and contiguous departments could partly eliminate this problem, however, in this case the risk of adding subjective opinions; it has to be borne in mind that for the problem owner his problem is the most important and significant, which does not add any objectivity to the process of problem ranking. Similar situation could be encountered as a result of addressing the experts, since their prognoses come true with such probability, which is no higher than random guessing. [2]
In the course of practical introduction of TRIZ at production facilities of aluminum industry the authors discovered the demand for a tool of objective selection of directions for search and subsequent ranking of identified problems.
In order to solve this problem, the authors of the article introduced the new notion of «problem situation density», developed the method for analysis of problem density, based on this notion and developed an algorithm for using this method, which was tested on several cases from the project activity of the authors.
REVIEW OF REFERENCE SOURCES
The problem of correct selection of tasks was stated by G.S. Altshuller back in 1961. "Each of these machines could be improved and in each of them new inventions could be made. It is also possible to create other machines - completely new, such that don't exist. But where to begin? Which problem should be solved first?" [1].
In TRIZ the ranking elements are found, for example, in benchmarking, value engineering analysis (VEA) and flow analysis.
The goal of VEA is the lowering of expenditures per unit of useful effect. [3,4] However, this approach could yield erroneous conclusions in such cases, when the analyzed technical system has complicated incorporated structure of subsystems and formation of expenditures is not restricted by one element. At the same time the creation of a precise value engineering model requires a higher number of resources and the result can be improbably cumbersome, which would hardly simplify the identification of problems, which are necessary to solve. [5]
Similar situation could be encountered in performing flow analysis. For example, identified leak of useful flow at the upper level of the system, if it is analyzed at the lower system level, will appear to be scattered throughout the multitude of elements of subsystems. And the answer to the question, with which element of the subsystem it is necessary to interact further, unfortunately could be less definite as we might wish.
It is necessary to note that outside the ambience of TRIZ the necessity for problem ranking is also vital and the ranking methodologies, which are verified and workable, are in no lesser demand. In the article «Methodology for quality problem ranking...» it is said that «...elimination of failures, which were observed in a complicated technical object, should be regulated in terms of labor intensity and expenditures. .As the practice shows, the ranking of quality problems based on one feature only (for examples, massive scale or expenditures) is not efficient. It is necessary to perform corrective action both as applied to mass, but insignificant defects and as applied to significant (expensive) defects, discrepancies or breakdowns. .In order to define the priority for elimination of this or that problem, it is necessary to determine a complex indicator, which would include at least the evaluation of mass involvement and evaluation of significance». [6] The methodology proposed in [6], is applicable for selection of priority problems, but has a specific applicable character and requires the accumulation of statistical data regarding failures. The tool becomes cumbersome and labor intensive.
Moreover, the problem of selection and ranking of problems at an enterprise is a private manifestation of a broader problem of decision-taking. Therefore, it would be a disadvantage of the authors not to take into account the achievements of the theory of decision-taking.
In the article by the Nobel-prize winner G.Symon and his colleague A.Newell [7], the authors single out so-called well-structured and poorly structured problems. Well-structured problems are understood as those, which can be expressed in figures and symbols, i.e., have a parametric form of expression. Poorly structured problems, for their part, have no parametric expression and the decision is taken based on subjective preferences of the person, who takes the decision.
The increase in the degree of the problem being structured is possible due to the enhancement of decision-taking criteria. At that both classical methods of solving a multi-criteria problem [8, 9], and the modern methods based on these above-mentioned previous developments exist. [10]
However, the application of support system based on these methods is rather labor consuming. These methods and systems are most often applied for solving such problems, in which the expenditures on development and implementation of these systems are compensated. To these problems refer, for example, the problems for planning the activity of the corporations, design of complicated technical systems, selection of variants for operation of expensive equipment, etc. Application of complicated mathematical models is first of all economically feasible in such cases. [11] However, with systems, which are less equipped with resources the problem of decision-taking is less vital due to such a simple reason that they are less tangible to wrong solutions against the background of costs incurred on taking this decision. Therefore, the demand for a simple and objective methodology for selection and ranking is vital.
THE BASIC PART
As it was marked in the reference review, the existing methodologies can either lead erroneous conclusions, or require significant preliminary preparation and creation of information structure on gathering various data.
In order that the problem stated in this article should be clearer, let us quote an example. There is a technical system, in which the overall amount of expended energy dominates over other kinds of energy used. However, the technical system consists of a large number of different users, who use electric energy and the statement of problem of energy saving will lead to the fact that it will be necessary to solve not one problem, but many problems at once applied to each consumer, while taking into account the restrictions, mentioned in the introduction to the present article, the feasibility of such approach is problematic. In order to avoid such mistakes, the authors of the article stated the goal of developing such a method, which would enable to take feasible decisions regarding the selection of this or that direction and the task for decision-taking. The method is based on the notion of problem situation density (problem density).
The density of problem (of problem situation) is a numeric indicator of a problem situation, which shows specific value of this or that characteristic of the object. The characteristic could be understood as consumed (or generated) energy, expendable materials, amount of reject, labor payment fund, number of failures, expenditures on operation, maintenance and other parameters, which numerically characterize this or that analyzed object and the problem situation associated with it.
The specific value of this parameter could be reduced to the area, at which the cumulative value of this parameter is formed, to the unit of the overall amount of equipment, overall number of steps or elements, to the unit of time, to the overall number of operation and servicing personnel, etc. Based on the value of problem situation density, it is possible to rank different problem situa-
tions and stated problems. The higher the specific density of the problem, the more vital could be the analyzed problem and the more promising could be the solution of it for the implementation of the project or for the improvement of parameters of the enterprise as a whole.
The authors also noted that in problem ranking one can use not only the density of problem characteristic, but also the density of useful characteristic.
The following algorithm was developed in order to formalize the method and to obtain a possibility to transfer it for future application by other users:
• Select a complex of objects or processes for analysis. Selected objects or processes should have one measurable parameter, which should be common for all of them (physical, economic, social, etc.). For example, equipment or technological process of a particular manufacturing plant is selected as an object.
• To define the goal of analysis: ranking of objects or processes for the purpose of singling out problems and placement of tasks; defining the potentiality of certain concepts or known solutions. For example, to decrease the expenditure of energy, increase productive capacity, decrease the violations of discipline, define the potentiality of the energy source, etc.
• To define the parameter values, according to which the ranking will take place. On the one hand, they should embrace all objects or processes chosen for selection, while on the other hand they should match the selected goal of analysis. Selected parameter values should correspond to system level of TS. For example, the selected parameter values can take the form of energy consumption or energy production, expenditures or profits, number of violations of law or of prizes, expenditures on operation, repair or maintenance, etc.
• To select the parameter values of objects and processes, the measuring unit of which will control in defining the density. It could be presented in the form of spatial, temporal and quantitative characteristics. For example, the area of a workshop, the length of a pipeline, the number of energy consumers (in units), changes in sales volume in time, number of personnel, etc.
• With characteristics selected in item 3 it is necessary to calculate specific characteristics (of density) per unit of characteristics selected in item 4. If we are talking about the distribution according to the number of units, it is necessary to take into account if they belong to the same type or not (i.e., if we shall have to solve the problems separately). For example, the specific expenditure of gas with each workshop per year averagely or in each season. Or the amount of reject at similar products made with the aid of similar equipment.
In case the selected analyzed object is the process with processing stages (steps) or the flow, it is necessary to take into account the influence of cumulative effect of expenditures in the flow upon the density of the problem.
• In order to compare several characteristics (densities of characteristics), benchmarking should be used.
• To analyze the results of ranking from the viewpoint of correspondence to problems stated in item 2. If the goals are not attained, it is necessary to pass over to item 1.
EXAMPLES OF APPLICATION OF THE METHOD
The method was tested on practical examples from the experience of project activity of the authors.
One of the products of the plant was selected as the object for analysis. In the course of analysis of the enterprise the data was obtained on costs for manufacturing this product, while the characteristic was understood as energy expenditures on the manufacturing of the product. The distribution of energy expenditures is presented in Fig. 1
22,9
^ Electric energy > Gas v Circulating water «Compressed air Figure 1. Distribution of energy expenditures Rbs/t
Based on the diagram presented in Fig. 1, it is possible to make a conclusion that first of all it is necessary to deal with issues of energy saving. Having recorded this conclusion, let us pass over to problem density analysis and after that we shall compare the conclusions obtained.
In order to perform the analysis, the authors received from the enterprise not only the data on product costs, but also process design, which characterizes product manufacturing and includes the list of equipment used. The number of energy consumers was selected as a characteristic, the unit of which will define the density in the course of further analysis. Table 1 quotes the obtained data and the results of calculating density values with regard to similar equipment:
Table 1
Type of energy Cost, Rbs/tonne Number of individual Problem density,
consumers, units. Rbs/tonne*unit.
Electric energy 327 12 27.3
Gas 212.1 1 212.1
Circulating water 74.5 2 37.3
Compressed air 22.9 2 11.5
As it is seen from the Table 1, the introduction of specific value redistributed the priorities from electric energy in favor of gas and circulating water.
Similar approach could be used in the analysis of losses in water pipeline networks. The situation is worse not in the water network, in which more water is lost, but in that one, with which the relative parameter value of water losses per length of the water network or per overall volume of water consumption in this network is higher. Let us consider it on example of two cities - Lisbon and Tokyo.
In terms of absolute parameter values the situation with water losses in Lisbon is much better - the number of losses is 9.3 times less than in Tokyo. However, if we include the specific characteristic with our analysis, the situation fundamentally changes - the losses per 1 km of network are 70 % higher than in Tokyo (see Fig. 2)
Figure 2. Comparison of absolute and specific characteristics of losses in water network
in Lisbon and Tokyo
Another example is proposed for analysis: data on technological waste in manufacturing of a certain product. Fig. 3 shows the expenditures incurred by the manufacturing plant at each of the alterations of the technological sequence of manufacturing as well as normative percentage of technological wastes remaining after each step of manufacturing.
Figure 3. Comparing the expenditures due to manufacturing alterations with the amount
of obtained technological wastes
A conclusion could be drawn from the quoted diagrams that the efforts first of all should be directed at the work with alteration 3, since the highest expenditures in conventional units are concentrated there, or with alteration 2, since there we have the highest percent of obtained technological wastes. But what alteration shall we begin with?
If we address the characteristic of problem density, which takes into account the expenditures of all previous alterations, we shall obtain a different result, which will certainly direct us towards the top-priority direction of work:
Table 2
Alteration 1 Alteration 2 Alteration 3 Alteration 4 Alteration 5
Expenditures on alterations 1.4 1.2 2.4 2.2 1
Waste after alteration 10% 15% 9% 8% 6%
Cumulative expenditures on wastes after alterations 0.14 0.39 0.45 0.58 0.49
Thus, it is possible to draw a conclusion that the most «morbid» alteration is exactly the alteration 4, in which based on the parameter of problem density, the greatest losses of the enterprise are concentrated. Let us note that before introducing the notion of problem density we never spoke about alteration 4 and the choice was made between alterations 2 and 3.
In the course of developing and practical testing of this methodology the authors noticed that in calculating the problem density (the density of useful characteristic) it is also necessary to take into account the effect of replication - expenditures on the search of solution are not multiplied by the number of similar or identical objects. Let us quote a practical example.
Table 3 quotes the structure of expenditures on the manufacturing of products according to the type of expenditures. It is seen that after introducing the notion of problem density with regard to the number objects, the structure underwent changes: the transition was from the concentration of attention on consumers of electric energy to users of crude oil.
Table 3
Percentage of expenditures in rubles, % Number of sites according to these characteristics Index of characteristics with regard to density
Gas consumers 18% 5 26,2%
Electric energy consumers 57% 21 19,7%
Crude oil consumers 12% 2 43,6%
Compressed air consumers 8% 7 8,3%
Cooling mixture consumers 5% 17 2,1%
Now let us introduce a correction regarding the effect of replication. Table 4
Characteristics of objects Identical objects from the viewpoint of planned analysis Groups of identical objects Index of density with regard to replication
Gas consumers 5 1 56.9%
Electric energy consumers 4 18 2.4%
Crude oil consumers 0 2 31.6%
Compressed air consumers 4 4 4.5%
Cooling mixture consumers 17 1 4.6%
It is to be seen from Table 4, that with regard to identical objects the priority of directions also changed - first of all it is necessary to deal with gas consumers. The consumers of electric energy were practically devoid of attention due to a great number of various consumers, they should be the last consumers to deal with.
As it is seen from the quoted example, the method, which is rather easy to master, enables to precisely define the priority directions, which are characterized by maximum potentiality of efficient use of resources for solving problems.
CONCLUSION
• The authors identified and formulated the reasons, due to which the identification of problems and setting the tasks at the enterprises is a key issue and preconditions both the future of individual projects in particular and the tempo of implementing TRIZ at the enterprise on the whole;
• It was shown by the authors that the existing methods for decision-taking and for selection of directions could either yield erroneous conclusions or require significant preliminary preparation, creation of information infrastructure, involving the gathering of comprehensive data and use of complicated mathematical apparatus, which significantly reduces their application in practice;
• For the first time the authors introduced the notion of problem situation density (problem density) and the density of a useful characteristic;
• The authors developed the method and the algorithm for using this method based on the notion of problem situation density and the density of useful characteristic for ranking the directions for improvement as well as task ranking;
• The authors demonstrated the practical use of newly introduced notions and the developed algorithm for analyzing the density of problem characteristic for obtainment of results, which are unexpected and different from those results, the obtainment of which is based on standard logics for task selection;
• The authors noted, that in calculating the density of problem situation (useful characteristic) it is necessary to take into account the number of identical objects, with which it is possible to replicate the found solutions;
• The authors noted, that in ranking according to several characteristics with regard to their density it is possible to use benchmarking or other known methods for ranking according to several characteristics.
The authors suggest the following directions for further research of developed method and its optimization:
• Testing the analysis of problem situation density (useful characteristic) on a large number of cases, formulation and formalization of constraints of this method;
• Improvement of algorithm of using the method involving the formalization of selection of typical sets of characteristics depending upon the field and kind of activity;
• Enhancement of precision of the method through introduction of corrective multipliers of the lifecycle stages of a technical system with the further optimization of algorithm for using this method.
REFERENCE
1. G.S.Altshuller. Way of learning to invent. - Tambov: Book publishers, 1961. - P. 113-123. https://www.altshuller.ru/triz/triz68.asp
2. Nate Silver. The Signal and the Noise - Penguin Books, 2015. - 576 pages.
3. V.M.Gerasimov; V.S.Kalish; M.G.Karpunin, A.M.Kuzmin, S.S.Litvin. Main provisions of methodology for performance of value engineering analysis: Methodological recommendations. - M.: Inform - FSA, 1991. - 40 p.
4. M.S.Rubin, O.M.Gerassimov. "Concerning the methods of analysis of problem situations and selection of problems. An attempt at review." https://triz-summit.ru/confer/tds-2007/203814/203837/
5. T. Yoshikawa, J.Innes, F.Mitchell. A Japanese case study of functional cost analysis, Management Accounting Research,Volume 6, Issue 4, December 1995, Pages 415-432.
6. Vladimir N. Kozlovskiy, Aleksey V. Zayatrov, Natalya V. Afinogentova. METHODOLOGY FOR QUALITY PROBLEMS RANKING FOR HIGH-TECH PRODUCTS IN ENGINEERING BY ECONOMIC CRITERIA. Mathematical methods, models and information technologies in economics, 2016 http://www.irbis-nbuv.gov.ua/cgi-bin/irbis_nbuv/cgiirbis_64.exe? C21COM=2&I21 DBN=UJRN&P21DBN=UJRN&IMAGE_FI LE_DOWNLOAD= 1 &Image_file_name=PDF/ape_2016_3_41.pdf
7. H.Simon, A.Newell Heuristic problem solving: the next advance in operations research/Operations Research, 1958, v.6.
8. M. Ehrgott, Multicriteria Optimization, Springer, 2005.
9. K. M. Miettinen, Nonlinear Multiobjective Optimization, Kluwer Academic Publishers, 1999.
10. Joseph Gogodze. Ranking-Theory Methods for Solving Multicriteria Decision-Making Problems, Hindawi Advances in Operations Research Volume 2019, Article ID 3217949, 7 pages https://doi.org/10.1155/2019/3217949
11. A.V.Lotov, I.I.Pospelova. Multi-criteria problems involving decision-taking: Manual. - M.: MAKS Press, 2008. - 197 p.
