Decision-making Model for Designs
Editor | On 05, Oct 2009
By Manabu Sawaguchi
Abstract
This paper features the decision-making model to measure the ideality of products, based on the Theory of Inventive Problem Solving (TRIZ). These products are designed using innovative activities from the point of view of evolution toward increased ideality as one of the patterns of technical evolution in the TRIZ methodology.
In order to give customers (especially high-end customers) highly-valued products, companies have to chase the ideal final result (IFR). The decision-making model, therefore, uses idealized conceptual designs to realize highly-valued products rationally from the standpoint of chasing the IFR.
The basis of the decision-making model is the analytic hierarchy process (AHP).4 It is devised to evaluate ideality from various aspects corresponding to useful functions (such as product design parameters) and harmful effects (such as side effects with useful functions).
Both the hierarchy diagrams (the AHP) and the series of useful functions and harmful effects in an object product use the decision-making model to measure the ideality index.
Introduction
The purpose of this study is to propose the decision-making model for evaluating the highly-valued product for customers in product development activities from the standpoint of evolution toward increased ideality.
According to the evolution toward increased ideality, the useful functions performance of main parameters will be improved and harmful effects such as weight, space, noise and cost will be decreased in technical systems. Over time the technical systems advance toward increased ideality. By using the AHP from the improvement of ideality in a TRIZ field a technique for design proposals can be used.2,7 In particular, TRIZ expert Vladimir Petrov referred to the idealization level behavior through nine cells of the contradiction matrix and explained the levels of idealization with examples.7
Outline of Proposed Model
In order to measure the ideality index for design proposal alternatives from the viewpoint of evolution toward increased ideality, the author suggests drawing two types of hierarchy structure diagrams. This is due to the proposed decision-making model based on the AHP: one is for useful functions and another is for harmful effects. Based on these two types of diagrams, the coefficient can be expected to be measured by showing the degree of ideality rationally for design proposals to be considered through TRIZ activities.
First the decision-maker unit has to estimate the weight (level of importance) for each design proposal by using the AHP and the desiring level. Next, calculate the weight (degree of incident) for each design proposal based on an acceptable level. Lastly, calculate for the ideality index for each design proposal by dividing the total sum of weight by each useful function and harmful effect to reach a design proposal.
The acceptable level is the minimum level that is adequate for customers. It is the opposite direction of the desiring level. The ideality index for each design proposal is bigger (more than 1) and it is valuable for customers (especially high-end customers).
Implementation Procedures of Proposed Model
The flowchart regarding implementation procedures of the proposed decision-making model is shown in Figure 1.
| |||||
Step 1: Making two types of hierarchy structure diagrams from the point of view of evolution toward increased ideality
Two types of hierarchy structure diagrams are drawn to improve required useful functions and to avoid reducing harmful effects.
Step 2: Estimation of weight (level of importance) of each useful function and harmful effect
Estimate the relative weight of each useful function and harmful effect by using the AHP (it is a type of relative measurement approach).
UWi: Weight of UFi (useful function) (level of importance for customers)
HWj: Weight of HEj (harmful effect) (degree of incidence against customers)
| |||||
| |||||
Step 3: Setting up a desiring level for each useful function at an acceptable level for each harmful effect
Define the desiring level for each useful function at an acceptable level for each harmful effect of a technical system (for example, a new product in the pipeline).
- minRi: Value of desiring level for UFi (useful function)
- maxURj: Value of AL (acceptable level) for HEj (harmful effect)
Step 4: Estimation of weight (degree of relative merit) of each design proposal for each useful function and harmful effect
Measure the outcome of each design proposal by estimating the achievement (weight) for the desiring level of each useful function .
- *minRi: DL (desiring level) of UFi (useful function ) normalized to 1
- UW (i, k): Weight of DPk (design proposal) for *minRi of UFi (useful function)
- *maxURj : AL (acceptable level) of HEj (harmful effect) normalized to 1
- EW (j, k): Weight of DPk (design proposal) for *maxURj of HEj (harmful effect)
Step 5: Estimation of total score of each design proposal for a series of useful functions and harmful effects.
Calculate the total score of each design proposal by method of weighted mean between the weight of each design proposal and each useful function. The weight of each useful function is estimated in Step 2. Then calculate the total score of each design proposal as it relates to each harmful effect by the same method.
- Uk: TS (total score) of DP (design proposal) k for a series of useful functions
| |||||
UminR: TS (total score) of DL (desiring level)
- *minRi: DL (desiring level) of UFi (useful function ) normalized to 1
| |||||
Hk: TS (total score) of DP (design proposal) k for a series of harmful effects
| |||||
HmaxUR: TS (total score) of AL (acceptable level),
- *maxURj: AL (acceptable level) of HEj (harmful effect) normalized to 1
| |||||
Step 6: Estimation of the ideality index of each design proposal
Calculate the ideality index of each design proposal based on the total score of each design proposal to be estimated from Step 5.
IkII (ideality index) of DPk (design proposal)
| |||||
Min I: DL (desiring level) of II (ideality index)
Example: Evaluation of Three Types of Paper Cups for Coffee1,3
Making two types of hierarchy structure diagrams according to Step 1
Two types of hierarchy structure diagrams were drawn that focus on a paper cup for drinking coffee while walking – a mobile paper cup. One hierarchy structure diagram is for useful functions (shown in Figure 9) and another hierarchy structure diagram is for HFs (shown in Figure 10).
| |||||
| |||||
The hierarchy structure diagram for useful functions is the diagram showing the required useful functions of a mobile paper cup. On the other hand, the hierarchy structure diagram for harmful effects is the one showing the predictable side effects.
The objective of making these hierarchy structure diagrams is to evaluate each design proposal (or alternative of a mobile paper cup) rationally at the next step or later (Steps 2 through 6).
Figures 11, 12 and 13 illustrate the three design proposals (alternatives of a mobile paper cup) for evaluating the ideality index:
| |||||
| |||||
| |||||
Estimating weight of each useful function and harmful effect according to Step 2
When estimating the weight of each useful function and harmful effect use the AHP (relative measurement approach).5 In order to use the analytic hierarchy process (AHP) correctly, keep in mind the meaning of weight for useful functions and harmful effects. The weight of each useful function is the level of importance for users (customers). It is the estimated weight of each useful function from a useful function improvement standpoint. The weight of each harmful effect is the reduction of harmful effects; it is the degree of incidence against the users.
A paired comparison of judgments from the AHP is applied by homogeneous elements. For example, a series of useful functions and harmful effects are homogeneous elements of the hierarchy structure diagram for useful functions and harmful effects (shown in Figures 9 and 10). The fundamental scale of absolute values for representing the intensities of judgments is shown in Table 1.
| Table 1:The Fundamental Scale6 | |
| Intensity of Importance | Definition |
| 1 | Equal importance |
| 2 | Weak |
| 3 | Moderate importance |
| 4 | Moderate plus |
| 5 | Strong importance |
| 6 | Strong plus |
| 7 | Very strong or demonstrated importance |
| 8 | Very, very strong |
| 9 | Extreme importance |
| Reciprocals of above | If i has one of the above numbers assigned to it when compared with activity j then j has the reciprocal value when compared with i |
The scale shown in Table 1 has been validated for effectiveness through applications and theoretical comparisons with other scales such as the scales focusing on useful functions and harmful effects in TRIZ.
According to basic theory of the AHP the numbers are used to represent how many times the larger of the two elements dominates the smaller one with respect to a common property or criterion. The smaller element is the reciprocal with respect to the larger one.6
Based on the fundamental scale, the weight of each useful function is shown in Table 2:
| Table 2:Weight of Each Useful Function (UF)of a Mobile Paper Cup | |||||
| Three UsefulFunctions (three items) | Consistency Index | If CI < 0.1, it is OK In this case, CI = 0.02 | |||
| UF1 Easiness to Drink | UF2 Heat-retaining Property (coffee stays hot) | UF3 Insulating the Heat of Hot Coffee | GeometricAverage | UWi (i = 1-3) | |
| UF1 Easiness to Drink | 1.000 | 1.500 | 1.000 | 1.145 | 0.380 |
| UF2 Heat-retaining Property (coffee stays hot) | 0.667 | 1.000 | 1.200 | 0.928 | 0.308 |
| UF3 Insulating the Heat of Hot Coffee | 1.000 | 0.833 | 1.000 | 0.941 | 0.312 |
Total | 3.014 | 1.000 | |||
The estimated weight of each harmful effect on the basis of the same scale is shown in Table 3:
| Table 3:Weight of Each Harmful Effect (HE) of a Mobile Paper Cup | |||||||
| Five HarmfulEffects (five items) | Consistency Index | If CI < 0.1, it is OK In this case, CI = 0.07 | |||||
| HE1 Waterproof Property | HE2 Mobility of Coffee | HE3 Economical Efficiency | HE4 Environmental Policy | HE5 Safety | GeometricAverage | HWj (j = 1-5) | |
| HE1 Waterproof Property | 1.000 | 0.333 | 2.000 | 1.000 | 2.000 | 1.059 | 0.196 |
| HE2 Mobility of Coffee | 3.000 | 1.000 | 3.000 | 2.000 | 1.500 | 1.933 | 0.357 |
| HE3 Economical Efficiency | 0.500 | 0.333 | 1.000 | 1.200 | 0.333 | 0.582 | 0.107 |
| HE4 Environmental Policy | 1.000 | 0.500 | 0.833 | 1.000 | 1.000 | 0.839 | 0.155 |
| HE5 Safety | 0.500 | 0.667 | 3.000 | 1.000 | 1.000 | 1.000 | 0.185 |
Total | 5.414 | 1.000 | |||||
Setting up a desiring level for each useful function and an acceptable level for each harmful effect according to Step 3
The definitions for a desiring level for each useful function and an acceptable level for each harmful effect from the perspective of new product activities in the pipeline are shown in Tables 4 and 5.
| Table 4: Each Desiring Level for Each Useful Function (UF)for the Mobile Paper Cup | |
| Useful Functions | Desiring Level (way to advance “＋”) Concrete Performance Measures |
| UF1: Drink coffee while walking (easiness to drink) | Makes it easy to bring coffee to the mouth with one hand while walking slowly (about 50 m/min). It is expected to be possible to bring coffee (in one hand) while walking, even more quickly (speed is about 90m-100m/min on foot). |
| UF2: Prevent decreasing temperature (heat-retaining property) | It is possible to keep at 80 degrees Celsius for five minutes. It is expected to be possible to keep it hot for a longer time (more than five minutes). |
| UF3: Holding the cup’s body with hot coffee inside for a while (insulating the heat of hot coffee) | It is expected to insulate the heat of hot coffee It is possible to hold the cup’s body even at a high temperature (coffee is more than 90 degrees Celsius) for a while (approximately ten minutes). |
| Table 5: Each Acceptable Level for Each Harmful Effect (HE)for the Mobile Paper Cup | |
| Harmful Effect | Acceptable Level (way to advance to zero) |
| HE1: Prevent leaking coffee between glued connections of the cup’s body (waterproof property) | Leaking some coffee between glued connections of the cup’s body should be realized at less or equal to hundredth part of probability. (quality of paper cup) |
| HE2: Prevent spilling coffee from mouth while drinking (mobility of coffee) | An individual can move coffee (fluid) from the cup to the mouth without spilling coffee from the mouth while drinking. (zero per occurrence ) |
| HE3: Inexpensive coffee cup (economical efficiency) | Cost of each cup is less than five yen. (usual coffee cup) |
| HE4: Environmentally-friendly design (environmental policy) | The ratio of waste material is less than ten percent. (the ratio of reuse is more than 90 percent ) |
| HE5: Prevent dominant hand from burning while drinking (safety) | Prevent dominant hand from burning while drinking. (percentage of risk: less than one percent) |
Estimating weight (degree of relative merit) of each design proposal for each useful function and harmful effect according to Step 4
In order to estimate the effectiveness of the three design proposals (shown in Figures, 11, 12 and 13) and before producing them as possible to a customer, use the modified AHP (including desiring level and acceptable level shown in Tables 4 and 5). The results of the paired comparison table of each design proposal by the AHP including the desiring level include:
| Table 6:Paired Comparison Table Including a Desiring Level of Useful Function 1 of Design Proposals (DPs) | |||||||
| Easiness to Drink | Consistency Index | If CI < 0.1, it is OK In this case, CI = 0.011 | |||||
| DP1 | DP2 | DP3 | minR1 | GeometricAverage | Weight | UW (1, k) K = 1-3 | |
| DP1 | 1.000 | 9.000 | 9.000 | 2.000 | 3.568 | 0.549 | 1.604 |
| DP2 | 0.111 | 1.000 | 1.000 | 0.143 | 0.355 | 0.055 | 0.160 |
| DP3 | 0.111 | 1.000 | 1.000 | 0.143 | 0.355 | 0.055 | 0.160 |
| minR1 | 0.500 | 7.000 | 7.000 | 1.000 | 2.225 | 0.342 | *minR1 = 1.000 |
Total | 6.502 | 1.000 | |||||
Function 1 of Design Proposals (DPs)
| Table 7: Paired Comparison Table Including a Desiring Level of Useful Function 2 of Design Proposals (DPs) | |||||||
| Heat-retaining Property | Consistency Index | If CI < 0.1, it is OK In this case, CI = 0.017 | |||||
| DP1 | DP2 | DP3 | minR2 | GeometricAverage | Weight | UW (2, k) K = 1-3 | |
| DP1 | 1.000 | 0.500 | 0.500 | 2.000 | 0.841 | 0.197 | 1.414 |
| DP2 | 2.000 | 1.000 | 1.000 | 2.000 | 1.414 | 0.332 | 2.378 |
| DP3 | 2.000 | 1.000 | 1.000 | 2.000 | 1.414 | 0.332 | 2.378 |
| minR2 | 0.500 | 0.500 | 0.500 | 1.000 | 0.595 | 0.139 | *minR2 = 1.000 |
Total | 4.264 | 1.000 | |||||
The following tables by the AHP include harmful effects:
| Table 8:Paired Comparison Table Including a Desiring Level of Useful Function 3 of Design Proposals (DPs) | |||||||
| Insulating the Heat of Hot Coffee | Consistency Index | If CI < 0.1, it is OK In this case, CI = 0.081 | |||||
| DP1 | DP2 | DP3 | minR3 | GeometricAverage | Weight | UW (3, k) K = 1-3 | |
| DP1 | 1.000 | 0.500 | 5.000 | 2.000 | 1.495 | 0.300 | 1.093 |
| DP2 | 2.000 | 1.000 | 6.000 | 1.000 | 1.861 | 0.373 | 1.361 |
| DP3 | 0.200 | 0.167 | 1.000 | 0.143 | 0.263 | 0.053 | 0.192 |
| minR3 | 0.500 | 1.000 | 7.000 | 1.000 | 1.368 | 0.274 | *minR3 = 1.000 |
Total | 4.987 | 1.000 | |||||
| Table 9:Paired Comparison Table Including an Acceptable Level of the Harmful Effect 1 of Design Proposals (DPs) | |||||||
| Waterproof Property | Consistency Index | If CI < 0.1, it is OK In this case, CI = 0.003 | |||||
| DP1 | DP2 | DP3 | maxUR1 | GeometricAverage | Weight | EW (1, k) K = 1-3 | |
| DP1 | 1.000 | 0.333 | 0.333 | 0.500 | 0.485 | 0.109 | 0.577 |
| DP2 | 3.000 | 1.000 | 1.000 | 2.000 | 1.565 | 0.351 | 1.861 |
| DP3 | 3.000 | 1.000 | 1.000 | 2.000 | 1.565 | 0.351 | 1.861 |
| maxUR1 | 2.000 | 0.500 | 0.500 | 1.000 | 0.841 | 0.189 | *maxUR1 = 1.000 |
Total | 4.457 | 1.000 | |||||
| Table 10:Paired Comparison Table Including an Acceptable Level of the Harmful Effect 2 of Design Proposals (DPs) | |||||||
| Mobility of Coffee | Consistency Index | If CI < 0.1, it is OK In this case, CI = 0.088 | |||||
| DP1 | DP2 | DP3 | maxUR2 | GeometricAverage | Weight | EW (2, k) K = 1-3 | |
| DP1 | 1.000 | 0.125 | 0.167 | 1.000 | 0.380 | 0.059 | 0.874 |
| DP2 | 8.000 | 1.000 | 5.000 | 7.000 | 4.091 | 0.641 | 9.410 |
| DP3 | 6.000 | 0.200 | 1.000 | 4.000 | 1.480 | 0.232 | 3.405 |
| maxUR2 | 1.000 | 0.143 | 0.250 | 1.000 | 0.435 | 0.068 | *maxUR2 = 1.000 |
Total | 6.385 | 1.000 | |||||
| Table 11: Paired Comparison Table Including an Acceptable Level of the Harmful Effect 3 of Design Proposals (DPs) | |||||||
| Economical Efficiency | Consistency Index | If CI < 0.1, it is OK In this case, CI = 0.013 | |||||
| DP1 | DP2 | DP3 | maxUR3 | GeometricAverage | Weight | EW (3, k) K = 1-3 | |
| DP1 | 1.000 | 2.000 | 3.000 | 1.000 | 1.565 | 0.359 | 1.107 |
| DP2 | 0.500 | 1.000 | 2.000 | 0.500 | 0.841 | 0.193 | 0.595 |
| DP3 | 0.333 | 0.500 | 1.000 | 0.500 | 0.537 | 0.123 | 0.380 |
| maxUR3 | 1.000 | 2.000 | 2.000 | 1.000 | 1.414 | 0.325 | *maxUR3 = 1.000 |
Total | 4.357 | 1.000 | |||||
| Table 12:Paired Comparison Table Including an Acceptable Level of the Harmful Effect 4 of Design Proposals (DPs) | |||||||
| Environmental Policy | Consistency Index | If CI < 0.1, it is OK In this case, CI = 0.020 | |||||
| DP1 | DP2 | DP3 | maxUR4 | GeometricAverage | Weight | EW (4, k) K = 1-3 | |
| DP1 | 1.000 | 1.000 | 2.000 | 1.000 | 1.189 | 0.286 | 0.841 |
| DP2 | 1.000 | 1.000 | 1.000 | 0.500 | 0.841 | 0.203 | 0.595 |
| DP3 | 0.500 | 1.000 | 1.000 | 0.500 | 0.707 | 0.170 | 0.500 |
| maxUR4 | 1.000 | 2.000 | 2.000 | 1.000 | 1.414 | 0.341 | *maxUR4 = 1.000 |
Total | 4.151 | 1.000 | |||||
| Table 13:Paired Comparison Table Including an Acceptable Level of the Harmful Effect 5 of Design Proposals (DPs) | |||||||
| Safety | Consistency Index | If CI < 0.1, it is OK In this case, CI = 0.038 | |||||
| DP1 | DP2 | DP3 | maxUR5 | GeometricAverage | Weight | EW (5, k) K = 1-3 | |
| DP1 | 1.000 | 0.167 | 0.200 | 1.000 | 0.427 | 0.081 | 0.855 |
| DP2 | 6.000 | 1.000 | 0.500 | 4.000 | 1.861 | 0.351 | 3.722 |
| DP3 | 5.000 | 2.000 | 1.000 | 4.000 | 2.515 | 0.474 | 5.030 |
| maxUR5 | 1.000 | 0.250 | 0.250 | 1.000 | 0.500 | 0.094 | *maxUR5 = 1.000 |
Total | 5.303 | 1.000 | |||||
Keep in mind the five tables are based on a series of harmful effects. When using the AHP including the acceptable level an individual will have to consider that the more the design proposal weighs, the more harmful effects decrease and the value of each design proposal increases. This is because the direction of an acceptable level advances to zero (shown in Table 5). The weight of each harmful effect, therefore, is reciprocal in value compared to the one of each useful function.
Estimating total score of each design proposal for a series of useful functions and harmful effects according to Step 5
Table 14 shows the total score of each design proposal for a series of useful functions:
| Table 14: Total Score of Each Design Proposal for a Series of Useful Functions | ||||
| UF1 Easiness to Drink | UF2 Heat-retaining Property (coffee stays hot) | UF3 Insulating the Heat of Hot Coffee | Total Score (TS) | |
| Weight of UFi | 0.380 | 0.308 | 0.312 | 1.000 |
| *minRi | 1.000 | 1.000 | 1.000 | 1.000 |
| DP1: U1 | 1.604 | 1.414 | 1.093 | 1.386 |
| DP2: U2 | 0.160 | 2.378 | 1.361 | 1.218 |
| DP3: U3 | 0.160 | 2.378 | 0.192 | 0.853 |
After that, calculate the total score of each design proposal for a series of harmful effects by the same method shown in Table 15.
| Table 15:Total Score of Each Design Proposal for a Series of Harmful Effects | ||||||
| HE1 Waterproof Property | HE2 Mobility of Coffee | HE3 Economical Efficiency | HE4 Environmental Policy | HE5 Safety | Total Score (TS) | |
| Weight of HEj | 0.196 | 0.357 | 0.107 | 0.155 | 0.185 | 1.000 |
| *maxURj | 1.000 | 1.000 | 1.000 | 1.000 | 1.000 | 1.000 |
| DP1: H1 | 0.577 | 0.874 | 1.107 | 0.841 | 0.855 | 0.832 |
| DP2: H2 | 1.861 | 9.410 | 0.595 | 0.595 | 3.722 | 4.568 |
| DP3: H3 | 1.861 | 3.405 | 0.380 | 0.500 | 5.030 | 2.627 |
Calculating ideality index of each design proposal according to Step 5
Based on the total score of each design proposal for a series of useful functions and harmful effects, calculate the ideality index:
| Table 16:The Ideality Index of Each Design Proposal (DP) | |||
| ΣUF (useful function) Total Score (TS) | ΣHE (harmful effect) Total Score (TS) | Ideality Index | |
| Min I | 1.000 | 1.000 | 1.000 |
| DP1: I1 | 1.386 | 0.832 | 1.665 |
| DP2: I2 | 1.218 | 4.568 | 0.267 |
| DP3: I3 | 0.853 | 2.627 | 0.325 |
Conclusion
The result of Table 15 shows that design proposal 1 is expected to be the most attractive mobile paper cup for users. Consumers (or users) using consumable goods like paper cups decide what paper they like. An engineer decides on new product activities such as improving ideality since they have to chase the IFR.
The importance of the ideality index calculated through the proposed decision-making model is for engineers (with new product activities) and TRIZ practitioners to use this model. It is possible to predict the most valuable design proposal (alternative) before proceeding to the production stage. The most valuable design proposal (based on TRIZ-oriented thinking as it corresponds to the highest ideality index is more than one (it means desiring level of ideality index). It could possibly solve the serious contradictions before they appear. The solo lid and insulating sleeve shown in Figure 11 is a highly-valued idea for solving contradictions for a usual paper cup under the condition of walking while drinking coffee.3 Practitioners of TRIZ should use the proposed decision-making model when predicting the most valuable design proposals in TRIZ activities.
References
- Manabu Sawaguchi, On the Roles of TRIZ at the Workshop Focusing on Innovation Based on Cooperation Between Various Industries, Proceedings of The Tenth Annual Conference of the Altshuller Institute for TRIZ Studies, Kent State University, April 13-15, 2008.
- Stan Kaplan, An Introduction to TRIZ the Russian Theory of Inventive Problem Solving, Ideation Inc., 1996.
- Starbucks corporate social responsibility annual report, 2006.
- Thomas L.Saaty, The Analytic Hierarchy Process, RWS Publications, 1992.
- Thomas L.Saaty, The Analytic Network Process, RWS Publications, 1996.
- Thomas L.Saaty, The Analytic Network Process, RWS Publications, pp. 23-25, 1996.
- Vladimir Petrov, Progress and Ideality, TheTRIZ Journal, 2006.
Note: This paper was originally presented at The Altshuller Institute’s TRIZCON2009.