We apply this method to evaluate functions and costs and propose an optimizing process for improvement design so that the next target for improvement can be ...
Available online at www.sciencedirect.com
ScienceDirect Procedia Engineering 131 (2015) 569 – 576
World Conference: TRIZ FUTURE, TF 2011-2014
Optimizing Process for Improvement Design Using TRIZ and the Information Integration Method Heikan Izumia*, Manabu Sawaguchib a
IZUMI Products Companny,3039 Sasaga Matsumoto Nagano 399-8721,Japan b WASEDA University, 3-4-1 Okubo Shinjyuku Tokyo 169-8555, Japan
Abstract Functional improvements and cost reductions are requested in products’ improvement design during its maturity stage from competitions in markets. Many engineers face several difficulties in satisfying both these expectations simultaneously. We propose a design process called Improvement Design Process (IDP) that compares different types of features by using TRIZ and the Information Integration Method (IIM). IIM is an evaluation method, based on Shannon’s information theory, that lets designers compare different types of features by using a common measure of “information.” We apply this method to evaluate functions and costs and propose an optimizing process for improvement design so that the next target for improvement can be automatically selected without the requirement of any subjective inputs. We confirmed IDP’s effectiveness in a case study involving an electric shaver’s improvement design. by Elsevier Ltd.by This is an open © 2015 2015 Published The Authors. Published Elsevier Ltd.access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the Scientific Committee of TFC 2011, TFC 2012, TFC 2013 and TFC 2014 – GIC. Peer-review under responsibility of the Scientific Committee of TFC 2011, TFC 2012, TFC 2013 and TFC 2014 – GIC
Keywords: Development Management Technology, Improvement Design Process, Information Integration Method;
List of abbreviations IIM
Information Integration Method
IDP
Improvement Design Process
1. Introduction Staying competitive in today’s markets requires making lifelong product improvements, even during the product’s maturity stage [1]. Not only functional improvements but also cost reductions are required from the market
1877-7058 © 2015 Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the Scientific Committee of TFC 2011, TFC 2012, TFC 2013 and TFC 2014 – GIC
doi:10.1016/j.proeng.2015.12.451
570
Heikan Izumi and Manabu Sawaguchi / Procedia Engineering 131 (2015) 569 – 576
competition, which are difficult for engineers to simultaneously satisfy. Many engineers develop products through trial and error, thereby trying to strike a balance between the different types of features. However, this trial-and-error process leads to redundancy of the design and makes it difficult to compare features that are evaluated using different measures [2] [3]. In this study, we propose a new design process called the Improvement Design Process (IDP) that attempts at realizing different types of features by using TRIZ and the Information Integration Method (IIM). On the basis of the concept of Shannon’s information theory, IIM, an evaluation method, lets designers compare different types of features with a common measure [4]. We apply IIM to our method to evaluate functions and costs so that the next improvement can be automatically selected without the requirement of any subjective inputs. We illustrated IDP’s effectiveness using a case study involving an electric shaver’s improvement design. This study borrows the practical design processes from our previous studies: “Practical Cost Reduction Methods based on TRIZ”[5] and “Functional Improvement Methods based on Expert Engineers’ Thinking Way”[6]. 2. Background of study Product development is categorized into three main steps: product planning, structuring design, and detailed design[7]. In this study, we focus on an optimizing process for improvement in structuring design . Relations between this study and other current studies are described in Figure 1.
Fig.1Relations between this study and current studies
Following are the issues faced during improvement in conceptual design: (1) Evaluation method Although many conventional evaluation methods—such as the rader chart, point rating, or the elimination method—have been used to evaluate items requiring different measurements, such methods are usually dependent on subjective inputs. The Analytic Hierarchy Process (AHP) [8] reduces subjectivity by numerically presenting humans’ feelings. However, AHP is costly and time consuming. The concept of value (functions versus costs) is used in several cases [9], but it is still difficult to compare designs using different measures, such as performance and cost. In this study, we apply IIM to compare these different types of features. (2) Procedure in IDP In IDP, cost reduction and some of functional improvements are requested simultaneously. Engineers need to keep balance of these features. Unfortunately in some cases, improvement designs fall into repeating developments as there is no clear design procedure to satisfy all of them.(Figure 2) We also propose an optimizing process for improvement design based on IIM. By comparing evaluation items in this method, next target for improvement can be automatically selected without personal subjectivity.
571
Heikan Izumi and Manabu Sawaguchi / Procedia Engineering 131 (2015) 569 – 576
For the next age of innovation, new tools allowing computer-aided artifact creation are anticipated [10] and in recent studies, some of such formalization approaches for computer-aided tools are proposed [11][12]. We are assuming that this optimizing process using TRIZ and IIM is also one of such approaches which are measurable and tangible. 3. Optimizing process 3.1. Information Integration Method (IIM ) IIM helps evaluate features requested for system optimization. In this method, all features are evaluated using a common measure called Information based on Shannon’s information theory. Information is defined in following mathematic formulas. Information (I) for communicating the status of feature a, which is associated with probability Pa, is given as follows:
I
ln
1 Pa
(1)
IIM expands this concept to measure the difficulties (Information, energy, or effort) required to satisfy the requested features in products design. The smaller the probability Pa, the more difficult it is to satisfy feature a. In IIM, the System Range is defined as the range of a feature in a product (or system); Design Range is defined as the range of a feature requested from markets or customers; and Common Range is where the System Range and Design Range overlap (Figure 2) .
Fig.2 Probability distribution of system parameter (1)
Fig.3 Probability distribution of system parameter (2)
A higher Design Range probability density is indicative of a higher level of satisfaction with the requested product feature. In IIM, Information (I) to obtain the Common Range in probability Pc is used as a measure for evaluating the design. It is assumed that the probability density of most product features can be approximated for uniform distribution. Information (I) can be defined as follows if the probability density is subject to continuous uniform distribution (Figure 3):
I
ln
1 Pc
ln
SystemRange(l 1) CommonRange(l2 )
(2)
572
Heikan Izumi and Manabu Sawaguchi / Procedia Engineering 131 (2015) 569 – 576
In IIM, Information (I) can be infinite when it is out of Design Range. Even if other features are satisfied with requested features, the design will not be selected if one of the features falls outside the Design Range. To address such situations, Common Range Coefficient k is proposed in IIM. As described in Figure 4, the Design Range falls between feature parameters a and b, and the feature parameter that should be satisfied is c. In this case, Common Range Coefficient k = 1, where the system feature parameter is between a and b. Common Range Coefficient k = 0 when the System Range parameter is between 0 and c. Common Range Coefficient k is described in equation 3 as follows, in which the System Range parameter is between a and c, assuming there is a small System Range with width w:
Fig.4 Common range coefficient
I
ln
w kw
ln
1 k
(3)
(4)
Total Information, IT, is the amount of features’ Information. IT is described as
IT
¦I
i
(i=1…n)
where the Information of feature i is Ii, and the number of features is n. In this study, IT is used as a common evaluation measure of the improved design. IIM has been used with the Nakazawa method in combination with the experimental design[13]. 3.2. Optimizing Procedures Using IIM, we optimized our IDP by studying the improvement design and reducing it to total IT. To further optimize the improvement design, this work needed repetition until the required specification was satisfied (total IT is zero), or no more improvements could be found. The following algorithm describes this optimizing process (Figure5):
Calculate total IT based on current design
Find improvement design to reduce total IT
Fig.5 Optimizing procedure for IDP
Repeat until total IT is zero or no more improvement design can be found
573
Heikan Izumi and Manabu Sawaguchi / Procedia Engineering 131 (2015) 569 – 576
4. Case Study We applied the IDP to improve the flexible head motion of an electric shaver (Figure 6). In current electric shavers, a higher level of flexibility for the shaver head is desired to better conform to the face while shaving, but without increasing the cost or noise level. Therefore, we used IIM to evaluate the selected features: noise level, cost, and flexibility. Although there are more electric shaver features we could evaluate, here we focus on these three because they directly relate to flexibility motion. Table 1 shows the current and requested status of each feature. Each feature value is described as a percentage, and the current value is 100. Flexibility motion is the total degree of shaver head motion.
Fig.6 An electric shaver Table 1 Requested specifications for improvement design
Parameters System range Design range Common range probability p Information value I
Cost 100 105 or less 1.00 0.00
Evaluation items Noise level Flexibility motion 99–101 98–102 101 or less 110 or more 1.00 0.00 0.00
Total( IT)
Table 2 Information value of current design
Requested status
Evaluation items Cost
Noise level
Flexibility motion
Current feature (System range)
100
99–101
98–102
Expected feature (Design range)
105 or less
101 or less
110 or more
110
105
110
Absolute feature
Table 2 shows the Information value of the current status. For each Information value, Information of flexibility motion is infinite. Therefore, it is selected as a study target. We applied TRIZ to get ideas regarding how to increase the shaver head’s flexibility. From the TRIZ contradiction matrix, principle “1 Segmentation” was selected by assuming that “3 Length (angle) of movement object” was a feature offering improvement and that the feature “36 Device complexity” was not. Then, improved design 1 was found by dividing the shaver head and the shaver body, as described in Figure 7.
574
Heikan Izumi and Manabu Sawaguchi / Procedia Engineering 131 (2015) 569 – 576
Fig.7 Improvement design
After each feature, value based on improvement design 1 was checked by calculating the Information value. Table 3 shows the resulting Information value based on improvement design 1. Table 3 Information value of improvement design 1
Parameters System range (or d ) Design range
Evaluation items Cost 106 105 or less
Noise level
Flexibility motion
104
110
101or less
110 or more
Total( IT) -
Common range probability p
0.00
000
1.00
-
Absolute feature c
110
105
-
-
Common range coefficient k
0.8
0.25
-
-
Information value I
0.22
1.39
0.00
1.61
As total IT is still over 0 in improvement design 1, the noise level feature was selected next because the noise level increased when the shaver head and shave body were divided. From our previous study[6], we applied improvement design 2, described in Figure 8, to reduce the noise level.
Fig.8 Improvement design 2 for noise reduction
Table 4 shows the result of new Information value based on improvement design 2.
575
Heikan Izumi and Manabu Sawaguchi / Procedia Engineering 131 (2015) 569 – 576 Table 4 Information value of improvement designs 1and 2
Parameters System range (or d ) Design range Common range probability p
Evaluation items Noise level 102
105 or less
101 or less
110 or more
-
0.00
1.00
-
0.00
Flexibility motion 110
Total( IT)
Cost 107
-
Absolute feature c
110
105
-
-
Common range coefficient k
0.60
0.75
-
-
Information value I
0.51
0.29
0.00
0.80
Fig.9 Improved designĹ for cost reduction
As total IT was still over 0 in improvement designs 1and 2, cost feature was selected next because adding some parts for improved design 2 increased the cost. From the TRIZ technical contradiction matrix, principle “5 Merging” was selected to solve this technical problem by assuming “32 Easy manufacture” was an improved feature and “33 Easy of operation” was a worsening feature. Improved design 3 was found based on “5 Merging” (Figure 9). Table 5 shows the result of new Information value based on improvement design 1, 2, and 3. Table 5. Information value of improvement design 1, 2 and 3
Parameters System range (or d )
Evaluation items Cost 105 105 or less
Noise level 102
Flexibility motion 110
Total( IT) -
101or less
110 or more
-
100
0.00
1.00
-
Absolute feature c
-
105
-
-
Common range coefficient k
-
0.75
-
-
Information value I
0.00
0.29
0.00
0.29
Design range Common range probability p
Because total IT is still over 0 in improvement design 1, 2 and 3, the noise level feature was selected next. However, no more improvements could be found, so improvement design 1, 2, and 3 were selected as the optimized design.
576
Heikan Izumi and Manabu Sawaguchi / Procedia Engineering 131 (2015) 569 – 576
5. Conclusion and next activities Using IDP, we obtained a satisfactory degree of flexibility for the shaver head with a quicker development schedule than what is normally obtained. We assume this is because of the following reasons: (1) By using IIM for evaluation, let us compare multiple features with different evaluation measures using a common measure: “Information.” (2) Comparing each Information value at every step of the improvement design helped us select the next improvement feature (3) Visualizing the level of improvement can help engineers shorten the development schedule and clarify the target of the optimized improvement design. In the future, we plan to (1) Apply this IDP to more feasibility studies and update it to make it more practical, and (2) Apply it to new product developments, and not just improvement designs. Acknowledgements The authors would like to thank Hiromu Nakazawa for helpful advices about Information Integration Method. The authors would like to thank Enago for the English language review. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13]
Hiroyuki Ogino: “Over the cost reductions”, Nikkei Monozukuri (Japanese), June 2008, (2008), pp.50-79. Tsutomu Nakayama, “Seeking value without compromising”, Nikkei Monozukuri (Japanese), December 2009, (2009), pp.38-66. Tsutomu Nakayama, “Over-quality more than 50%” , Nkkei-Monozukuri (Japanese), September 2011, (2011), pp.83-86. Hiromu Nakazawa, Tetsuro Ohtsubo: “Decision System of Optimum Grinding Conditions Based Information Integration Method”, Journal of Japan Society for Precision Engineering, Vol.57-5 (1991), pp.863-868. Heikan Izumi, Manabu Sawaguchi, “Practical Cost Reduction Methods based on TRIZ”, Journal of Japan Association for Management Systems, Vol.29, No.2 (2012), pp.95-104. Heikan Izumi, Toshihiko Ohara, Manabu Sawaguchi: “Functional Improvement Methods based on Expert Engineers’ Thinking Way”, Proceedings of Japan Association for Management Systems Vol.48 (2012), pp.126-129. Katsufumi Araki, Katsuya Terashima, Makoto Senoo, Jun Kanie, “Product and Process Modeling for Product Development”, The Japan Society of Mechanical Engineering, Proceeding of 20th Design and system department 2010 (20), pp."2314-1"-"2314-5" Thomas L. Saaty, “Decision making with the analytic hierarchy process”, Int. J. Services Sciences, Vol.1 No.1 (2008), pp.83-98. Manabu Sawaguchi: “VE and TRIZ”, Dougakkan (2002) D. Cavallucci: “A research agenda for computing developments associated with innovation pipelines”, 2011 Computers in Industry 62 (4) , pp. 377-383 F.Rousselot , C.Zanni-Merk , D. Cavallucci :” Towards a formal definition of contradiction in inventive design”, 2012 Computers in Industry 63 (3) , pp. 231-242 S. Ikovenko : “TRIZ and computer aided inventing”, Building the Information Society156 (2004) pp.475–485 Hiromu Nakazawa: “Monozukuri no Kirifuda Nakazwa Method (Japanese)”, (2011), Union of Japanese Scientists and Engineers
