Advance Design Theories

8/6/2019 Advance Design Theories

1/16

1

Critical Analysis of DesignTheoriesP.S. Pad liya1, Y. Naikvade2, D.D. Ghosh3

Department of Mechanical and Aerospace Engineering, State University of New York at Buffalo,

USA

Abstract

Traditional design processes have been described to follow an Algorithmic approachwherein a set of systematic sequence of steps is to be followed. Dr. Nam Suhs conceptcreated a new school of design approach Axiomatic approach, which has the foundation on

the premise that there are generalised principles which govern the underlying behaviour of the

system. In this paper we have given an introduction of the axiomatic design process andother methodologies Pahl and Beitz Systematic Design, Pughs Total Design and DecisionBased Design., This will be followed by the critical analysis of the afore mentioned methods to

find out shortcomings (if any), ways to mitigate them and finally attempt to adapt thestrengths of these methods along with other techniques to develop a more robust method ofdesign.

Key words axiomatic design, systematic approach, total design, and decision based design

Authors to whom all correspondence should be addressed1 Person Number [37410540], Email [email protected] Person Number [37351725], Email [email protected]

Person Number [37420971], Email [email protected]

Introduction

There are two ways to approach design:-

Algorithmic: Prescribes the proper design process. Axiomatic: Provides generalised principles that govern the behaviour.

Axiomatic design elevates engineering design to a science, governed by a few basic rules, and

from what has been an art integrated with engineering analysis. Axiomatic design approach is

based on the interplay between what we want to achieve and how we choose to achieve it

[1].

This paper can be divided in to five sections. The first section explains the Dr. Suhs Axiomatic

Design in detail. However, due to the limitations on the scope of this paper only the basic

procedure has been focused upon. Section 2 explains the framework of other methodologies

briefly. Section 3 presents the comparison of the methodologies. In this section we have also

attempted to highlight the salient features of each method that we agree with, and point out

the lacunae in these methods. The solutions of these lacunae have been touched upon. Section

4 talks about the conclusion that we have drawn after studying the various methods i.e. the

design methodologies are not competing in nature instead, if integrated together they would

lead to better results. This section thus proposes a better strategy to design. It should be


2/16

2

noted that even after integrating the four techniques, some of the lacunae still persist. Hence

the adoption of some other powerful techniques has been advised. Section 5 summarizes the

findings and presents the conclusions.

1.Overview of Suhs Axiomatic Approach1.1 Preliminary concepts of Axiomatic DesignAxiomatic approach consists of the following concepts [1]:

Design world consists of distinct domains: Consumer, Functional, Physical and Process The design process involves mapping between the domains Each domain is defined (or characterised) by a characteristic vector which can be

decomposed by zig - zagging between the domains. The characteristic vectors

associated with each domain are:

i) Consumer Attributesii) Functional Requirementsiii)Design Parametersiv)Process Variables

The mapping process involves creative conceptualization which must satisfy the designaxioms

Figure 1 shows the vectors associated with each domain along with the mapping

Figure 1 Domains in Axiomatic Design and Mapping [1]

The characteristics of the mappings are as follows:

Consumer to Functional domain mapping is an interpretation of the needs ofcustomers which result in the formulation of the objective functions

Functional to Physical domain mapping is interpreted as core design activity. Thedesign parameter will represent the concept of the design

Physical to Process domain mapping describes the process by which the requireddesign parameter can be obtained

Mapping between each domain must satisfy the design axioms Independence Axiom and

Information Axiom.


3/16

3

1.2 Design AxiomsAxioms are fundamental truths that are always observed to be valid and for which there

are no counter examples or exceptions. According to axiomatic design the process of

mapping should conform with the following two axioms [1]:

Independence Axiom: During the mapping process one to one mapping betweenthe attributes of two domains should be maintained.

Information Axiom: Minimize the information content of the design, that is selectthe simplest product.

The applications of the two axioms are pre-emptive in nature.

1.3 Domain mapping process [1][2]As per the independence axiom there should be one to one mapping between the attributes

of two domains. We can describe each domain in the form of attribute vector. Figure 2

shows the mapping process.

Figure 2 Mapping Process [1]

Let [FR] = [FR1 FR2 FR3]T

and [DP] = [DP1 DP2 DP3]T

be the attribute vector of functionalrequirements and design parameters respectively and x denote dependence of attribute

in one domain on the attribute of other domain while 0 denote no dependence. Then

according to Independence theorem the mapping relation can be described by the

following matrix relationship

[FR] = [A] [DP]

1 2 3 0 00 00 0

123Here [A] is the design matrix. Similar mapping process exists between design parameter

matrix [DP] and process domain [PV], where [B] is the design matrix.

Depending on the type of design matrix systems can be classified as follows [2]:

Decoupled system Partial decoupled system Coupled system

Figure 3 shows the three systems depending on the type of design matrix


4/16

4

Figure 3 Mapping Process

As per the independence axiom a good design will always have case (a) as the design

matrix, but complex systems may have case (b) although it is not desirable. Case (b)

shows that the solution set depend on the order in which it is solved. The design matrix is

formulated using the decomposition process.

1.4 Decomposition ProcessThe highest level of design equation is called Design Intent. Decomposition is a process of

transforming Design Intent into Realizable design details [3]. This process involves zig-

zagging from what domain to how domain until FR is satisfied. At each level, one DP is

selected to satisfy one FR. Subsequently, that DP imposes a constraint on the next level

down. The process stops when the next level is obvious. This level is known as the Leaf

[3].

Figure 4 Decomposition process

During the processes of domain mapping and decomposition the independence axiom

should always be satisfied. Once we get alternate design we apply information axiom to

select the best design.

1.5 Independence Axiom [1]The information axiom provides a method of quantifying the best design out of all the

possible choices that satisfy the independence axiom. This axiom compares the design

range stated by the FR to the system design range which is dependent on the DP meeting


5/16

5

the requirement. As stated in [4], the information content I for an uncoupled design with n

FRs can be expressed as: log where pi is the probability of DPi satisfying FRi.The probability of success can be computed by specifying the Design Range for the FR and

by determining the System Range that the proposed design can provide to satisfy the FR.

This is shown in Figure 5. The overlapping region called as the Common Range is the only

region where the design requirements are satisfied. The information content can then be

expressed as I = log (Asr/Acr) where Asr denotes the area under the System Range and Acr

is the area under Common Range. As Asr = 1 in most cases, log

Figure 5 Design Range, System Range and commonRange (Information Axiom) [1]

2.Overview of other design processes2.1 Overview of Pahl and Beitz Systematic DesignPahl and Beitz [5] developed a systematic function to form method of engineering design.

In this method the overall function is broken down into sub-functions. Individual solutions

are then obtained for each sub-function. These individual solutions are then combined toachieve overall objective.

Figure 6 Pahl and Beitz problem solution

decomposition structure [5]Figure 7 Pahl and Beitz model of the design

process [6]


6/16

6

2.2 Overview of Pughs Total DesignStuart Pugh took design one step further by introducing the theory of Total Design.

According to Stuart Pugh, design is the integration of the two cultures Arts and Sciences.

He states that Design is not like Mathematics or Physics; it does not represent a body of

knowledge; it is the activity that integrates the bodies of knowledge present in Arts,

Sciences and their Derivatives [6] and further quotes that ... it is only the balance and

distribution of arts and science contents which distinguishes one from another [6]. Stuart

Pugh understood the importance of Total Design and sensed the need to integrate

academia and industry. Pughs concept of total design states that marketing,

manufacturing, finance, research are all part of the design process and every aspect should

be considered in design.

Figure 8 Pughs Activity Model of Total Design [6]

2.3 Overview of Decision Based Design (DBD)A relatively new development in the field of design is the emergence of Decision Based

Design (DBD). DBD focussed on the primary aim of any artefact producing firm: to make

profit. DBD has the following two activities at its core [7]: Determine all possible options Choose the best option

The framework of DBD, proposed by Hazelrigg [7] is based on the utility theory. DBD cares

about uncertainty and risk in the design process.

Figure 9 Decision Based Design system framework [7]


7/16

7

3.Comparison and Constructive criticism of Design methodologiesIf we closely look at Systematic Approach design model we can compare it to a waterfall

model, where the result is obtained in a systematic manner in a top down approach,

although iteration is allowed in the process but it is not recommended.

3.1 Salient features of Systematic ApproachSome of the key basic concepts which can be universally adopted as per our opinion and

stated by Pahl and Beitz are [5]:

Design is about converting three things :- Energy, Materials, Signals Types of designs :- Original, Adaptive, Variant Design methodology should reflect the findings of cognitive psychology and

ergonomics. Although we do consider that Systematic Approach has not explained

the technique to integrate these in detail.

3.2 Lacunae of Systematic Approach

Some of the lacunae which we would like to point are: Manufacturing considerations not taken into account Lack of using prototyping method to test the design Absence of the explanation of bottom up approach to integrate the solution Absence of mechanism of choosing between alternate designs Impact of design processes on other engineering activities and activities pertaining

to the organization not being taken into account.

Since the core for the Systematic approach is the function to form matchup, thecriteria of a successful design is that the form and structure of the final design

realises the requirements of the function. The method doesnt take into

consideration how the form looks like and the process to obtain the form in

physical. The evaluation system is also inadequate and qualitative in nature.

Systematic approach has a very strict design algorithm which can reduce thecreativity involved and lead to stagnancy.

The system is inflexible and not good for drastic changes.3.3 Salient feature of Total Design MethodTotal design method however takes into account many other activities and is an integration

of people, products and organisations. Being a Total approach, the model could overcome

many of the lacunae of Systematic model. The major impact we consider is the use of the

evaluation matrix to compare and evolve with new design solutions while in Systematic

approach it was based on economic and technical criteria, VDI guidelines [5] and never

involved group discussions to evaluate the alternatives. Considering the Total Design model

many designs which were engineering marvels and results of the Systematic approach

were actually a failure if we evaluate according the Total Design approach. The best

example of this can be the Sinclair C5 vehicle model [6].

3.4 Lacunae in Total Design MethodTotal design also had some lacunae, some of them as per our opinion are:

Being a Total design approach and taking the larger picture into consideration, themodel is too vast to meet deadlines which form a very critical part of market

currently.


8/16

8

The model heavily depends on group discussions and qualitative comparisons tocome to a conclusion on choosing a particular concept that can be carried forward.

Studies in economics and Decision Field theory have shown that a number ofirrationalities come into play during the process of group decision making.

Although Total design gives a framework for design, it lacks in defining the clarityof steps and how to transition between the steps. Explanation of information

transfer between the steps has not been provided clearly.

The approach gives the designer an insight into various spheres which he/sheneeds to consider but the approach speaks little about how to execute the steps.

Because of the dynamics within the group many advantages of working in groupsare often lost.

Groups with same structure may come up with different styles and produce adifferent output which shows the qualitative attribute of evaluation method.

However we do observe potential advantage in this, as scope of exploration and

evaluation increases and better design solutions can be obtained.

The selection of concept as a datum for evaluation is one of the most critical stepsand is also a very difficult step.

3.5 Salient features of Axiomatic ApproachFrom the above analysis it can be clearly seen that a need of a quantitative evaluation

procedure was needed in the design activity. Axiomatic approach fulfils this requirement to

some extent. Some of the salient features of Axiomatic approach over the above two

according to us are:

Above methods can be considered as a confluence of art and design principleswhere art has an analogy with creativity and design principles corresponds to the

modelling of the components. Need of a scientific and mathematical approach to

design was fulfilled by Axiomatic Design [3].

Design practices prior to Axiomatic approach were based on empirical relationsrelying on trial & error and heuristics and hence time consuming [3].

The algorithmic approach including the above two methods were difficult to applyat conceptual level and were more suitable for level of detailed design and were

less effective if many functional requirements were to be satisfied [3].

It can be clearly seen that Axiomatic method was influenced by the concept of Totaldesign as it took into consideration the people factor and organisation factor in the

process of product creation and did not just focus on the design activity.

The two axioms of Axiomatic approach proved to be a great tool in evaluatingdesign solutions in a quantitative manner.

In the above two methods integration of the sub systems has not been explicitlymentioned which has been taken care of in the Axiomatic Design as shown in

Figure 10 [8].

Axiomatic Approach is scalable and can be used for Flexible systems as well [3]. Since Axiomatic Design is modular and each module corresponds to independent

functional requirement, it provides a very clear demarcation of each functional

requirement. So axiomatic design is very suitable for customisation [3]. For

example in case of laptops, where the end products remains the same i.e. a laptop,

but the configuration of the laptop can be different as per the customers

requirement. So, if a product is to be designed with some FRs different, axiomatic

design addresses this design and integration very well .


9/16

9

Figure 10 V model approach of Axiomatic Design [8]

Every product according to product life cycle graph will face a stagnancy phase asshown in Figure 11 and there is a need of revitalization.

Figure 11 P roduct Life Cycle Graph

The Systematic approach and the Total design approach have not explicitly

mentioned about revitalization which is important from an organisations point of

view as no organisation would like to see it product facing the decline phase. Suh

provided a mechanism wherein he states that if there is scope of providing

additional functionality in a product and is not needed currently, then do not

provide it and make it available later [3]. This approach is adopted by product

families. Example of this is Gillette Razor where the organisation initially came up

with Single blade razor, reinvigorated the sales by introducing Twin blade razor,

continued the trend with Triple blade razor and currently has come up with Fusion

Proglide Power razor which is made for intricate parts although all are part of the

same family of Razors. Although Systematic Design states the use of electronic media, it cannot be

implemented in an automated way. But electronic platforms can be used to assist

the methodologies like creating CAD models. The same goes for Total design model

as well, but Axiomatic design being mathematical in nature has been described in a

logical manner and implemented in a software framework. The architecture of the

software implementation is shown in Figure 11. This is one of the biggest

advantages and leaps in the design methodologies where an ideology has been

automated and implemented [2].


10/16

10

Figure 12 Software architecture of Axiomatic approach [2]

The commercial softwares available are [9]:

a. Acclaro DFSSb. Acclaro Schedulerc. Acclaro Sync

Axiomatic approach is suitable for different fields Manufacturing, Materials,Software, Organisation, Systems unlike the above methods.

Axiomatic approach takes into consideration the manufacturing processes. This canbe seen in the following equations:

[FR] = [A][DP] ......FR to DP mapping

[DP] = [B][PV] ......DP to PV mapping

Thus [FR] = [A][B][PV] ......FR to PV mapping

The above equation is known as the Design for Manufacturing [2]

In case of fixed system, the information content can be given as follows:Isystem = -log(pleaf) + Ia

Where, Ia is the information associated with the assembly of the modules [3].

Thus, Suhs theory clearly addresses the issue related with the integration of

various modules which is not addressed directly in any of the previous theories.

This minimizes the integration issues which would be created while assembling the

parts which are typically designed by different design teams in the industry.

3.6 Lacunae in Axiomatic ApproachAxiomatic approach being robust and versatile does have some lacunae and the following

are some of the lacunae according to us

Although Axiomatic approach is mathematical and quantitative in approach, theprocess of zig-zagging which defines the quality of mapping depends upon the

designers creativity and experience and hence is qualitative in this sense and

hence a new user designer would need time to adapt to the use of the technique

and produce good results.

According to Axiomatic approach the design matrix should ideally be a diagonalmatrix or a triangular matrix. But many times even a sparse matrix is also


11/16


12/16

12

Figure 14 In troduction of iteration steps from DP to CA

By using independence axiom designer is able to optimise individual attributes, butoften this axiom cannot be fully met in real design practice.

As stated earlier axiomatic theory tends towards automation and a major work isdescribed in [10] where the concept of a machine capable of developing designs is

conceived. For this the design knowledge of many designers must be stored in a

database to check for all possible solutions. This actually contradicts the term

thinking machine which is not able to make decisions and is merely a search engine

model. Hence although there is a scope of automation it is not possible to develop

a thinking machine as human factor which has been neglected by Suh should

always be considered.

One basic flaw in Axiomatic Design is independence which is subjective in natureand needs decision making which has not been considered by Suh.

Axiomatic Process does not include cost as a functional requirement but howeverstates that it should considered as a constraint. But sometimes according to the

customers voice cost has to be considered as a functional requirement.

3.7 Salient features of Decision Based DesignAxiomatic Approach introduced the concept of quantitative analysis of design, but has

not taken into consideration the human factor and uncertainty involved. Decision baseddesign realizes that decision making is a critical step involved in quantitative analysis.

Some of the salient features which we would like to state are below.

If we have a closer look at Systematic approach, it s observed that the customersviews and requirements are not taken into account adequately. Although there is a

scope of iteration stated but it is not recommended. On the other hand demand

modelling, customer views and requirements are the most important factors in

design according to the DBD process. The evaluation system is all about satisfying

the customers need as well as generating profit.

The authors clearly see the influence of Total Design on DBD as it takes intoconsideration the overall impact. DBD suggests that design is not only a multi-

disciplinary process; it is rather an omni-disciplinary process [7].

Total Design explains group discussions to evaluate selection criteria, i.e., groupwould arrive at a rational decision which is generally not the case as proved by the

Arrows Impossibility Theorem [7]. These decisions are often made based in

experience and instinct of the designer. There is no mathematical justification for

these decisions. Thus Total Design evaluation method as the authors have already

stated earlier is qualitative and subjective. This is the reason why the same design

tool may lead to completely different results depending on the person who is using

it and the decisions he/she makes during the process. On the other hand Decision

Based Design quantifies this decision making process using axioms that underlie

the value theory.


13/16

13

DBD realizes that while teaching engineering design, the students should be trainedto handle the uncertainty. This is very necessary since uncertainty is always

associated with design as products or utilities are always designed for the future.

Methods like Quality Function Deployment, Design for Manufacture, Design forAssembly, Design for Quality, Concurrent Engineering, Pahl and Beitz, Suh, etc. all

require functional specifications of product as the starting point. None of the

methods provide mathematically consistent and logically correct insights on optimal

specification of product nor do they address the issue of inevitable trade-off like

product cost and product performance. DBD gives a mathematical basis for making

these trade-offs under conditions of uncertainty and risk, thus enabling the

designer to utilize the previously mentioned design technique efficiently.

DBD provides a method for modelling the uncertainty and ranking the variousalternatives by taking into consideration the uncertainty and risk involved.

DBD provides data as to which analytic equations are relevant and areas where theexperimentation should be focused to improve the analytic model. In other words it

pin-points the design options or variables within an option which should be focused

on to achieve maximum benefits in trade-offs.

DBD helps to find the design variable which will result in the best combination ofthe desired attributes.

DBD does not provide an optimization methods or algorithms but helps informulating the objective function.

3.8 Lacunae in Decision Based DesignLike other methods even DBD has some lacunae. Some of them according to us are

mentioned below.

DBD does not resolve the central problems of classical group decision making likechange in the result depending on the voting method used [11].

Although it models the uncertainty, it cannot provide solution or framework toeliminate it completely from the design. In other words it cannot provide a method

to make the design independent of the uncertainty [11].

DBD does not help in determining the constraints or range for the trade-offs. Also itdoes not provide the functional relationship between the design variables and

performance attributes. Experimentation and engineering analysis is still needed to

arrive at these values [11].

DBD cannot be employed during the creative or configuration stage, but can enablethe designer to think in terms of function rather than form [11].

Although a framework has been developed for DBD, this theory is still in thedevelopment stages and a number of questions need to be answered to facilitate

smooth implementation of this technique. Some of the prominent issues are ways

to develop flexible design representations so as to ensure that all design options

are explored, computational capabilities needed for the extensive optimisation,

inertia to change from the industry [12].

The integration of DBD approach in todays product design infrastructure wouldrequire some major changes which may not be accepted readily by the industry

[12].

3.9 Comparison for Original , Variant and Adaptive DesignAs stated in section 3.1 we completely agree with the three types of designs Original,

Adaptive and Variant. In the following table we have ranked the methodologies for

generating the above three design types


14/16

14

Systematicdesign

Total designAxiomatic

designDecision

based design

Original design XX XX XXXX XX

Adaptive design XXX XXXX XXX XXXX

Variant design XXX XXXX XXX XXXX

Table 1 Rating different methods for different types of design

Ratings: X=Poor, XX=Moderate, XXX=Good, XXXX=Very Good, XXXXX=Excellent

4.Unification of design methodologiesConsidering the above analysis it is evident that no single method is sufficient enough for a

perfect design. Hence we have made an attempt to unify the methods considering the strong

points and to eliminate or mitigate the lacunae of the others. We have not limited the work to

only the four design methodologies described above but have expanded to include other

methods based on the study and results of others research in the area.

As seen, Axiomatic theory does a good job in considering customer, functions, physical and

process domain, however lacks the influence of the company and the importance of decision

making. Interface with the company is important as the product is developed with theintention to generate profit for the company as stated in Total Design Theory and Decision

Based Design. As stated by Marston, et.al in [13], axiomatic approach in combination with

decision based design yields good results for variant design. Magrab has used axiomatic

approach for solving design problem in combination with Quality Function Deployment (QFD) in

[14] and stated that when appropriated, the design requirements may be classified based on

the functional requirements i.e., the requirements must be firstly established and used for

organising the design requirements. Theory of Inventive Problem Solving (TRIZ) is a powerful

technique and the links between axiomatic approaches has been established by Yang, et.al in

[15]. We adapt some features of the above work in the proposition of new framework. Figure

15 shows the proposed model

Figure 15 Proposed model of a new framework

As seen from the Figure 15 the basic framework consists of Company (Company Objectives),

Customer Domain, Functional Domain, Physical Domain and Process Domain. We use the QFD

method during the mapping process from customer domain to functional domain where the

requirements are ranked according to the relative importance. During the zig-zagging process

when a designer has selected an FR and wants to identify alternative DPs to achieve it, TRIZ


15/16

15

can be helpful. The axioms of axiomatic approach should be satisfied. During these processes

when faced with uncertainty, the process moves towards the DBD framework to quantitatively

solve the uncertainty. As stated in section 3.5, Systematic approach is more suitable for

detailed design stage. Hence during the formulation of the design parameters the systematic

approach should be adopted. In this case as well, under uncertainty the framework adopts

decision theory. As stated in 3.6 axiomatic process lacks the interaction of the physical domain

and the customer domain, hence in the framework we have introduced this interface. We again

adopt TRIZ during the mapping of Physical and Process Domain. Process Domain considers

processes like manufacturing which may involve variation and failures. We have introduced the

application of Statistical process control and Failure Mode and Effect Analysis in this domain to

solve the problems arising due to variation and to ensure quality. As the goal of developing the

design is related to the company, all the domains have an interface with the company domain.

Since the qualitative nature of evaluation is minimum and mathematical in nature, there is

scope of automation of this framework. This framework is also versatile in all fields like

manufacturing, software, materials, etc.

5.ConclusionIn this paper we presented a brief overview of four design methodologies Systematic

Approach to Design, Total Design Methodology, Axiomatic Approach, and Decision Based

Design. We listed some of the salient features and lacunae in each of the methodologies and

proved that none of these methods are sufficient to produce a good design on their own. We

strongly feel that these methods are not competing with each other and can produce better

results if used in combination with each other. We feel that Axiomatic Approach provides the

most robust framework, hence we have used it as a foundation and integrated it with other

powerful techniques adapted from Systematic Approach, Total Design, Decision Based Design,

Quality Function Deployment, Statistical Process Control, Failure Mode and Effect Analysis to

propose a new framework for design activity which is not only concentrated on the detailed

design procedure but also emphasizes on the interface between customer i.e., people,

company objectives i.e., organisation, products, uncertainty i.e., decision making and quality.

References

[1] D.A. Gebala, N.P. Suh, An Application of Axiomatic Design, Research in Engineering

Design, 1992

[2] N.P. Suh, The Principles of Design, Oxford University Press, 1990

[3] N.P. Suh, Axiomatic Design Advances and Applications, Oxford University Press, 2001

[4] N.P. Suh, Axiomatic Design of Mechanical Systems, ASME Journal of MechanicalDesign, June 1995, Vol. 117/5

[5] G. Pahl, W. Beitz, L Wallace, L. Blessing, Engineering Design: A Systematic Approach,

Third Edition, Springer, 2007

[6] S. Pugh, Creating Innovative Products using Total Design: The Living Legacy of StuartPugh, Addison Wesley Publishing Company, Reading MA, 1996

[7] Hazelrigg, G.A., An Axiomatic Framework for Engineering Design, ASME Journal ofMechanical Design, 1999

[8] Web Link of MIT Courseware - http://ocw.mit.edu/courses/mechanical-engineering/2-882-system-design-and-analysis-based-on-ad-and-complexity-theories-spring-

2005/lecture-notes/lec309.pdf, accessed on 24-Feb-2011


16/16

16

[9] Web Link Axiomatic Design Solutions INC

http://www.axiomaticdesign.com/products/default.asp, accessed on 24-Feb-2011

[10] N. Suh, Design of Thinking Machine, Annals of CIRP, 39/1, 145-148, 1990

[11] D.L. Thurston, Real and Misconceived Limitations to Decision Based Design withUtility Analysis, ASME Journal of Mechanical Design, 2001

[12] H.J Wassenaar, W. Chen, An Approach to Decision Based Design with Discrete Choice

Analysis, ASME Journal of Mechanical Design, 2003

[13] M. Marston, B. Bras, F. Mistree, The Applicability of the Axiomatic and Decision BasedDesign Equations in Variant Design, Proceeding of DETC: ASME Design EngineeringTechnical Conferences, 1997

[14] E.B Magrab, Integrated Product and Process Design and Development: The ProductRealization Process, New York, USA, CRC Press, 1997

[15] K. Yang, H. Zhang, A Comparison of TRIZ and Axiomatic Design, TRIZ Journal, 2000

Advance Design Theories

Documents