N89 25208
`DECISIONBASED PERSPECTIVE
`
`for the
`DESIGN OF METHODS
`
`for
`SYSTEMS DESIGN
`
`Farrokh Mistree and Douglas Muster
`Systems Design Laboratory
`Department of Mechanical Engineering
`University of Houston
`
`Houston Texas
`
`Jon
`Shupe
`BF Goodrich Company RD Center
`9221 Brecksville Road
`Brecksville Ohio 44141
`
`Janet
`
`Allen
`
`Janco Research
`4501 University Oaks Bouevard
`Houston Texas 77004
`
`Breckenridge Exhibit 1030
`Breckenridge v. Research Corporation Technologies, Inc.
`
`PRECEDING PAGE BLANK NOT FILMED
`
`1111
`
`

`
`MOTIVATION
`
`is
`
`construct which characterizes
`By system definition we mean the establishment of
`the needs and
`The importance of system definition and
`requirements of
`system for
`particular application
`concept design in the development of new major engineering systems cannot be overemphasized
`popular representation that has been used to illustrate this importance with respect
`to life cycle cost
`reproduced below As can be seen design freedom or management leverage rapidly decreases once
`In other words the freedom to make modifications
`becomes
`is underway
`to
`concept
`project
`significant amount of the life
`increasingly expensive as one gets further
`In effect
`into the project
`the time when relatively little
`cycle costs have been committed at
`the object of
`about
`knowledge
`design has been generated Usually this occurs by the end of the conceptual
`and preliminary design
`phases
`We believe that the design research community ---
`and engineers in
`the theorists the academics
`industry --- must all contribute to the common goal of striving to develop
`recognized science of
`In an ultimate sense the purpose of developing the science of design is to ensure that our
`design
`industries can become more effective as well as efficient and that
`the designers
`manufacturing
`manufacturers and maintainers of our products will be working in an environment where their
`subdisciplines are considered simply as parts of the continuous technological spectrum which spans
`what we have come to call
`life cycle engineering
`Our recent work and research interests suggest
`there is
`viewpoint of design research
`that
`and tightly organized within itself which should be considered The issues we include
`three-faceted
`in our view of design research do not exist separately but as
`single interactive entity They are
`
`the way in which we define and partition
`Meta-design
`discipline-independent modeling techniques
`Computer-based design supports holistic or systems thinking
`way of learning through doing
`Adaptive Action Learning
`
`problem using generic
`
`We believe this tripartite view of design research is unique and is essentially congruent with the
`principal elements required to establish the philosophy and practice of the science of design which
`when accepted and used in industry and academe will ensure the continued growth and improved
`productivity of our industries
`
`LI CYCLI
`
`4IHIIIIIIIJ
`
`cot
`
`1T1ARI4
`
`PL
`
`SI
`
`1112
`
`__
`
`_________
`.t 61
`
`1OTN
`OPWTP
`
`

`
`ORGANIZATION OF MATERIAL
`
`Everyone designs who devises courses of action aimed at changing existing situations into
`preferred ones Simon
`1291
`The preceding definition is not discipline specific
`It can be used as the basis for categorizing the
`activities of groups of individuals in other science-based disciplines than engineering for example
`The
`management science systems science economics and the social and behavioral sciences
`members of the groups are designers in the context of Simons definition They design artifacts and
`machines engineers industrial organizations managers including their communication
`and
`information networks behavioral scientists and experts in information science and accounting
`information systems accountants managers and experts in information science We subscribe to
`Simons definition of
`In this paper our comments are directed principally towards
`designer
`The organization of the matenal
`is given below
`engineering design but are not
`limited to it
`
`DEFINITIONS
`
`Decision-Based Design
`Heterarchy and Hierarchy
`System
`
`THE DECISION SUPPORT PROBLEM TECHNIQUE CONCEP11JAL MODELS
`
`Short Term Goal Design that can be Produced and Maintained
`long Term Goal Design Manufacture
`as Continuous Process
`and Maintenance
`
`DECISION-BASED DESIGN
`
`Meta-Design Computer.Based Design and Adaptive Action Learning
`The Charadenstics of Decisions
`
`Decision Activities to Decision Entities
`
`Types of Design
`
`THE DECISION SUPPORT PROBLEM TECHNIQUE
`
`STATUS
`
`Designing for Concept and Designing for Manufacture
`Designing for Concept
`Scenario
`Status Software Decision Hierarchies and Applications
`
`ISSUES THAT NEED TO BE CONSIDERED TO FOSTER DEVELOPMENT
`
`1113
`
`

`
`DECISION-BASED DESIGN
`
`is
`
`new focus from which
`term we have introduced to provide
`Decision-Based Design 123
`In the context of Decision-Based Design we assert that
`design methods can be developed
`the
`role of an engineer is to make decisions associated with the design of an artifact This
`principal
`seemingly limited role ascribed to engineers is useful
`for developing
`to provide
`starting point
`of decisions made by
`design methods based on paradigms that spring from the perspective
`designers who may use computers as opposed to design that is assisted by or based on the use of
`computers optimization methods computer-aided design optimization or methods that evolve
`from specific analysis tools such as finite element analysis In other words we do not consider
`Decision-Based Design as
`subset or superset of Computer-Aided Design or Computer-Based
`Design We see it
`in another role Many design approaches were developed originally for
`purposes and uses now considered outmoded Their continued use by designers is contingent
`largely upon custom tradition and familiarity and the innate conservatism of most engineers
`decision-based process offers designers
`Enter Decision-Based Design considering design as
`for viewing established approaches and provides them with the basis
`new and different perspective
`for extending and developing anew these established tools of the trade
`of DBD can take many forms One implementation of Decision-Based
`The implementation
`Design is the Decision Support Problem Technique
`
`new focus from which to develop methods
`
`new term to provide
`that support
`systems thinking and
`the making of decisions by designers of engineering systems
`
`Principal role of engineer
`the design of an artifact
`
`in DBD is to make decisions associated with
`
`Starting point for developing design methods based on paradigms
`that spring from the perspective of decisions made by designers
`with or without computers
`
`DBD has as its content
`heterarchical set of constructs that embody
`researchers perception of the design environment and the real world
`definition of system
`types of design original adaptive and variant
`open and closed environments
`the nature of decisions and the type of decision activities
`
`There is NO SINGLE unique TECHNIQUE or METHOD for the
`implementation of DBD The development of major class of
`design technique or method will be
`result of
`researcher selecting
`hierarchy between them
`subset of constructs and establishing
`
`1114
`
`

`
`DECISION-BASED DESIGN DEFINITION
`
`We define Decision-Based Design as
`heterarchical set of constructs that embody
`developers perceptions of the design environment and the real world
`
`products life-cycle are the products market
`The heterarchical constructs associated with
`the product the design must meet or exceed the criteria related to the products function
`its capability for being manufactured in serial and when it
`reaches its
`meeting its market
`it be free of unreasonable dangers its manufacture tooling and assembly its
`market that
`and its subsequent
`retirement
`portion of the heterarchical set of constructs
`maintenance
`products life-cycle are shown below The relationships between the constructs are
`for
`not ordered and hence not directed
`
`DECISION-BASED DESIGN is
`heterarchical set of constructs
`designers perception of the real world
`that embody
`
`r.-4-
`
`hA
`B.
`
`12A
`a.
`
`13A
`
`17
`
`HETERARCIIY
`UM
`
`Sska
`
`VPi
`rkeduS
`un cturlni Eeeiiui
`12A n..tr1I V.P i.wlC
`13 Workeri
`
`O11ce
`
`gaIT
`
`31
`
`30
`
`29
`
`55
`
`58
`
`53
`
`so
`
`3S Tho wkpco
`31 T1
`
`Delg
`53 Mat.I
`53 MansAcWre
`
`38 Mi.toi.ce
`
`lnMrum.aUon
`
`91
`
`98
`
`107
`
`102
`
`Ecs...k denLh.ua
`1St Peidi coldorolions
`1$ MarteU col0.todous
`107 CaucatIoo
`riitIo
`
`119
`
`106
`
`ORIGINAL PA
`POOR QUALITY
`
`1115
`
`

`
`DECISION-BASED DESIGN HIERARCHY
`
`Different hierarchies can result from the same heterarchy
`heterarchy is transformed
`for the transformation is identified and the subsystems
`hierarchy once the goal
`into
`that can contribute to the achievement of
`are selected and placed in the
`the goals
`hierarchy
`
`HIERARCHY OF CONSTRUCTS FOR THE DEVELOPMENT
`OF TECHNIQUES AND METHODS IN
`DECISION-BASED DESIGN
`AN EXAMPLE
`
`ASSERTION
`
`The development of major dass of design
`
`method will be the result of
`
`researcher
`
`subset of constructs
`sdecting
`HETERARCHY and establishing
`HIERARCHY
`
`from
`
`DESIGN
`MANUFACTURE
`
`MAINTENANCE
`
`UNIFIED
`
`CONTINUOUS PROCESS
`
`DESIGN
`
`______________
`
`rMANUFACTURE1
`
`______________
`KNOWLEDGE
`
`PRODUCT
`
`__________
`
`PRODUCT
`
`MAINTENANCE
`_________________
`
`KNOWLEDGE
`ABOUT
`PERFORMANCE
`
`1116
`
`

`
`THE DECISION SUPPORT PROBLEM TECHNIQUE
`GOALS AND REPRESENTATION
`
`The implementation of Decision-Based Design can take many forms
`comprehensive approach called the Decision Support Problem Technique
`the University of Houston
`is being developed and implemented at
`to provide
`support for human judgment
`
`from the same heterarchy
`Different hierarchies can result
`heterarchy is
`transformed into
`hierarchy once the goal for the transformation is identified In
`is to unify the processes of design manufacture and
`the long term our goal
`maintenance
`
`for the UI-I Decision Support Problem Technique is to
`In the short term our goal
`develop processes and tools to support
`the making of decisions associated with the
`In the long term we
`design of an artifact
`that can be produced and maintained
`would like to develop the capability to design manufacture and maintain as
`the representations for the two cases are
`unified continuous process Note that
`different
`
`LONG TERM GOAL
`
`REPRESENTATION
`
`______
`
`SHORT TERM GOAL
`
`REPRESENTATION
`
`DESIGN
`Of AN
`ARTIFACT
`THAT CAN RE
`MANUFACTURED
`
`MAINTAINED
`
`MANUFACTURE
`
`1117
`
`

`
`DECISION-BASED DESIGN THE FIRST CONSTRUCT
`
`We define the term system to mean
`groupini of associated entities which is
`The terms in this definition were selected for
`by mental construct
`characterized
`they can convey The term grouping conveys
`the specific meanings and associations
`the impression that an act of forming and arranging is involved Associated is used
`to indicate that there is an association among or that relationships may exist between
`grouping without
`indicating the precise natures of the association or
`the entities in
`relationships The entities could be any thing with an essential nature that can be
`ideas symbols and objects in the
`conceptualized including other systems concepts
`real world The term characterized is meant
`the characterization of
`to convey
`that
`is coupled to the grouping and mental construct
`the grouping is unique and that
`which have been selected Only with both can
`mental
`image of the system be
`complex idea resulting from synthesis of simpler ideas
`created
`Construct
`The redundant qualifier mental serves to highlight the involvement of the human
`This definition is as will become
`mind in the process of creating
`construct
`evident from the following Sections of primary importance for the development of
`methods rooted in Decision-Based Design
`
`is
`
`it
`
`STE
`
`grouping of associated entities which is characterized
`by mental construct
`
`THE PROCESS OF DECISION MAKING AS
`
`SYSTEM
`
`Entltl
`
`Decisions
`DSPs and Decision Blocks
`
`Intormatlon and knowledge
`
`Parent or
`
`DomIflant
`
`DSP
`
`ge
`
`HETERARCHICAL REPRESENTATION
`The relationships
`between
`decision blocks are not ordered
`and hence not directed
`
`HIERARCHICAL REPRESENTATION
`The relationship between decision
`blocks are ordered and hence are
`directed
`
`1118
`
`

`
`CONCEPTUAL MODEL
`DESIGN THAT CAN BE MANUFACTURED AND MAINTAINED
`
`system En the DSP
`process according to our earlier definition can be modeled as
`Technique unification of process is sought through the harmonious hierarchical
`integration
`of
`
`DESIGN
`process of converting information that characterizes the needs and
`prototype of the product
`into knowledge about
`product
`requirements for
`MANUFACTURING
`process in which the knowledge about
`is converted into replicates of the product and
`the product
`MAINTENANCE process in which informationthat characterizes the performance of
`is monitored and
`in terms of its function and its effects on its environment
`product
`analyzed in order to
`
`prototypical version of
`
`maximize the performance/cost
`
`ratio thereby enhancing customer satisfaction
`
`gain knowledge for design modifications thereby increasing industrial competiveness
`
`This hierarchical construct of design for the life-cycle provides the conceptual model of
`design see below for which the Decision Support Problem Technique is being developed
`is clear that the conceptual model can be modeled in its entirety using the entities of DSPs
`and information/knowledge
`
`It
`
`conceptual model representing the short
`Technique is shown below
`
`term goal for the Decision Support Problem
`
`Continued on next page
`
`DECISION BASED HIERARCRY
`
`DE4
`
`ULIGV
`
`____
`CMA1tN
`
`wI_
`
`1119
`
`

`
`CONCEPTUAL MODEL
`UNIFIED DESIGN MANUFACTURE AND MAINTENANCE
`
`Continued from previous page
`
`The relation between the processes of design manufacturing and maintenance is below and
`in the next slide Within the checkered box the three processes contain knowledge about
`the interaction between them This establishes
`process-based hierarchy in the creation of
`an artifact Outside the checkered box the rectangles circles and ovals are used to
`represent collections of related decisions within design manufacturing and maintenance
`respectively These decisions correspond to systems and subsystems when dealing with
`design and plans and subplans when dealing with manufacturing and maintenance
`The
`interaction between the DSPs within each of the processes
`is represented by the lines
`the rectangles circles and ovals The lines represent
`the passing of information
`connecting
`The patterns of
`and knowledge
`the networks
`shown are possible hierarchical
`These
`representations of the decision process in design manufacturing and maintenance
`hierarchies are based on the types of decisions made in each process Hence they are called
`decision-based hierarchies which engineers recognize today as being common to the three
`linked processes of design manufacturing and maintenance
`
`conceptual model
`representing the long term goal
`Technique is shown below
`
`for the Decision Support Problem
`
`DECISION BASED HIERARCHY ______________
`
`__________
`
`DESIGN
`
`ART
`nm
`
`MANUFACrURZ
`
`g4vu
`1Lptcoct
`
`MAINTiANcE
`
`INOWLIb
`
`1\
`
`1120
`
`

`
`DECISION-BASED DESIGN SOME OBSERVATIONS
`
`The design of most real-life engineering systems is characterized by the foLlowing descriptive
`sentences
`
`The problems are multi-leveled multi-dimensional and multi- disciplinary in nature
`Most of the problems are loosely defined and open virtually none of which has
`singular unique solution but all of which must be solved The solutions are less than optimal
`and are called satisficing solutions
`There are multiple measures of merit
`may not be equally important
`All the information required may not be available
`Some information may be hard that
`is based on scientific principles and some information
`may be soft being based on the designers judgment and experience The design environment
`is invariably fuzzy
`Design is the process of converting information that characterizes
`of
`about
`the system itself
`system into knowledge
`
`for judging the goodness of the design all of which
`
`the needs and requirements
`
`The design of
`complex engineering system involves partitioning of the system into smaller
`manageable parts which in turn require the formulation and solution of
`series of problems
`involving decisions to be made by the designer
`
`Decision-Based Design Technique is based on the following assertions
`
`series of decisionssome of which may be made sequentially and others that
`Design involves
`must be made concurrently
`
`Design involves hierarchical decision making and the interaction between these decisions must
`be taken into account
`
`Design productivity can be increased through the use of analysis visualization and synthesis in
`complementary roles and by augmenting the recognized capability of computers in analysis to
`include the use of expert systems with limited at present capability in synthesis
`
`Symbols are processed to support human decisions
`Analog/signals
`Numbers
`
`Graphs/Pictures/Drawings
`Words
`
`technique that supports human decision making ideally
`
`must be process-based and discipline-independent
`must be suitable for solving open problems that are characteristic of
`and
`must facilitate self-learning
`
`fuzzy environment
`
`1121
`
`

`
`META-DESIGN COMPUTER-BASED DESIGN AND
`ACTION LEARNING
`
`Central
`research
`
`to the development of Decision-Based Design are the following major areas of
`
`This consists of two parts namely partitioning and planning
`Meta-Design
`Partitioning deals with the way in which we define and partition
`problem using
`generic
`discipline-independent modeling technique Planning involves the way in which we
`organize the expertise of individuals the information and knowledge embodied in
`databases and computers
`
`Computer-based Design The use of discipline-independent
`processes to facilitate the
`information and knowledge that are needed to negotiate
`generation domain-dependent
`satisficing solutions to problems
`
`AdaptivAction Learning
`
`way of learning through doing
`
`.djgn Meta-design
`In each of the areas of overlap in multidisciplinary program engineers and
`Partitioning
`scientists from one discipline bring with them the intellectual baggage of the technical culture
`in which they have been trained They work as say engineers who have knowledge of the
`problems and methods
`in another area but
`they tend to abide in their discipline and use its
`approaches and methods without changing their mindset Recently as designers move towards
`each other and seek out common ground they have redefined their problems and in the process
`them This meta-level
`represents the common
`meta-level on which to approach
`defined
`ground on which say engineers and managers can meet
`These meta-engineers
`and
`level where the commonahies and only the commonalties of their
`meta-managers operate at
`thus they are of this common ground with mutually understood mindset and
`disciplines exist
`not simply as in the case of multidisciplinary approach to
`problem engineers and managers
`working in the overlapping areas of each others disciplines with the mindset from which they
`come
`
`consists of two parts namely partitioning and planning
`
`The process of planning decisions is crucial
`Planning
`for effective implementation of
`The decisions themselves are not made in this phase
`Decision-Based Design
`rather the
`decisions that need to be made to convert
`information that characterizes the needs and
`into kaowleg about
`that can be
`prototype of
`for
`product
`requirements
`product
`manufactured and maintained are placed in
`decision plan This plan is created with the
`knowledge of what will be needed in implementing
`designers tasks and their relationships
`one to another and on the knowledge gained from meta-engineering
`the design organization
`and its resources the time scaie and the anticipated costs
`
`The pervasive influence of computer-based thinking has spread to
`Computer-based Design
`every part of every science-based discipline like
`benign virus creating an environment
`that
`the parallel growth of systems thinking and an appreciation of the practical
`has encouraged
`limits of analysis-based science in design These events have encouraged designers to look
`approach we
`afield for new paradigms and new approaches and methods
`The computer-based
`in some later discussions here and elsewhere
`espouse is captured in its essence
`In
`term which is used
`is the antithesis of computer-aided design in detail at
`sense it
`least for us
`to characterize methods of automating calculations
`and visualization
`essentially without
`interaction by the designer Our computer-based approach
`to design requires the constant
`human designer and
`interaction between two entities
`computer
`
`1122
`
`

`
`META-DESIGN COMPUTER-BASED DESIGN AND
`ACTION LEARNING .. continued
`
`to
`
`Adaptive Action Learning The focus of effort at present
`in Decision-Based Design
`the object of design early on and to develop computer
`tools for
`increase the knowledge about
`supporting human decision making in the very early stages of project
`In our
`initiation
`improve design practice only under certain
`opinion the development of design theory will
`conditions We offer for consideration one condition that we feel
`is of paramount
`importance
`fields of science academics in design must be concerned with
`Unlike the practitioners in other
`the pedagogical aspects of how design skills associated with both theory and practice can be
`passed on to their students We believe that in the long term only that portion of theory that
`can be taught or as we prefer to say learned to
`large number of students will
`influence
`design practice We have found that what we call adaptive action learning with its emphasis
`on the synergistic effects associated with teamwork 1471 is an essential
`in our
`ingredient
`research and in assuring that our students do in fact understand the approach methods and
`design philosophy we espouse Over two thousand years ago Confucius is quoted as having
`said
`Tell me and will forget
`Show me and will
`remember
`Let me do it and will understand
`
`This captures our feeling and belief
`that only through
`with participation in
`goal-oriented design process
`designers we want
`them to be
`
`hands-on learning process coupled
`truly become the
`can our students
`
`Central to the development of Decision-Based Design are the following major areas of reearcb
`
`META-DESIGN
`Pwiuioii.eg deals with the way in which we define and partition
`inoddilno techniaue
`
`dlscjpline-independent
`
`problem using
`
`genetic
`
`Pkaaiitg Involves the way in which we ore anize the expertise or individuals the informatioui
`and knowledge
`embodied in databases and computers
`
`COMPLTItR.BASED DESIGN
`
`domain-dependent
`The use of disdnline.indcprndent processes to facilitate
`the generation
`to problema Our
`information and knowtede that are needed to negotiate satisflcing solutions
`svstema thinkinp and requires the constant
`interacthm
`computer-based
`approach
`supports
`human designer and
`
`ttweeniseo entities
`
`computer
`
`ADAYFIVE ACTION LEARNiNG
`Confucius is quoted as having said
`Tell me and will forget
`Show mc and will remember
`Let me do It and will understand
`
`1123
`
`

`
`LTHE DECISION SUPPORT PROBLEM TECHNIQUE PROCESS
`
`The principal
`role of any Decision-Based Design process is to convert
`information
`that
`the needs and requirements for
`characterizes
`about
`product
`into knowledge
`the product
`The DSP Technique facilitates the conversion of information for the product
`itself
`into
`that can be used for its manufacture In the DSP Technique
`knowledge about
`the product
`identification decomposition organization and synthesis are used
`
`to identify the information that characterizes the needs and requirements for the design
`and is necessary for the process of design
`to partition and decompose
`design problem into appropriate Decision Support
`Problems
`to organize the domain dependent
`form suitable for solution and
`information in
`to synthesize the component solutions into one system solution and thereby gain
`knowledge about the product being designed
`
`In the DSP Technique the process for converting information into knowledge Consists of
`two phases meta-design and design and six steps as shown below These steps are valid
`for any stage in the design process and the DSP Technique can be used for designing
`systems and components
`
`Our efforts to date have been directed to developing the second phase namely design
`In
`the process we have identified various decision
`hierarchies see later and developed
`software to solve them see later These developments we believe are of value to industry
`classroom.since Decision Support Problems can be
`and are appropriate for use in
`formulated and solved as an activity in any other design scheme particularly if designing for
`is involved
`concept
`
`PHASE
`Meta-Design
`
`IDETIFY
`STEP
`Chant proM dory
`
`Technical brief
`
`Partltlo
`
`problem
`dieclpunc.lndepcsdcnt
`.oddlng icheme
`
`adas
`
`PHASE
`
`Design
`
`STRUCTURE
`STEP
`Oiiee dacpeudant
`mratlo and rorinulala
`DSP taniplatea the woid
`rorauIatl.
`the mathaniajcaj
`
`STEP
`
`SOLVE
`
`Solve the rtsp
`
`PLAN
`STEP
`Identity the Decielen
`Support Problem. and
`
`POST.SOLUflON
`STEP
`Validate .olutlo.
`InvhIatadTed
`
`.olutIo of
`
`tar quen
`
`Dderl.ew$iether
`
`Iteratia Ii
`
`1124
`
`ORIGINAL PAGE IS
`OF POOR QUALITY
`
`

`
`CHARACTERISTICS OF DECISIONS
`
`fine line between developing and maintaining
`designer often walks
`In engineering
`model that
`is amenable to solution and one whose results yield information and knowledge
`is usable in practice
`In engineering decisions
`involving design are characterized by
`that
`the following descriptive sentences
`
`Decisions involve information that comes from different sources and disciplines
`
`Decisions are governed by multiple measures of merit and performance
`
`The measures of merit may not be of equal
`conflict with each other
`
`importance to the final decision and may
`
`All of the information for making an adequate decision may not be available to
`designer
`Some of the information may be hard based on scientific principles some may be soft
`based on the perceptive judgment of the designer and some may be partially soft
`in nature
`empirical
`
`The problem for which the decision is being made is open
`form of the DSPs and govern the type of
`These characteristics dictate the mathematical
`solution algorithms appropriate for solving them For example these characteristics
`the use of traditional single objective optimization in Decision-Based
`virtually rule out
`Design
`
`JiFoMA11ON
`
`____________
`
`FROM
`INFORMATION
`MULTIPLE
`DISCIPLINES
`
`MULTI DIMENSIONAL
`
`CONFUC11NG
`MEASURES OF MERIT
`
`____
`
`____________________________
`
`_____
`
`INFOMA1I4q4
`
`_____
`
`INCOMPLETE
`iNFORMATION
`
`SOFF
`HARD AND
`INFORMATION
`
`OPEN.ENDD
`
`1125
`
`

`
`DECISION ACTIVITIES TO DECISION ENTITIES
`
`An abstract conceptual
`the relationship between decisions as people discuss them and
`statement about
`how they could be structured for computer-based
`solution is made in in the figure below 1281 The
`Decision Support Problems and Decision Blocks are also shown in the Figure The transformation of
`is based on the inherent characteristics of decisions in
`decision activities
`into decision blocks
`engineering design The transformation of DSPs into DBs depends on the type separability and order
`in which the decisions have to be made to effect
`solution As is evident from the figure the process
`of this transformation is modeled using the decision and knowledge/information
`entities described
`earlier Further the transformation process complies with our earlier definition of system The set of
`Decision Support Problems shown in the figure is incomplete
`
`PRELIMINARY SELECTION
`The selection of the top-of-the-heap concepts for further development
`
`SELECTION
`The indication of
`
`alternatives
`
`preference based on multiple attributes for one amongst several
`
`feasible
`
`COMPROMISE
`The improvement of an alternative through modification
`
`HffiRARCHICAL
`Decisions within decision entity in which both selection and compromise occur
`Decisions between decision entities which involve subplan
`interaction and compromise
`
`CONDITIONAL
`Decisions in which the risk and uncertainty of the outcome are taken into account
`
`HEURISTIC
`Decisions which are made on the basis of
`
`knowledge
`
`base of facts and rules of thumb
`
`ACTIYITIESD
`
`IIcIJ
`
`CONVERT ACTIVITIES INTO DSPs
`and
`DECISION ENTI11ES
`
`N.k1SUC
`
`C0WIUI1Sl
`
`cost I1o.
`
`ID.ák..IllIhs
`
`SETIZA1CaC1
`
`FIve According to Decision Plan
`
`1126
`
`

`
`TYPES OF DESIGNS
`
`The design specification
`for the system may require the system
`Original Design
`new function altogether An original solution
`to perform the same similar or
`desired system and used to create the knowledge
`and
`principle is defined for
`information so that it can be manufactured and maintained
`
`An original design is adapted to meet modified set of
`Adaptive Design
`The solution principle in the main remains the same but
`the
`specifications
`can meet
`be sufficiently different
`product will
`the changed
`so that
`specifications
`
`it
`
`The size and/or arrangement of parts or subsystems of the
`Variant Design
`set of specifications more cheaply
`chosen system are fine-tuned say to meet
`the specifications and solution principle remain the same
`
`ADAPTIVE DESIGNJ
`
`___
`
`f.ORIc
`
`LSPEC IFI CATIONS
`rr iu ii
`
`DigEDJ
`
`ri
`
`VARIANT DESIGN
`
`FINE-TUNED
`
`CHARACTERISTICS OF DESIGN
`GUIDANCE SYSTEM
`
`IT MUST BE COMPATIBLE WITH ALL THREE TYPES OF DESIGN
`
`ELEMENTS OF IT MUST BE USABLE INTERCHANGEABLY
`
`1127
`
`

`
`DESIGNING FOR CONCEPT
`
`information that
`role of any design process is to convert
`According to our definition the principal
`into knowledge about
`characterizes the needs and requirements for
`itself
`the product
`product
`is safe to assume that because of the complexity of the product an engineering system
`Further it
`the conversion of information into knowledge will have to be accomplished in stages
`In the
`traditional design process names have been given to the stages such as feasibility conceptual
`preliminary and detail The names and the number of stages from the standpoint of the information
`necessary for making decisions in each of the stages are not important What
`is that
`is important
`the types of decisions being made e.g selection and compromise are the same
`in all stages and
`
`the amount of hard information increases as the knowledge about
`increases
`
`the product
`
`It appears to us that in Decision-Based Design the ratio of soft to hard information available is
`human designer needs as he/she negotiates
`key factor in determining the nature of the support
`that
`design problem Our current efforts are focused on understanding what
`is needed
`solution to
`project We
`and developing the tools to support human decision making in the early stage of
`is available to make
`is possible based on the ratio of hard to soft information that
`that
`assert
`distinction between designing for concept and designing for manufacture see below Based on
`computer-based aids into two categories namely tools
`is possible to categorize
`this distinction it
`that provide support for human decision making and tools for automation
`
`it
`
`DESIGNING FOR CONCEPT AND MANUFACTURE
`DECISION-BASED PERSPECTIVE
`
`Hard jlor.atpo./pofl
`
`iforatão
`
`Ideatks
`
`tk Support
`Ars .cp4ies
`Sdtios
`Ccpro
`kpid proiotyp1
`
`Atonuto
`
`CAD
`Sotd o4elli
`Toera.c
`cthrl.II prv.-
`Ma
`
`1TTTTT.7TT
`
`Co.trptaI
`
`Detail uay
`cung marnafaclzuubanyl
`
`netii
`Dr.wmp
`
`for
`Maau.ctare
`SertaI
`
`Eeooik
`
`V.blty s7
`
`Dimrs.I
`Syatha
`
`Pro4otype
`Teig
`imcnj
`
`1128
`
`

`
`ORIGINAL PAGE Is
`OF POOR QUALITY
`
`FOR CONCEPT
`
`SCENARIO
`
`In designing for concept we seek
`and then
`wide net
`is generate many concepts
`to cast
`that
`that meets its functional specifications and can be produced and
`systematically home-in on
`concept
`maintained In other words we are involved in the process of converting information that characterizes
`the needs and requirements for
`into specific knowledge that can be used in designing for
`product
`In designing for manufacture we attempt to ensure that
`manufacture
`the product can be manufactured
`cost-effectively Of course we recognize that
`in practice iteration will occur and this for convenience
`has not been illustrated in the figure
`
`is shown below Let us assume that the process is underway
`schematic of designing for concept
`This permits ideation that results in the identification of
`and the problem definition is available
`alternative ways concepts of achieving the objectives embodied in the problem definition Ideally
`large number of concepts should be generated At this stage most of the information will be soft and
`preliminary selection DSP being formulated and
`there should be many concepts We envisage
`solved to identify the more promising top-of-the-heap or most-likely-to-succeed
`concepts
`In
`the end product of ideation We evaluate the concepts
`preliminary selection we start with concepts
`judgment or soft
`based on criteria The criteria are quantified using experience-based
`information
`only The solution to the preliminary selection DSP involves the rank ordering of concepts Therefore
`one cannot automatically infer from the rankings by how much one concept
`is preferred to another
`is injudicious to use this approach to identify the best concept At this stage we expect
`and hence it
`