system development in dynamic environments: …system architecture selection of the types of system...

Copyright © 2009 Reich / Research in SE / Technion 1

System Development in Dynamic Environments: System Development in Dynamic Environments: Tools for architecture development and process Tools for architecture development and process

managementmanagement

Prof. Yoram ReichProf. Yoram ReichSchool of Mechanical EngineeringSchool of Mechanical Engineering

Faculty of Engineering, Tel Aviv UniversityFaculty of Engineering, Tel Aviv University

With collaboration ofWith collaboration ofDr. Dr. AvnerAvner Engel, Dr. Engel, Dr. ArieArie Karniel, Yuri Karniel, Yuri BelskyBelsky


Outline

Unpredictable dynamic environments

Product development paradox

Product architecture

Robust product concept

Development processes

Conclusions


Unpredictable dynamic environments

Dynamic environments could be unpredictable

Environment changesChanges could be unpredictable, incomprehensible, due to our incomplete knowledge of the worldChanges lead to further environmental changes


More SE → better performance


Concept/architecture design is one of the 2 most critical steps in product development

Resource spending on this project stage is <5%-10% of budget

More upfront SE → more impact

Time into product life-cycle/project

20

80

60

40

100

Knowledge

Cost committed/Quality determined

Cost incurredSpec

ifica

tion

deve

lopm

ent

Con

cept

desi

gn

Prod

uct

desi

gn


The product development The product development fantasyfantasy


20

80

60

40

100Knowledge

… there are no changes in the product, no design iterations

… design can be done in parallel, reducing time-to-market


Cost incurred

Knowledge is available to make the best choices when decisions are required

Spec

ifica

tion

deve

lopm

ent

Con

cept

desi

gn


The product development The product development paradox paradox (PD(PDPP))


Design freedom20

80

60

40

100

Knowledge

(more) Familiar figure


Cost incurred

Knowledge is lacking when critical decisions are made

… by the time we have knowledge, design freedom diminishes, cost have been committed and changes are costly


The product development The product development paradox paradox (PD(PDPP) (2)) (2)

ת ו צאו ת הרסניו ת ז מ ן פ ית וח , על ות גבוהה–אי ן חו פ ש פע ולה לבצע שינ ו י ים כש הם נדרשים

ס יכ ו ן גדל , מ ת מ שךבזבוז מ שאבים, הפ סד הזדמ נ ו י ות –ידע נ וצר אינ ו מנ ו הל

Time into decision life-cycle/project

PotentialDecisionfreedom

20

80

60

40

100Knowledgegenerated

Usableknowledge

AvailableDecision freedom

(less) Familiar figure


The product development The product development paradox paradox (PD(PDPP) (3)) (3)


20

80

60

40

100Useable Knowledge

Preliminary PD Detailed PD

Increase useable knowledge


20

80

60

40

100Preliminary PD Detailed PD

Decision freedom

Maintain decision freedom


20

80

60

40


Cost of change

Reduce cost of change


20

80

60

40


# changes

Move changes earlier


מעקרונות לגישת פתרוןמעקרונות לגישת פתרון: : פרדוקס האי כותפרדוקס האי כות

מטרות אינ טג רציה של מקורות רבים –העשרת הידע הזמ ין

וניהול היד ע ניהול דרגות החופ שהקטנת ע לות ש ינו יי ם

יכולת לבצע ש ינו י ים בתחילת התהליך בעלות נמוכה מינ ו ף ש ינו יי ם כהזדמנוי ות למידה ויצי רת ידע נו ס ף –

תהליך פיתוח אידאליTime into decision life-cycle/project

20

80

60

40

100Useable Knowledge

Preliminary PD Detailed PD

Decision freedom

Cost of change

# changes

גישהשי מו ש בשי ט ות תכן מ שולבותניהול דינמי של תהליך הפיתוח

י שלוב השי ט ות בתהליך "ניהול ידע עסדור

1. Feasibility study

2. Problem formulation

3. Concept generation

Dynamic process management

iterations

Information management


י ישום הגישה

RQFDRQFD (Resource/Robust QFD)

WPWP (Wicked Problem formulation)

APAPSOSSOS ((Concept generation)

DPDPDPDP (dynamic product development process management)

Acknowledgements:DPDP: Arie KarnielRQFD: Eyal Levy, Amir PazWP, n-dim: Prof. Eswaran Subrahmanian, n-dim groupSOS: Dr. Amir Ziv-AvAP: Dr. Avner EngelAPSOS: Dr. Avner Engel, Yuri Belsky

עקרונותי ביצוע ני תוחי רגישות"ח סינ ות הפתרון ע

מש וב מיד י מפ ורט מאפשר ס בבי הגדרת ) e.g., WP+SOS( בעיה ופתרון מהירי ם

קלות שי מ וש בגלל שי מו ש בנתונים ג סי ם

nn--dimdim (n dimensional information modeling)

1. Feasibility study

2. Problem formulation

3. Concept generation

Dynamic process management

iterations

Information management

גישהשי מו ש בשי ט ות תכן מ שולבותניהול דינמי של תהליך הפיתוח

י שלוב השי ט ות בתהליך "ניהול ידע עסדור


System architecture

Selection of the types of system elements, their characteristics, and their arrangement

[INCOSE SE Handbook]

SoShierarchicalInteractingcomplex

Environment:Economic crisisOil price fluctuatingIntense competition


AP: Assumption Promises (Dr. Avner Engel)

Ensuretraceabilitycompatibilitycompleteness

UAV System

ssumptions romises

Ground station

Communication

Air vehicle

Payload


AP: Assumption Promises (Dr. Avner Engel)

External and internal interaction between assumption and promises


SOS – robust product concept innovation

Subjective Objective SystemSubjective Objective System for Optimal Robust Concept GenerationPhD work of Amir Ziv AvAmir Ziv Av, being continually extended in different directionsUsed in Ziv Av Engineering

References:Y. Reich and A. Ziv-Av, “A comprehensive optimal product concept generation framework,” in Proceedings of the 15th International Conference on Design Theory and Methodology (DTM), (New York, NY), ASME, 2003. A. Ziv-Av and Y. Reich, “SOS – Subjective objective system for generating optimal product concepts,” in CD-ROM Proceedings of the 14th International Conference on Engineering Design (ICED), The Design Society, 2003.A. Ziv-Av and Y. Reich, “SOS – Subjective objective system for generating optimal product concepts.” Design Studies, 2005.Y. Reich and A. Ziv-Av, “Robust product concept generation”, in CD-ROM Proceedings of the 15th International Conference on Engineering Design (ICED), The Design Society, 2005.


SOS and robust product concepts

Inputs to SOS:Product objectives

Customer requirementsManufacturer objectives… government regulations, etc.

Building blocksConstraints

Influence of building blocks on attaining product objectives


Example car design: Product objectives

Customer – objectiveStability and handlingSafety in accidentsLow ownership cost…

Manufacturer – subjectiveMinimal investment/risk in development (for Peugeot, Citroen)Minimal mass production costEase of maintenance …

Objectives have relative weight ww which define a market segment


Example car design: Building blocks (BB)

Front engineRear engineDriver before front wheelsDriver behind front wheelsFront wheel driveRear wheel drive4x4 wheel driveCentral transmissionRigid axelsIndependent suspensions

Value of building blocks – D = 0,1


Example car design: Constraints b/w BBs

1. Front engine or Rear engine(Front engine) + (Rear engine) =1

2. Front, Rear, or 4x4 wheel driveFront + Rear + 4x4 = 1

3. 4x4 drive requires central transmission and vice versa(4x4 drive) – (central transmission) = 0

4. Driver behind front wheels and Front wheel drive require independent suspension(Independent suspension) – ((Driver behind front wheels) + (Front wheel drive)) ≥ -1

From our experience in numerous projects, all constraints could be represented by simple linear expressions


Example car design: Influence of BBs on product objective

1st order

2nd order

LI( Front engine ) = 1

LI( Rear engine ) = -1

Minimal investment/risk in development (for Peugeot, Citroen)

LI( 4x4, central transmission) = -1

LI( Rear engine, Front wheel drive) = -1


Organization of information into layers

Building block variables


Mathematical Formulation (basic form)

subject to

( )1, , , 1, ,k n kg D D b k R≤= =≥

K K

∑ ∑= =

⋅⋅=n

jk

n

kljkjl DLIDLV

1 1

njD j ,,1,1,0 K==

∑=

⋅=m

lll NLVwQ

1max

11

=∑=

m

llw

min

max minl l

ll l

LV LVNLV

LV LV−

=− The user of SOS is

unaware of this formulation.

After entering the data, SOS outputs the optimal concept.


APSOS: integrating SOS with AP (Yuri Belsky)Product

Component 1 Component 3Component 2

ln1

ln2 ln3

Set5 Set6 Set7

ln4

ln5

ln7

ln9ln6

Sub-product

Sub-product

Component 4

ln10

ln11

ln12

ln14ln13

Set1 Set2

Component 5

ln15

ln16

ln17

ln19ln18

Set3 Set4

ln8


Dynamic product development processes (Dr. Arie Karniel)

Need to account for dynamic system development projectsConstantly evolving in an unpredictable manner

Make sure fast changes are soundPlan based on simulations of randomized events (albeit from a predictable ensemble)


Dynamic new Product Development Processes (DPDP)

ProcessGenerator

ProcessEngine

Process Status

Product Design Knowledge

Process Scheme

PredefinedProcess

&Changes

Requirements

Constraints

AgendaAgenda


DPDP

InputPredefined process schemeVarious constraintsReal-time project status and knowledge

OutputNew process scheme optimized to desired objectivesRecord of design process


DPDP: Framework

ProcessGenerator

ProcessEngine

Process Scheme

PredefinedProcess

Process change source listAgendaAgenda


DPDP: Framework

ProcessGenerator

ProcessEngine

Process Status


Process Scheme

PredefinedProcess

AgendaAgenda


DPDP: Framework

ProcessGenerator

ProcessEngine

Process Status


Process Scheme

PredefinedProcess

Process change source listAgendaAgenda


DPDP: Framework

ProcessGenerator

ProcessEngine

Process Status


Process Scheme

PredefinedProcess

&Changes

Requirements

Constraints

AgendaAgenda


DPDP: Process Planning CycleFrom product knowledge to process plan

Define product Components

Define Component parametric links

Chassis & Cover

Drum

Drum Rotation system

Laser head

Writing head

drive sys

Load /unload

unit

Registration & punching

unit

Control system

A B C D E F G H

Chassis & Cover

A Weight 3

Size 9

Weight 1 Size 3 Max velocity 1

Weight 1 Weight 1 Size 3

Drum

B Inertia

Moment 1 Weight 1

Punch force 1 Beam pressure 1

Drum Rotation system

C Weight 1

Inertia Moment 9

Power 1 Exposure time 1

Gripper force 6 Beam pressure 3

Laser head

D Speed 1

Size 3 Processing

time 1

Writing head drive

sys

E Weight 1

Size 6 Exposure time 3

Processing time 1

Load /unload unit

F

Registration & punching

unit

G Size 6

Surface 1 Roughness

Plate load speed 1

Control system

H Rotation

accuracy 1 Resolution 1 Acceleration 3

Diodes 1 Exposure time 1

Resolution 3

Load/Unload rate 1 Load/Unload Time 1

System type 3

DSM Representation

RearrangingDSM

Probability DSM

Planned Process


End

DesignConceptual

Design

Begin Specification

PDR CDR

Production

Cancel Project

Iteration Iteration

DPDP: New product process example


DPDP: sound and unsound processes

Changes in the process scheme must lead to sound processes

(a) Live lock process (b) Sound process

Iterations of parallel paths


DPDP: Process simulations and decision making

Process simulations of random changes allows making better process design decisions

Difference function

P1 – Serialized

P2 – Design

block


Conclusions

Product development paradoxProduct development paradox

Integrated suite of tools for architecture development Integrated suite of tools for architecture development and dynamic process managementand dynamic process management

SOS moves robustness into conceptual designRobustness is addressed by systematically simulating conceptual design with varying conditions and maintaining traceability and compatibility between componentsDPDP supports dynamic changes.


Additional information:email: [email protected]: www.eng.tau.ac.il/~yoramgoogle.com: yoram reichCourse Material: Product development: Introduction and methods

1. A. Karniel and Y. Reich, Managing dynamic new product development processes, CD proceedings of the INCOSE International Symposium (INCOSE 2007), San Diego, CA, 2007. (DPDP)

2. A. Karniel and Y. Reich, Coherent interpretation of DSM plan to PDP simulation, in CDROM Proceedings of the 16th International Conference on Engineering Design (ICED), The Design Society, 2007. (DPDP)

3. A. Karniel and Y. Reich, “From DSM-based planning to Design Process Simulation: A review of process-scheme logic verification issues”, IEEE Transactions on Engineering Management, in press, 2009. (DPDP)

4. E. Kolberg, Y. Reich, and I. Levin, Design of design methodology for autonomous robots, RoboCup 2007, Atlanta, 2007.5. E. Kolberg, Y. Reich, and I. Levin, Express engineering change management, in CDROM Proceedings of the 16th International

Conference on Engineering Design (ICED), The Design Society, 2007.6. Y. Reich and E. Levy, Managing product design quality under resource constraints, International Journal of Production Research,

42(13):2555-2572, 2004. (RQFD)7. Y. Reich, E. Kolberg, and I. Levin, Designing contexts for learning design, International Journal of Engineering Education, 22(3):489-

495, 2006. 8. Y. Reich and A. Paz, Managing product quality, risk, and resources through resource quality function deployment, Journal of Engineering

Design, 19(3):249-267, 2008. (RQFD)9. Y. Reich and A. Ziv-Av, Robust product concept generation, in CDROM Proceedings of the 15th International Conference on

Engineering Design (ICED), The Design Society, 2005. (robust concept design using SOS)10. Y. Sered and Y. Reich, Standardization and modularization driven by minimizing overall process effort, Computer-Aided Design,

38(5):405-416, 2006. 11. E. Subrahmanian and Y. Reich, Advancing Problem Definition and Concept Generation for Improved Product Life Cycle, International

Conference on Trends in Product Life Cycle, Modeling, Simulation and Synthesis, PLMSS-2006, 18-20 December, Bangalore, India, 2006. (Invited Talk) (WP + SOS)

12. A. Ziv-Av and Y. Reich, SOS – Subjective objective system for generating optimal product concepts, Design Studies, 26(5):509-533, 2005. (SOS)

13. Y. Reich, Preventing Breakthroughs from Breakdowns, Proceedings of the 9th Biennial ASME Conference on Engineering Systems Design and Analysis ESDA2008, Haifa, Israel, 2008. (describes the PDP)


Thank you for your attentionThank you for your attention


רקע –יורם רייך ' פרופאבי ב-תל' א ו נ , בהנדסה מכ נ ית) בהצט י נות(וש נ י ) בהצ טי נ ות י תרה(תואר רא ש ו ן

מ"פ ותי ב" ש נ ים בחיל הים בתכנו ן הנדסי כולל הקמ ת מדור מו6Carnegie Mellon University, תואר של י שי ב ש י מו ש י ב י נ ה מלאכו תית בהנ דסה

; Engineering Design Research Center ,Carnegie Mellon University -עב ודה בDuke University ; שבתון ב- Design Center ,Stanford University

ר התא הישר אלי של הא יג וד הבין לאומ י " חבר הנהלה ויו–בעב ר SME – Society of Manufacturing Engineers ,

פע יל בתא היש ראל י של האי גוד הבי ן לאומ י ל הנדסת מ ערכ ות –כיום מא מרים מד ע יים 170 -פרסום למ עלה מ

Research in Engineering Designע ורך רא ש י של ה ירחו ן המ דעי ירחו נ ים מדע יים אחרים5 -חבר מ ער כת ב

Design Society -ר מ ש ו תף של ה תא לתור ת תכן ש ל ה"יוחבר ו עדות מד ע יו ת של כ נ סים רב י ם בע ולם

תל אביב ' א ו נ, תוכנ ית ל ימ ודי ת ו אר ר א ש ו ן בהנדס ה מכנ ית, מ"רא ש מס לול תכן ותיבת ל אביב ' א ו נ, רא ש המ עבדה למ ערכ ות מ מוח שב ו ת

, כר יי ת נת ו נים , בי נה מלאכ ותית, מ ע רכות מ י דע בהנד סה, ניהול יד ע, התמחות בש יט ות תכן הנדסימ"תיב

system development in dynamic environments: …system architecture selection of the types of system...

Documents