Modeling and Managing Interdependent Complex Systems ofSystems Modeling and Managing Interdependent Complex Systems of Systems Yacov Y. Haimes Thiseditionfirstpublished2019 ©2019JohnWiley&Sons,Inc. Allrightsreserved.Nopartofthispublicationmaybereproduced,storedinaretrievalsystem,or transmitted,inanyformorbyanymeans,electronic,mechanical,photocopying,recordingor otherwise,exceptaspermittedbylaw.Adviceonhowtoobtainpermissiontoreusematerialfrom thistitleisavailableathttp://www.wiley.com/go/permissions. TherightofYacovY.Haimestobeidentifiedastheauthorofthisworkhasbeenassertedin accordancewithlaw. RegisteredOffice JohnWiley&Sons,Inc.,111RiverStreet,Hoboken,NJ07030,USA EditorialOffice 111RiverStreet,Hoboken,NJ07030,USA Fordetailsofourglobaleditorialoffices,customerservices,andmoreinformationaboutWiley productsvisitusatwww.wiley.com. Wileyalsopublishesitsbooksinavarietyofelectronicformatsandbyprint-on-demand.Some contentthatappearsinstandardprintversionsofthisbookmaynotbeavailableinotherformats. LimitofLiability/DisclaimerofWarranty Thepublisherandtheauthorsmakenorepresentationsorwarrantieswithrespecttotheaccuracyor completenessofthecontentsofthisworkandspecificallydisclaimallwarranties;includingwithout limitationanyimpliedwarrantiesoffitnessforaparticularpurpose.Thisworkissoldwiththe understandingthatthepublisherisnotengagedinrenderingprofessionalservices.Theadviceand strategiescontainedhereinmaynotbesuitableforeverysituation.Inviewofon-goingresearch, equipmentmodifications,changesingovernmentalregulations,andtheconstantflowofinformation relatingtotheuseofexperimentalreagents,equipment,anddevices,thereaderisurgedtoreviewand evaluatetheinformationprovidedinthepackageinsertorinstructionsforeachchemical,pieceof equipment,reagent,ordevicefor,amongotherthings,anychangesintheinstructionsorindicationof usageandforaddedwarningsandprecautions.Thefactthatanorganizationorwebsiteisreferredto inthisworkasacitationand/orpotentialsourceoffurtherinformationdoesnotmeanthattheauthor orthepublisherendorsestheinformationtheorganizationorwebsitemayprovideor recommendationsitmaymake.Further,readersshouldbeawarethatwebsiteslistedinthisworkmay havechangedordisappearedbetweenwhenthisworkswaswrittenandwhenitisread.Nowarranty maybecreatedorextendedbyanypromotionalstatementsforthiswork.Neitherthepublishernor theauthorshallbeliableforanydamagesarisingherefrom. LibraryofCongressCataloging-in-PublicationData Names:Haimes,YacovY.,author. Title:Modelingandmanaginginterdependentcomplexsystemsofsystems/ byYacovY.Haimes. Description:Hoboken,NJ:JohnWiley&Sons,2018.|Includes bibliographicalreferencesandindex.| Identifiers:LCCN2018000550(print)|LCCN2018009974(ebook)|ISBN 9781119173700(pdf)|ISBN9781119173694(epub)|ISBN9781119173656 (cloth) Subjects:LCSH:Systemsengineering.|Systemanalysis. Classification:LCCTA168(ebook)|LCCTA168.H282018(print)| DDC003–dc23 LCrecordavailableathttps://lccn.loc.gov/2018000550 CoverDesign:Wiley CoverImage:©Digital_Art/Shutterstock Setin10/12ptWarnockbySPiGlobal,Pondicherry,India PrintedintheUnitedStatesofAmerica 10 9 8 7 6 5 4 3 2 1 v Contents Foreword vii Acknowledgments xv 1 ModelingandManagingInterdependentComplexSystemsofSystems: Fundamentals,TheoryandMethodology 1 2 Modeling,Decomposition,andMultilevelCoordinationofComplex SystemsofSystems 51 3 HierarchicalHolographicModelingandMultilevelCoordination ofComplexSystemsofSystems 111 4 ModelingComplexSystemsofSystemswithPhantomSystem Models 141 5 ComplexSystemsofSystems:MultipleGoalsandObjectives 183 6 HierarchicalCoordinatedBayesianModelingofComplexSystems ofSystems 229 7 HierarchicalMultiobjectiveModelingandDecisionMaking forComplexSystemsofSystems 279 8 ModelingEconomicInterdependenciesamongComplexSystems ofSystems 363 9 GuidingPrinciplesforModelingandManagingComplexSystems ofSystems 411 10 ModelingCyber–PhysicalComplexSystemsofSystems:FourCase Studies 447 vi Contents 11 GlobalSupplyChainasComplexSystemsofSystems 527 12 UnderstandingandManagingtheOrganizationalDimension ofComplexSystemsofSystems 559 13 SoftwareEngineering:TheDriverofCyber–PhysicalComplex SystemsofSystems 607 14 InfrastructurePreparednessforCommunitiesasComplexSystems ofSystems 647 15 ModelingSafetyofTransportationComplexSystemsofSystems viaFaultTrees 695 Appendix 739 AuthorIndex 773 SubjectIndex 779 vii Foreword Philosophical and Historical Perspectives on Understanding Commonalities Characterizing Complexity Thegrowinginterestbythesystemsmodelingcommunityintheconceptandin theliteratureoncomplexitydeservesafreshreflectiononitsessenceandonits evolving definitions and characterizations. For systems modelers, the starting point begins by focusing on what constitutes complexity and how to under- stand, model, and manage it. The English language fails to provide a succinct definitionofthetermcomplexityinoneshortorlongsentence.Thisisbecause each of the two words – “modeling and managing” – used in the title of this book has multiple connotations, interpretations, and associations of the term complexitydependingontheindividualsusingthetermsandthespecificcon- text inwhich they are used. Wedefineandmodelcomplexityinthisbookviatheinterdependenciesand interconnectedness (I-I) characterizing complex systems of systems (SoS) (Complex SoS). We further model and quantify the I-I by building on the shared/commonstatesandotheressentialentities(shareddecisions,resources, functions, policies, decision makers, stakeholders, and organizational setups) within and among the subsystems that, in their totality, constitute Complex SoS.Indeed,theabove,alongwithhierarchicaldecompositionandhigher-level coordination, encompass the essence of the modeling, theory, methodology, andpracticeespousedinthisbook.Webuildonthefactthatalloutputsfrom asystemarefunctionsofthestatesofthatsystemandthusalsoofthedecisions andallotherinputstothesystem.Thisfactisofparticularsignificancetomod- elingComplexSoS.Forexample,Chen(2012)offersthefollowingsuccinctdef- initionofstatevariable:“Thestatex(t )ofasystemattimet istheinformation o o att thattogetherwiththeinputu(t),fort≥t ,determinesuniquelytheoutput o o y(t) forall t≥t .” o viii Foreword Indeed,thestatesofasystemarecommonlyamultidimensionalvectorthat characterizes the system as a whole and plays a major role in estimating its future behavior for any given input. Thus, (i) the behavior of the states of the system as a function of time enables modelers to determine, under certain conditions,thesystem’sfuturebehaviorforanygiveninput,orinitiatingevent– and (ii) the shared states and other essential entities within and among the subsystems and systems constitute the essence of the multifarious attributes of the I-I characterizing Complex SoS. Thus, inmodeling Complex SoS, we exploit the I-I characterizing Complex SoSthataremanifestedviasharedstatesandotheressentialentitiesinmultiple ways. The following sample of modeling methodologies beyond Chapter 1 includes (i) decomposition and multilevel-hierarchical coordination (Chapters 2 and 4) with a primer on modeling risk and uncertainty in Part II of Chapter 2; (ii) hierarchical holographic modeling (HHM) (Chapter 3); (iii) multiple conflicting, competing, and noncommensurate goals and objec- tives and the associated tradeoffs (Chapter 5); (iv) hierarchical coordinated BayesianmodelingofComplexSoS(Chapter6);(v)hierarchical-multiobjective modelinganddecisionmakingofComplexSoS(Chapter7);(vi)modelingeco- nomicinterdependenciesamongComplexSoS(Chapter8);(vii)guidingprin- ciples for modeling and managing Complex SoS (Chapter 9); (viii) modeling cyber–physicalComplexSoS–fourcasestudies(Chapter10);(ix)globalsupply chainasComplexSoS(Chapter11);(x)understandingandmanagingtheorgan- izationaldimensionofComplexSoS(Chapter12);(xi)softwareengineering– thedriverofcyber–physicalComplexSoS(Chapter13);(xii)infrastructurepre- paredness for communities as Complex SoS (Chapter 14); and (xiii) modeling safety of highway Complex SoS via fault trees (Chapter 15). Throughoutthisbook,weintroducethereader,viaexamplesandcasestudies, todecomposition,hierarchicalmodeling,multileveldecisionmaking,andopti- mizationandtomultiobjectivetradeoffanalyses.Decompositionisemployedto decoupletheI-IcharacterizingComplexSoS.Wepostulatethatdecisionsmade atthesubsystem’slowerlevelsofthehierarchycanserveasapretextthatthey are “independent.” The discrepancies and conflicts, fundamental differences, andtheassociatedtradeoffsareharmonizedatthehighestlevelsofthemodel’s hierarchical decision-making process. Startinginthe1960s,manyscholarsaimedatidentifyingthefundamentalcom- monalitiesthatcharacterizemodelingandmanagingComplexSoS.Mostofthe theory and methodology that were developed employed decomposition using pseudo-variablesatthelowerlevelsofthehierarchicalmodelsandwereultimately harmonizedatahigherlevelofthehierarchy.Overtheyears,wecontinuedto studyandimproveourmodelingperspectivessupportedbynewtoolsandmeth- odologiesthatledtoabetterunderstandingandmoreusefulmodelingoftheI-I that constitute Complex SoS. In the past, modeling the I-I was directed at the coupled decisions and decision makers that characterized Complex SoS. This Foreword ix wasmostlyachievedbythedeploymentofpseudo-variables,whichenabledthe reliance on decomposition at lower levels of the hierarchy, and a higher-level hierarchicalcoordinationoftightlyinterdependentandinterconnectedsystems andsubsystems. PreviousmethodsdevelopedformodelingComplexSoSwereaimedatadvan- cingtheoryandmethodologyforuncouplingtheI-Ithatcharacterizethem.In thisbook,wewillalsostudyandidentifyinterdependenciesandinterconnections byseekingabettercomprehensionoftheiressenceandtheirdominantcontri- butionstothecomplexityofSoS.WeaddressthischallengebyidentifyingtheI-I ofComplexSoSmanifestedviasharedstatesandotheressentialentities.Wealso embracethefactthatalloutputsfromasystemarefunctionsofthestatesofthat systemandthelatterarefunctionsofalldecisionsandallinputstothesystem. ThisnotionisalsoofparticularsignificanceandcentraltomodelingComplex SoS.Forexample,todeterminethereliabilityandfunctionalityofacar,onemust knowthestatesofthefuel,oil,tirepressure,andothermechanicalandelectrical systems.Allsystemsarecharacterizedatanymomentbytheirrespectivestates andtheconditionsthereof,andtheseconditionsaresubjecttocontinuousvar- iationandfluctuation.Similarly,thestatesofhealthofahumanaremultifaceted, includingbloodcompositionandpressure,amongmyriadothers,andtheI-Ithat existamongthestatesofbiologicalsystems. Thetimeframehasalwaysbeenrecognizedasamajordriverofwhatweterm complexity.Thisisduetothefactthatallsystemscontinuetoevolve,emerge, and thuschange, whilethe capability of ourmodelingtools tokeep pacewith these changes continues to lag behind. Our inability to model the dynamic changes that characterize Complex SoS remains an impediment that charac- terizesandimpairsourmodelingandmanagingtheI-IcharacterizingComplex SoS. We embrace the fact that complexities cannot, by their essence and definition, be compounded, packaged, understood, or modeled via one “straightjacket” modeling schema. Rather, we have to keep building on what wehavelearnedfrompastcontributionsdevelopedbyotherscholars,research- ers,andpractitioners,andaugmentthispastknowledgeintoourcurrentthink- ing, thereby creating new and improved theories and methodologies. Furthermore,seekingtodiscoverwhatmakestheI-IofComplexSoSsodifficult tomodelwillultimatelyhelpusbettermanagethem.Thisisnotafatalisticview ofmodelingcomplexity,ratherasoberunderstandingoftherealitycharacter- izing Complex SoS. Complexity, Interdependency, Interconnectedness, and Reinvention of Fault Trees Fordecadesengineersandscientistshaveexploredthemodelingpoweroffault trees in their quest to study and discover connections between two or among x Foreword several systems that may lead to catastrophic failure of safety-critical systems. The fundamental difference characterizing the previous use of fault trees and our present reinvention stems from the basic characteristics of the two approaches. In this book we investigate and identify the genesis of the I-I by exploring the shared/common states and other essential entities within the systemsandsubsystemsthatcompriseComplexSoS.Bydoingso,wealsodis- cover and quantify the genesis of potential failure of the entire Complex SoS, whether the interdependencies and interconnections are manifested by connections in series and/or in parallel. In this book we also benefit from decadesofexperiencethatengineersandscientistshavegainedfromtheintrin- sicpoweroffaulttrees.Furthermore,tomodelandimproveourunderstanding of the I-I that characterize Complex SoS, we have reinvented the use of fault trees via an innovative interpretation of the contributions that they offer systems modelers. We further exploit the I-I characterizing Complex SoS by tracing (via fault trees) prospective and inevitable failures due to theirinherentspecificconnectionsviasharedstatesand/orothersharedessen- tial entities. This process enables us to determine early in the modeling cycle “what not to do” during planning, design, and future decision making. By investigating the essence of the I-I characterizing Complex SoS, we can discover future failures that could be avoided. In the parlance of fault-tree analysis, the shared states and other essential entities are translated into systemsconnected inseriesorinparallel,ratherthanbeingseenascompletely independent. Thereexistsaninsightfulcorrelationandlessontobelearnedfromthespread ofdiseaseinthehumanbodyduetotheI-Ithatareenabledbythecontinuous flow of blood nourishing every cell (subsystem) of every organ (system) and ultimately of the entire body as Complex SoS. Similarly, all cyber–physical infrastructuresare,intheiressence,ComplexSoS,andtheirmodeling,under- standing,andmanagementcanbecharacterizedbyusingtheirsharedstatesand other essential entities (e.g. communication channels, decisions, decision makers, resources, and organizational setups). Our ability to observe, study, andlearnfromthebehavioroftheanimalkingdomasComplexSoSanddevelop knowledgebasedonlessonslearnedhavebeencentraltotheinsightfromwhich we benefit today. Althoughtheaboveobservations,aswellasthetheoreticaldiscoveries,seem obvioustousnow,theydoshedlighton,andprovideinsightfulunderstanding of,thegenesisoftheI-Icharacterizingbothlivingentitiesandcyber–physical ComplexSoS.Thisfindingconstitutesanotherbuildingblockintherepertoire ofthetheory,methodologies,andtoolsthatenablemodelersofComplexSoSto gaininvaluableinsightintodecipheringthegenesisoftheI-Ithatcharacterize Complex SoS. Consider the nearly two-decade-old perspectives on complexity offered by scholars inthe 1999 Special Issue of the journal Science: