Understanding, Modeling, and Managing Interdependent Complex Systems of Systems 20 OCTOBER 2017

Transcription

1 Understanding, Modeling, and Managing Interdependent Complex Systems of Systems 20 OCTOBER 2017 Yacov Y. Haimes Lawrence R. Quarles Professor of Systems & Information Engineering and of Civil & Environmental Engineering Founding Director (1987), Center for Risk Management of Engineering Systems University of Virginia 1

2 : The world is totally connected Jacob Bronowski [1978] Whatever explanation we invent at any moment is a partial connection, and its richness derives from the richness of such connections as we are able to make. There is no nerve without the muscle and no muscle without the nerve in the total animal. This is the same statement as the one I made about the total connection of the world 2

3 What is the common denominator among the following? (i) Universities; (ii) Supply Chain; (iii) Electricity and Communications; (iv) Families and Communities (v)democracy and Free Press? (vi) International Traveel; etc. 3

4 The growing interest in complexity by the professional community deserves a fresh reflection on its essence and on its evolving definitions and characterizations. For systems modelers, the starting point begins by focusing on what constitutes complexity and how to understand, model and manage it. 4

5 This seminar offers an understanding of complexity through the discovery of its specific attributes, thereby enhancing our ability to effectively manage complexity with new analytical models. 5

6 We define and model complexity via the interconnectedness and interdependencies (I-I) within and among the systems characterizing complex systems of systems ( Complex SoS ). 6

7 System A System E Complex SoS System D System B System C 7

8 Current models for emergent complex SoS are insufficient because too often they fail to incorporate the complexity derived from the networks of interdependencies and interconnectedness (I-I) characterizing complex systems of systems. This requires a reevaluation of the way we model, assess, manage, communicate, and analyze the risk thereto. 8

9 The key to modeling and managing Complex SoS lies in understanding the genesis of characterizing the interconnectedness & interdependencies manifested through shared: States, decisions, resources, functions, policies, decisionmakers, stakeholders, organizational setups, and others. 9

10 Why do farmers irrigate their crops in non-rainy seasons? 10

11 Why do farmers fertilize their crops? 11

12 Why do farmers irrigate their crops in non-rainy seasons? The answer is fundamental to the role that state space plays in systems modeling; and to understanding the definitions of vulnerability and resilience of, and thus the risk to systems. 12

13 Exogenous Variables Price of fertilizer Random Variables Sunlight Precipitation Input Water from upstream Power demanded Soil Moisture Soil Nutrients (The Farm) Decision Variables When to irrigate and fertilize, And by how much? Output Crops yield (Objectives) Maximize profit Minimize soil erosion 13

14 States of the System All decisions are made: (i) to control (retain or change as appropriate) the levels of the essential states of the system; (ii) to meet specific desired outputs (goals and objectives); (iii) At acceptable tradeoffs; and (iv) At acceptable time frame. 14

15 States of a System Given a system s model, the states of a system are the smallest set of independent system variables such that the values of the members of the set at time t 0 along with known inputs, decisions, random and exogenous variables completely determine the value of all system variables for all t t 0 (under certain conditions). 15

16 Modeling I-I Complex Systems of Systems (i) Models are built to answer specific questions; (ii) They ought to be as simple as possible and as complex as required; (iii) Modeling is an amalgamation and symbiosis of the arts and the sciences; (iv) Artists reconstruct images and ideas, scenes, people and structures; Similarly modelers of SoS decompose and restructure complex systems of systems. 16

17 Understanding and Modeling Complex Systems of Systems (Complex SoS) with Hierarchical Holographic Modeling (HHM) 17

18 Hierarchical Holographic Modeling (HHM) Political Decomposition 18

19 Hierarchical Holographic Modeling (HHM) Hydrological Decomposition 19

20 HHM Features Flipping Perspectives Hydrological Perspective Geographic Perspective Multiple perspectives allow for complete coordination 20

21 Flipping Perspectives Of Complex Systems of Systems A key feature of HHM is the concept of flipping, or changing perspectives, to highlight interconnectedness and interdependencies characterizing complex systems of systems. 21

22 HHM Features: Flipping Perspectives The Maumee River Basin is described by: Five planning sub-areas each consisting of several counties political/geographic decomposition Eight watersheds crossing state and county boundaries hydrological decomposition Seven major objectives identified by an advisory group functional decomposition Three planning time horizons temporal decomposition Geographic Hydrological Functional Temporal 22

23 Modeling Complexity: HHM Features: Flipping Perspectives Geographic Perspective County 1 County 2 County 3 County 4 County 5 Watershed 1 Watershed 5 Watershed 6 Agriculture Industry Aquatic Life Municipality County 2 contains three watersheds Protecting fish and wildlife may be of interest for the three planning periods 23

24 Modeling Complexity: HHM Features: Flipping Perspectives Hydrological Perspective Watershed 1 Watershed 2 Watershed 3... Watershed 8 Agriculture Industry Aquatic Life Municipality County 1 County 3 County Watershed 3 has serves several large corporations and transcends three counties Planning periods are considered for all three of those counties 24

25 Matrix Organization of a Production System Manufacturing Manager Plant 1 Manager Plant 2 Manager Plant 3 Marketing Manager Product A Manager Product B 25

26 Product-Plant Decomposition Third Level Higher Level Coordinator Second Level Product Manager A Product Manager B First Level Plant 1 Plant 2 Plant 3 Plant 1 Plant 2 Plant 3 26

27 Plant-Product Decomposition Third Level Higher Level Coordinator Second Level Plant Manager 1 Plant Manager 2 Plant Manager 3 First Level Product A Product B Product A Product B Product A Product B 27

28 Defense Infrastructure Sectors C3 ISR Financial Services Transportation Public Works Space Logistics Health Affairs Personnel D.I.I. C3: Command, Control, and Communications ISR: Intelligence, Surveillance and Reconnaissance DII: Defense Information Infrastructure 28

29 Civilian Infrastructure Sectors Telecommunications Electrical Power Systems Gas/Oil Trans. & Storage Banking and Finance Transportation Water Supply Systems Emergency Services Continuity of Government 29

30 Civilian Infrastructure Telecommunications Elec Power Systems... i... Continuity of Government Regional Decomposition Region 1 Region 2... Region j... Region m Temporal Decomposition Period 1 Period 2... Period k... Period n Functional Decomposition Function 1 Function 2... Function l... Function o 30

31 Civilian Infrastructure Civilian Infrastructure Telecommunications Elec Power Systems... i... Continuity of Government Defense Infrastructure C3 ISR... j... DII Assets Asset 1 Asset 2... Asset k... Asset m Missions Mission 1 Mission 2... Mission l... Mission n 31

32 Hierarchical Holographic Modeling of Complex Systems of Systems The more we look The more we find The more we find The more comprehensive is our modeling 32

33 Attributes of HHM Adds more realism to the entire modeling process by recognizing the limitations of modeling complex systems of systems via a single model; Provides more responsiveness to the inherent hierarchies of multiple objectives/sub-objectives and multiple decisionmakers. 33

34 Modeling Complex Systems of Systems via HHM The more we look/model The more we find/understand The more we find/understand The more comprehensive is our coverage 34

35 Risk Modeling, Assessment, Management, and Communication for Complex Systems of Systems: We must answer the questions for each system and the SoS as a whole Risk Modeling Causal Relationships Among All State Variables 1.What can go wrong? 2.What is the likelihood that it could go wrong? 3.What are the consequences? 4.What is the time frame? 1.What can be done and what options are available? 2.What are the associated trade-offs in terms of all costs, benefits, and risks? 3.What are the impacts of current management decisions on future options? Risk Communication Stakeholders and Knowledge Integration 35

36 Systems-Based Risk Analysis for the NextGen Integration of the Communication, Navigation, and Surveillance Complex System of Systems A Case Study Performed for the U.S.federal Aviation Administration (FAA) 36

37 Technical Understanding of CNS as System of System (SoS) Flight Management System (FMS) Air Traffic Controller s (ATC) Required Navigational Performance (RNP)_ 37

38 Technical Understanding of CNS as Complex SoS ATN Router System Overview 38

39 RECALL: The key to modeling and managing Complex SoS lies in understanding the genesis of characterizing the interconnectedness & interdependencies manifested through shared: States, decisions, resources, functions, policies, decisionmakers, stakeholders, organizational setups, and others. 39

40 Shared States: Current Configuration The shared states determined are highlighted in Red States Communica-on Naviga-on Surveillance LF Signal Integrity MF Signal Integrity HF Signal Integrity VHF Signal Integrity UHF Signal Integrity (Includes GPS) SATCOM Signal Integrity GPS Signal Integrity GPS Signal Availability GPS Signal Accuracy Radar Signal Integrity Comprehension Plane Authen=city/Iden=fica=on Signal Effec=ve Range Sta?on Authen?city/ Iden?fica?on Distance Calcula?on Accuracy Posi?on DriE Detec?on Availability Naviga?on System Error - NSE Flight Technical Error - FTE (Conformance) 40

41 Shared States: Dynamic Required Navigational Performance (D-RNP) Configuration States Communica-on Naviga-on Surveillance HF Signal Integrity VHF Signal Integrity UHF Signal Integrity SATCOM Signal Integrity GPS Signal Integrity GPS Signal Availability GPS Signal Accuracy Radar Signal Integrity Comprehension Plane Authen=city/Iden=fica=on Signal Effec=ve Range Sta?on Authen?city/Iden?fica?on Distance Calcula?on Accuracy Posi?on DriE Detec?on Availability Compliance 41

42 Shared Decisions Current Configuration Shared Decisions Decision Communica-on Naviga-on Surveillance Tuning of Frequency Human Informa?on Requests System Use INS Correc?on System Call Sign Assignment Flight Plan Al=tude Adjustments Weather Aversion 42

43 Shared Decision Makers Current Configuration Shared Decision Makers Decision Maker Communica-on Naviga-on Surveillance Pilot ATC Airport FAA ICAO ARINC AAC AOC 43

44 Shared Resources- Current Configuration Shared Factors Resource Communica-on Naviga-on Surveillance Radar Sites NAVAID Site Air Traffic Controllers Pilots Airplane Airport Satellites Naviga=on Charts (IAP, etc.) Flight Plans Weather Sta=ons Runways Flight Management Computer 44

45 Shared Decisions/Decision Makers Dynamic Required Navigational Performance (RNP) Configuration Decision \ Decision Maker Pilot ATC Airport Type of Approach Instrument Flight Rules Visual Flight Rules Approach System ILS Vectored Approach GPS (RNAV) Approach VOR Approach NDB Approach Runway Clearance AircraE Separa?on Approach Route Holding PaRern 45

46 Shared Decisions Required Navigational Performance (RNP) Configuration Type of Approach Decision Communica-on Naviga-on Surveillance Instrument Flight Rules Visual Flight Rules Approach System ILS Vectored Approach GPS (RNAV) Approach VOR Approach NDB Approach Runway Clearance AircraS Separa=on Approach Route Holding PaTern 46

47 Shared Resources Dynamic RNP Configuration Resource Communica-on Naviga-on Surveillance Radar Sites DME Substa?ons Satellites Air Traffic Controllers Pilots Airplane NACO Charts Flight Management Computer ATN B2 Receiver/Transceiver NAVAID Receiver 47

48 Fault Tree Analysis 1. Flight Management Control (FMC) Failure 2. Data Link Failure 3. Positioning System Failure Unable to Perform Dynamic RNP FMC Data Link Failure Positioning System 48

49 Fault Tree Analysis Continued Unable to Perform Dynamic RNP Communication Navigation FMC Data Link Failure Positioning System 49

50 Fault Tree Analysis Cont. d Flight Management Control FMC Unable to Perform Dynamic RNP Data Link Failure Positioning System Communication Navigation Fault Tree System Error Monitoring and Alerting Failure Human Error Software Error Software Failure Alerting Failure Monitoring Failure Incorrect Acceptance of Clearance Incorrect RNP Value Input Expanded Fault Tree with Failure Modes Caused by FMC Failure 50

51 Fault Tree Analysis Contued Data Link Fault Tree Unable to Perform Dynamic RNP Communication FMC Data Link Failure Positioning System Navigation ATN-B2 FANS 1/A (ACARS Data Link) On Ground Failure Signal Failure On Plane Receiver/FMC Failure Signal Degradation/ Failure On Plane Receiver Failure On Ground Failure Ground End System ATN Router VDL-3 VDL-2 SATCOM HF SATCOM VHF HFLD(?) FMC Failure Receiver Failure Radio Transceiver Network Failure AFEPS (Central Computer) Datalink Service Provider (a/c link) Expanded Fault Tree with Failure Modes Caused by DataLink Failure 51

52 Positioning System Fault Tree Unable to Perform Dynamic RNP Communication FMC Data Link Failure Positioning System Navigation DME/DME GNSS Accuracy Degradation Loss of Signal Accuracy Degradation Loss of Signal Signal Inteference Intentional Interference Insufficient Power Receiver/ Transceiver Issues Intentional Interference Receiver/ Transceiver Issues Signal Interruption Intentional Signal Blocking Receiver Failure Satllite Failure Intentional Spoofing Receiver Issues Lack of Continuity Satellite Issues Signal Interference Expanded Fault Tree with Failure Modes Caused by DataLink Failure 52

53 What Have We Learned? Throughout this seminar we defined and modeled complexity via the interconnectedness and interdependencies (I-I) characterizing complex systems of systems (SoS). 53

54 We further modeled and quantified the I-I by building on the shared/ common: (i) States; (ii) Decisions; (iii) Resources; and (iv) Decisionmakers. 54