Systems of Systems: Characteristics and Challenges Barry Boehm, boehm@usc.edu USC Center for Systems & Software Engineering http://csse.usc.edu October 25, 2006 Overview Definitions, Examples & Motivation SoS Characteristics and Challenges Conclusions 2 What is a “System-of-Systems”? Very large systems developed by creating a framework or architecture to integrate component systems SoS component systems independently developed and managed – New or existing systems Some outsourced Some externally evolving – Have their own purpose – Can dynamically come and go from SoS SoS exhibits emergent behavior not otherwise achievable by component systems SoS activities often planned and coordinated by a Lead System Integrator (LSI) Typical domains – Business: Enterprise-wide and cross-enterprise integration to support core business enterprise operations across functional and geographical areas – Military: Dynamic communications infrastructure to support operations in a constantly changing, sometimes adversarial, environment 3 What Does an SOS Look Like: Future Combat Systems 4 The Need for Net-Centric Systems of Systems (NCSOS) Lack of integration among stove-piped systems causes – Unacceptable delays in service – Uncoordinated and conflicting plans – Ineffective or dangerous decisions – Inability to cope with fast-moving events Increasing NCSOS benefits – See first; understand first; act first – Network-centric operations coordination – Transformation of business/mission potential – Interoperability via Integrated Enterprise Architectures 6 Overview Definitions, Examples & Motivation SoS Characteristics and Challenges Conclusions 7 SoS Characteristics and Challenges - I Complexity: Size and number of organizations, interfaces, suppliers, coordination groups – Scalability of processes, methods, and tools Dynamism: Number of changes/month, average time to process change – Rapid change impact analysis, change synthesis, negotiation, development, validation, implementation Emergence: requirements not pre-specifiable Build-to-spec processes, contracts infeasible – C2ISR a better metaphor for SoS acquisition than purchasing-agent – Command, control, intelligence, surveillance, reconnaissance 8 Complexity of SoS Solution Spaces Size: 10-100 MLOC Number of external interfaces: 30-300 Number of “Coopetitive” suppliers: 20-200 – Even more separate work locations Depth of supplier hierarchy: 6-12 levels Number of coordination groups: 20-200 – Reviews, changes, risks, requirements, architecture, standards, procedures, technologies, -ilities, integration, test, deployment, personnel, infrastructure, COTS,… – Key personnel spend 60 hours/week in meetings Unprecedentedness Emergence Rapid change Necessarily software-intensive… 9 Average Change Processing Time: 2 SoSs Average workdays to process changes 160 140 120 100 80 60 40 20 0 Within Groups Across Contract Groups Mods 10 Acquisition C2ISR Via Spiral OODA Loops Observe new/updated Example: objectives, constraints, alternatives ARPANet/Internet Spiral • Usage monitoring • Competition, technology, marketplace ISR Orient with respect to stakeholders priorities, feasibility, risks • Risk/Opportunity analysis • Business case/mission analysis • Prototypes, models, simulations Operate as current system Accept new system Act on plans, specifications Decide on next-cycle capabilities, architecture upgrades, plans • Keep development stabilized • Change impact analysis, preparation for next cycle (miniOODA loop) • Stable specifications, COTS upgrades • Development, integration, V&V, risk management plans • Feasibility rationale Life Cycle Architecture Milestone for Cycle 11 SoS Characteristics and Challenges - II COTS/Reuse/Legacy diversity – Many sources of incompatibility, changes – COTS: average 8-10mo/release; unsupported after 3 releases Multiple missions and stakeholders to support – Increment and change content requires negotiation Independently evolving systems – Often with “coopetitive” suppliers, interoperators More time needed for systems definition – Before and after source selection More time needed for teambuilding, partner coordination, supplier management, change management , integration and test 12 How much Architecting is Enough: COCOMO II Analysis 100 Percent of Time Added to Overall Schedule 90 10000 KSLOC 80 Percent of Project Schedule Devoted to Initial Architecture and Risk Resolution 70 Added Schedule Devoted to Rework (COCOMO II RESL factor) Total % Added Schedule 60 Sweet Spot 50 40 100 KSLOC 30 Sweet Spot Drivers: 20 Rapid Change: leftward 10 KSLOC High Assurance: rightward 10 0 0 10 20 30 40 50 60 Percent of Time Added for Architecture and Risk Resolution 13 COSOSIMO: I&T Sub-Model Size Drivers • # SoS interface protocols • # SoS scenarios • # unique component systems Cost Drivers • • • • • • • • • Requirements understanding Architecture maturity Level of service requirements SoS team capability Maturity of LSI processes Tool support Cost/schedule compatibility SoS risk resolution Component system maturity and stability • Component system readiness SoS Integration and Testing LSI I&T Effort 14 Conclusions Individual SoS attributes are highly challenging – Complexity, dynamism, emergence, uncontrollables, stakeholder diversity – Their combinations are even more challenging Acquisition management and negotiation skills are at least as important as systems engineering skills – C2ISR a better metaphor for SoS acquisition than purchasing-agent More time needed for systems definition – Before and after source selection 15 Backup Charts Complexity of Solution Spaces - Breadth, Depth, and Length Platform N • Width • • Platform 1 Infra C4ISR 1.0 2008 2.0 3.0 4.0 5.0 2010 2012 2014 2016 Command and Control Situation Assessment Info Fusion Sensor Data Management Sensor Data Integration Sensors Sensor Components : Depth … Length Legend: DOTMLPF C4ISR Doctrine, Organization, Training, Materiel, Leadership, Personnel, Facilities Command, Control, Communications, Computers, Intelligence, Surveillance, and Reconnaissance 17 Top-10 Risks: Software-Intensive Systems of Systems - CrossTalk, May 2004 1. Acquisition management and staffing 2. Requirements/architecture feasibility 3. Achievable software schedules 4. Supplier integration 5. Adaptation to rapid change 6. Quality factor achievability and tradeoffs 7. Product integration and electronic upgrade 8. Software COTS and reuse feasibility 9. External interoperability 10. Technology readiness 18 Effect of Unvalidated Requirements -15 Month Architecture Rework Delay $100M Arch. A: Custom many cache processors $50M Arch. B: Modified Client-Server Available budget Original Spec 1 After Prototyping 2 3 4 5 Response Time (sec) 19 Effect of Unvalidated Software Schedules Original goal: 18,000 KSLOC in 7 years – Initial COCOMO II, SEER runs showed infeasibility – Estimated development schedule in months for closely coupled SW with size measured in equivalent KSLOC (thousands of source lines of code): – Months =~ 5 * 3√KSLOC - KSLOC 300 1000 3000 10,000 - Months 33 50 72 108 •Solution approach: architect for decoupled parallel development; Schedule As Independent Variable (SAIV) process 20 The SAIV* Process Model 1. Shared vision and expectations management 2. Feature prioritization 3. Schedule range estimation and core-capability determination Top-priority features achievable within fixed schedule with 90% confidence 4. Architecting for ease of adding or dropping borderline-priority features And for accommodating past-IOC directions of growth 5. Incremental development Core capability as increment 1 6. Change and progress monitoring and control Add or drop borderline-priority features to meet schedule *Schedule As Independent Variable; Feature set as dependent variable. Also works for cost, schedule/cost/quality as independent variable. 21 Top-10 Risks: Software-Intensive Systems of Systems CrossTalk, May 2004 1. Acquisition management and staffing 2. Requirements/architecture feasibility 3. Achievable software schedules 4. Supplier integration 5. Adaptation to rapid change 6. Quality factor achievability and tradeoffs 7. Product integration and electronic upgrade 8. Software COTS and reuse feasibility 9. External interoperability 10. Technology readiness 22 COTS Upgrade Synchronization and Obsolescence Many subcontractors means an increasing number of evolving COTS interfaces Aggressively-bid subcontracts can deliver obsolete COTS – New COTS released every 8-9 months (GSAW) – COTS unsupported after 3 releases (GSAW) – An actual delivery: 120 COTS; 46% unsupported Emphasize COTS interoperability in source selection process Develop contract/subcontract provisions/incentives to ensure – Consistency and interoperability across contractor and subcontractordelivered COTS software products – Such products are recent-release versions Develop a management tracking scheme for all COTS software products in all NCSOS software elements Develop a strategy for synchronizing COTS upgrades 23 Emerging Scalable Spiral Process - For 21st century systems engineering and acquisition The C4ISR Metaphor for NCSOS Acquisition – Role of OODA loops – Example: Internet Spiral – Example: FCS Win Win Spiral Model Risk-Driven Scalable Spiral Model – Increment view – Life-cycle view – Role of anchor point milestones Acquisition management implications People management implications 24 Risk-Driven Scalable Spiral Model: Increment View Rapid Change Short Development Increments Foreseeable Change (Plan) Increment N Baseline Short, Stabilized Development of Increment N Increment N Transition/O&M Stable Development Increments High Assurance 25 Risk-Driven Scalable Spiral Model: Increment View Unforseeable Change (Adapt) Rapid Change Short Development Increments Agile Future Increment Baselines Rebaselining for Future Increments Deferrals Foreseeable Change (Plan) Increment N Baseline Stable Development Increments Current V&V High Assurance Resources Short, Stabilized Development of Increment N Artifacts Increment N Transition/O&M Concerns V&V of Increment N Future V&V Resources Continuous V&V 26 Risk-Driven Scalable Spiral Model: Life Cycle View System LCA System Inception System, DI1 LCA System Elaboration DI2 B/L LCA Changes Agile DI2 (OO&D) Rebaselining Plan-Driven DI1 Construction (A) DI1 V&V LCA: Life Cycle Architecture IOC: Initial Operational Capability OO&D: Observe, Orient and Decide V&V: Verification and Validation DI: Development Increment B/L: Baselined DI2 LCA Plan-Driven DI2 Construction (A) DI2 V&V 27 Risk-Driven Scalable Spiral Model: Life Cycle View System LCA System, DI1 LCA System Inception DI2 B/L LCA DI3 B/L LCA DI4 B/L LCA Changes System Elaboration Agile DI2 (OO&D) Rebaselining Plan-Driven DI1 Construction (A) DI1 V&V Changes Update Update DI1 IOC DI1 Trans’n DI1 Usage DI2 LCA Agile DI3 (OO&D) Rebaselining Plan-Driven DI2 Construction (A) DI2 V&V Changes Update DI2 IOC DI2 Trans’n DI2 Usage DI3 LCA Agile DI4 (OO&D) Rebaselining LCA: Life Cycle Architecture IOC: Initial Operational Capability OO&D: Observe, Orient and Decide V&V: Verification and Validation DI: Development Increment B/L: Baselined Plan-Driven DI3 Construction (A) DI3 V&V DI3 IOC DI3 Trans’n DI3 Usage . . . DI4 LCA ... 28 LCO (MS A) and LCA (MS B) Anchor Points Pass/Fail Criteria A system built to the given architecture will – Support the operational concept – Satisfy the requirements – Be faithful to the prototype(s) – Be buildable within the budgets and schedules in the plan – Show a viable business case – Establish key stakeholders’ commitment to proceed LCO: True for at least one architecture LCA: True for the specific life cycle architecture; All major risks resolved or covered by a risk management plan 29 Spiral Feasibility Rationale Deliverable LCO, LCA reviews not just UML/PowerPoint charts Need to show evidence of product and process feasibility Evidence provided by prototypes, production code, benchmarks, models, simulations, analysis – Sizing and cost/schedule model results for process feasibility Evidence provided in advance to LCO/LCA review team – Key stakeholders, specialty experts Lack of evidence risks destabilizing the process – Needs coverage by viable risk mitigation plan Key new progress metric – Feasibility evidence progress vs. plans 30 Acronym Definitions B/L C4ISR Baselined Command, Control, Communications, Computers, Intelligence, Surveillance, and Reconnaissance CCD Core Capability Drive-Through COTS Commercial Off the Shelf CRACK Collaborative, Representative, Authorized, Committed, Knowledgeable DI Development Increment DODAF Department of Defense Architectural Framework DOTMLPF Doctrine, Organization, Training, Materiel, Leadership, Personnel, Facilities ERP Enterprise Resource Planning FEAF Federal Enterprise Architectural Framework GSAW Ground System Architectures Workshop IESG Internet Engineering Steering Group IETF Internet Engineering Task Force IKIWISI I’ll Know It When I See It IOC Initial Operational Capability IPT Integrated Product Team IRR Inception Readiness Review LCA LCO LSI MLOC MS NCSOS OO&D OODA PM PRR SAIV SE SoS SOW SW Sys Engr V&V WMI Life Cycle Architecture Life Cycle Objectives Lead System Integrator Million Lines of Code Milestone Net-Centric System of Systems Observe, Orient, and Decide Observe, Orient, Decide, Act Person-Month/Program Manager Product Release Review Schedule As Independent Variable System Engineering System of Systems Statement of Work Software System Engineering Verification and Validation War-fighter machine interface 31