Transcript SSEI
Software Systems Engineering Initiative Sqn Ldr Mike Place Customer Interface Director Prof John A McDermid, FREng Technical Director ADF August 2008 Contents ■ Establishment of the SSEI ■ Current SSEI Programme ■ SSEI Future Developments ■ Conclusions SSEI and Personal Context DE&S DG Safety & Engineering Systems Engineering & Integration Group Hd Future Technology - Sqn Ldr Mike Place SSEI Customer Interface Director Current Responsibilities Future Technology Programme Management ■ Software Systems Engineering Initiative o www.ssei.org.uk ■ Capability Agility ■ Avionic Systems Standardisation Committee o www.assconline.co.uk ■ European Technology Acquisition Programme Provide first POC for: ■ Software issues ■ Standards – Avionic/software ■ Combat Systems - Maritime SSEI Vision “The SSEI is an MOD-funded strategic initiative intended to enhance through-life capability management for software intensive defence systems, and thus to reduce risks, delays and cost overruns.” Lt Gen Andrew Figgures DCDS(EC) Software Engineering is Different ■ Software is ‘pure design’ o No manufacture, no raw material, no cost to replicate & deliver o Potential for rapid, safe upgrade – not yet realised ■ Software problems are not easily absorbed because of o Complexity – unintended side effects o Integration – extensive side effects ■ Software science, technologies methods and tools are not stable or mature – e.g. aerospace: o 70’s, 80’s – Functional design + assembler + custom h/w o 80’s, 90’s – Object Oriented + HOL + COTS h/w o 90’s, 00’s – Model Driven + COTS o/s, h/w Why Software Engineering is Special ■ Software is inscrutable ■ Potential to be disruptive beyond its cash value ■ Range of development standards ■ Enables increase in system complexity Perception of Defence Software Systems ■ Failure to meet specification ■ Expensive ■ Always late Formation of the SSEI ■ BAE Systems (Military Air Solutions) was selected by MOD to hold the contract on behalf of an industry and academia consortium ■ Consortium formed from an amalgamation of all bidders o IBM, MBDA, AeI, QinetiQ, YorkMetrics, SEIC, EDS, Insyte o Universities of York, Newcastle, Loughborough, Oxford ■ Mixture of basic and applied research tasks plus an enabling contract for ‘additional services’ SSEI Management and Governance Joint Advisory Board Joint Working Group Technical Director Customer Interface Director Programme Manager John McDermid Mike Place Jane Fenn Theme Leads Commercial Management Business Manager Tony Dawe SEIG Leads Management IPT BAE SYSTEMS Management Research of Themes Software Systems Projects Academic Industry Tony Powell Barry White Software and Systems Architectures Integration of Software Systems Alan Grigg Alan Brown Steve Riddle John K Davies Development of Dependable Systems Tim Kelly Colin McDonald Properties of Software Systems Jim Davies Colin O’Halloran Systems Engineering Integration Group Richard Parsons Mark Hawksfield Current Status ■ Initial Research Programme o Understanding the problem o Identifying potential solutions ■ Agreed Technical Strategy o MOD endorsement • SIT, DE&S, DSTL o Industry endorsement Next Steps ■ Call for further research proposals o Based on Technical Strategy o Joint Industry/academia ■ Process for additional services o Trial run of process o Establish credibility/value added • Rapid exploitation of research outputs ■ Growth in SSEI membership o Core o Associates What Will Success Look Like? ■ SSEI has established itself o Centre of Excellence in the UK for Software Systems Engineering • • • • Provides leadership Coherent Industry and Academic capability Working beyond Defence A developing body of case studies o Diversity of funding • MOD • Wider Government • Industry o Recognised body for SSE accreditation • Individuals • Organisations o Expanding range of services offered Contents ■ Establishment of the SSEI ■ Current SSEI Programme ■ SSEI Future Developments ■ Conclusions SSEI Research Programme ■ Programme has five themes o Management, especially evidence based o Architecture, including SoS o Integration o Dependability, e.g. safety and security o Properties Initial three year programme funded, with 13 tasks Engagement possible via Special Interest Groups (SIGs) Current Research Tasks 1. IMS for Adaptive Systems (SEIC) 2. Evidence Based Management (York) 3. Evidence-Based Infrastructure (YorkMetrics) 4. Framework for Distributed Development and Integration (SEIC) 5. Model-Driven Integration of Software Systems (York) 6. Software Safety Cases – Establishing a Systematic Approach (York) 7. Dependability Explicit Metadata (Newcastle) 8. Dependable Use of FPGAs (York) 9. Analytic Techniques for the Predictability of Complex Systems (York) 10. Application of SOA to Dependable Systems (IBM) 11. Software Guidance for 00-56 (QinetiQ/York) 12. Software Language Policy (QinetiQ) 13. Managing the Complexity of FPGA Designs (SELEX) Management Theme Special Interest Group MOD, Industry, Academia Task 1: Evidence-based Management Principles, Properties Research to identify a balanced set of measures to support key decisions in software systems acquisition and trade-off analysis. The initial focus is likely to be on general software process measurement Management Theme Task 2: Evidence-based Infrastructure Practices, Frameworks Research to establish an appropriate measurement infrastructure to support decision making, including organisational practices and information sharing. Success Factors Value Added, Collaboration Task n: …. Skills, Risk, Forecasting Maturity, Capability, … The The Problem Solution Engineering Measurement Evidence-based Management Project Measurement • Engineering Performance • Project Performance • Earned Maturity • Earned Value • Proactive • Reactive Courtesy Tony Powell In more detail … Sources of ignition Smoke alarms Fire alarms Fires Engineering Capability Engineering Performance Engineering Status Financial Indicators Estimate with uncertainty £ * Mar Apr May Jun Jul Aug Sep S W BC P W AC WP BC Oct Time Causes Need to monitor drivers and pull control levers Consequences Performance not meeting plans Product not maturing fast enough Courtesy Tony Powell Behind schedule, unpredictable SSEI and Continued Airworthiness ■ Several relevant tasks o Software Safety Cases – Establishing a Systematic Approach (Task 6) o Software Guidance for DS 00-56 (Task 11) o Dependable use of FPGAs (Task 8) o Evidence-Based Management (Task 2) o …. Software Guidance for DS 00-56 ■ DS 00-56 Issue 4 is goal-based o Admits “appropriate” evidence and argument ■ For continued airworthiness, build on o o o o Evidence of operation (operational data) Evidence from development (where available) Evidence of change management Guidance will address combination, balance, … Other Tasks ■ Software safety cases o Longer term view ■ Evidence-based Management o Informed decision-making, reflecting balance of risk and benefit ■ Dependable use of FPGAs o Including processor replacement Contents ■ Establishment of the SSEI ■ Current SSEI Programme ■ SSEI Future Developments ■ Conclusions SSEI Development ■ Second phase of MoD research funding o Balance between themes o Address omissions o Open call, probably late third quarter ■ Additional services o Technical support, e.g. assurance, mentoring o Initially high integrity and safety critical applications Technical Strategy ■ Comprehensive analysis of o Defence requirements, e.g. defence industrial strategy • “Predictability of development time and cost should be accurate to 10%” • “New systems should be developed in the same time and cost, despite a predicted 5-fold increase in complexity” o Industrial needs, and assessment of industrial software development trends o Capabilities and tasks for each theme Focus for Technical Strategy ■ Technical Strategy comprehensive, but too extensive to fund, hence need for prioritisation ■ So-called “hard-problems” o Challenges which will deliver military benefit, if the research is successful o Identified route through to exploitation o Give focus for work in the five themes Hard Problems ■ Focus on six “problems” (TBC) o Supporting legacy o Secure system interoperation o Safety assurance at military tempo o Assurance of open systems o Model-based development o Integrated engineering management All have a technical and managerial element Fragment of Prioritisation Table Problem Supporting Legacy Secure System Interoperation Success Criteria Ability to preserve operational capability by providing ongoing support for legacy software Ability to assure secure services across different management domains Customers SENTRY, AH DG Air Systems, Ships, Subs Collaborators MODELPLEX ITA, SyIOP , DHS SwA initiative, ESII Management Prediction of Attributes Planning Engineering Management Product Maturity Assessment Prediction of Attributes Requirements Management Planning Engineering Management Product Maturity Assessment Architecture Unplanned Evolution Predictable Design Control Emergent Properties Unplanned Evolution Predictable Design Control Emergent Properties Supporting Legacy ■ Capabilities include o Legacy integration o Control/predicting emergent properties o Dependability/property trade-offs o Assessment and acceptance o Product maturity assessment o Planning engineering management Some new tasks, contingent on next call … Additional Services ■ Offering to be developed, covering o o o o o Acquisition support Training and mentoring Development of guidance Specialist input to MoD policy Assistance with technology transfer of non-SSEI research results (capabilities) Focus on high value-added activities Long Term Objective ■ After initial research programme o Stable independent organisation • Not dependent on core MoD funding o SSEI a key player in a global network • Focus for work in the UK • Arbiter of quality, but not sole source for advice o International collaboration, e.g. DMO o Remit broader than defence, e.g. OGD Contents ■ Establishment of the SSEI ■ Current SSEI Programme ■ SSEI Future Developments ■ Conclusions Conclusions ■ SSEI focused on an important problem o Need for national capability o But must work internationally, e.g. SEI & DMO • Engagement with SSEIAP ■ SSEI has many challenges, especially o Building critical mass o Achieving independent status Questions? Contacts: [email protected] [email protected] ADF August 2008