Main Heading/Title - Department of Defence

Download Report

Transcript Main Heading/Title - Department of Defence

Software Engineering in
the DMO: Where to from
here?
Matt Ashford
A/Director of Software Engineering
Electronic Systems Division (ESD)
professionalise | re-prioritise | standardise | benchmark | improve industry relationships and industry performance | lead reform
Opening Assertions
• Defence systems are getting more connected and increasingly
software intensive, and software is the critical element
• Engineering is fundamental to DMO’s core business, and is essential
to successfully acquire, develop and sustain complex Defence systems
In my opinion…
• The DMO can support its’ software engineers and improve its’ software
engineering capability through:
– Policy, Procedure & Guidance
Across SPO’s,
Branches,
– Training
Divisions and
– Knowledge management & collaboration
Groups.
– Peer networks and communication
– Career structures
– Functional experts (Grey Beards), mentoring, coaching
– Systems & Software Engineering related research & development
Directorate of Software Engineering (DSwE)
•
•
Mission:
– To sustain and improve ESD Software Engineering (SwE) capability and
performance
Objectives:
– Improve ESD SwE capability via:
• ESD Policy, Procedure and Guidance
• Knowledge sharing and collaboration
• SwE related research & development
– Improve ESD SwE performance via:
• Direct SwE support to ESD projects and SPOs
• Training
• Networks and communication
– Represent ESD SwE:
• ESD Continuous Improvement Group
• Across DMO Divisions
• DMO Materiel Engineering Council (MEC)
• ADF Technical Regulatory Agencies
• External groups
Proposed Work Plan (08/09)
• Policy, Procedure & Guidance
– State-of-the-Practice Survey
– Software Configuration Management
• Training
– DAU Software-Intensive Systems Acquisition Management (SiSAM)
Training
• Peer Networks & Knowledge Sharing
• Systems & Software Assurance
• Collaboration & Research
– Systems and Software Engineering Research Capability
State-of-the-Practice Survey
• Need to understand current state-of-the-practice of software
engineering in ESD, e.g.:
– How much and what type of software does ESD acquire & sustain?
– Who is doing what software related activities (SPOs, People)?
– How well are the current software Policies, Procedures and
Guidance being implemented?
• Software lists, PSM etc
– What are the biggest issues for our software practitioners?
• Survey will:
– Collect data
– Identify improvement opportunities
– Drive future DSwE initiatives
– Inform future research agenda
Software Configuration Management
• SCM identified as an immediate, systemic issue
– Especially for software-intensive support environments
• Broad, high level CM Policy available
• Lack of detailed implementation guidance and tool support
– Multiple software applications
– Different versions across multiple deployment environments
– Mixture of COTS and bespoke
• SCM task to work with most affected SPO’s to:
– Review current Defence & DMO Policy, Procedure & Guidance
– Identify SPO SCM needs
– Develop implementation level plans & procedures
– Develop & roll out ESD SCM Policy, Procedure &/or Guidance
• SCM stakeholder group / special interest group?
– Contact Matt Ashford if you want to be involved
Software Training
• Most DSWAR initiated training now discontinued
– Australian Software Professional Development Program (ASPDP)
– ISO 12207
– Practical System and Software Measurement
– SiSAM
• DMO Institute
– Plethora of Project Management and Logistics training
– Very little software related training
• DAU Software-Intensive Systems Acquisition Management (SiSAM)
training
– Planned for Dec 08
• Training Needs Analysis required to develop comprehensive curriculum
Peer Networks & Knowledge Sharing
• Few (if any?) technically focused DMO peer networks
– Are you aware of the “DMO Forum”?
• DMO Intranet Home → Tools and Resources → DMO Forum
• Very few collaborative knowledge sharing environments
(e.g. Wiki)
…and even fewer opportunities to interact
• Rely on external Centre’s of Excellence
– E.g. DGTA SCI
• Sometimes peer level support would suffice
– We have a lot of smart people
– A lot of problems have probably already been solved!
• Pockets of excellence already exist in DMO
– Just need to find them (when you need them)
System and Software Assurance
• A definition
– The level of confidence or trust that systems are free from
vulnerabilities, either intentionally designed into the system or
accidentally inserted at any time during its lifecycle and functions in
the intended manner (nothing more or nothing less) even in the
face of malicious activity.
• Exploitable vulnerabilities are an increasing concern:
– Outsourcing and Globalisation
– Prevalence of COTS, Open Source Software (OSS) and other
Software of Unknown Pedigree (SOUP)
• 2 types of problems
– Unintentional defects or vulnerabilities
• Majority of safety space
– Intentional, malicious attack or insertion (small minority)
• Includes the malicious use of pre-existing (innocent)
vulnerabilities
System and Software Assurance
•
•
•
The Problem:
– Current systems and software engineering and security accreditation
processes and techniques may not be adequate to assure the security of
modern software-intensive Defence systems
– (Perceived) lack of policy, procedure and guidance to address SA over the
full system lifecycle
– (Perceived) immaturity of SA related tools, technology and methodologies
– (Perceived) lack of awareness of potential threats and mitigation strategies
(especially with respect to systems operating within the deployed
environment)
– The multiple stakeholders (including authorities) with poorly defined
boundaries (especially with respect to systems operating within the
deployed environment)
Defence is not able to confidently state or defend the level of System
Assurance (with limited exception) in products, systems or capability currently
in acquisition, sustainment or in-service within the deployed environment.
Defence is not able to have a sufficient level of confidence in procured
equipment, systems and capabilities with respect to hosting undocumented
features.
System and Software Assurance
• Technical Regulatory Authorities concerned with item’s fitness for
service, safety and compliance with regulations for environmental
protection.
– Lots of focus on safety: relatively mature discipline
• System Assurance less mature
– Awareness of potential threats, vulnerabilities and mitigation
strategies is poor
– Lack of coherent policy and regulatory environment
• Fixed Vs Deployed
• White Vs Black
– Immature methodologies
• How often does DMO require Vulnerability Assessments or
Secure Coding Practices?
• Collaborating with US and UK
Is this different to System Safety?
• Similar methodologies / processes
– Different perspectives
• Safety and System Assurance Cases both contain
– Claims about a product or service (e.g. no exploitable buffer
overflows)
– Arguments (e.g. use static tool analysis)
– Evidence (e.g. test reports)
– Both reduce uncertainty, but not to zero
• Can be expensive, but best done from the beginning, retrospective
fixes are very expensive
• Everyone happy to share best practice with safety
– Do not always want to publicly share best practice for system and
software assurance
• Opportunity to leverage (or extend) extant safety practices
Rapid Prototype,
Development & Evaluation
•
RPDE Steering Gate Process
RPDE is:
– a collaborative venture
between the Defence
and industry
•
RPDE Operating Model
A RPDE Task is:
– A method to formally engage
industry to investigate a Defence
problem
– Around 12 -18 months duration
www.rpde.org.au
RPDE Task 24 –
System Assurance
• DMO has sponsored a RDPE Task to investigate SA problem
– CIOG now co-sponsor
• Question:
How can Defence ensure that deployable systems and assets have
an appropriate level of understood and quantifiable confidence that
the systems are assured and remain so and thus will continue to
perform in the manner expected even in the face of malice?
• Current status:
Entry Criteria
Question
Review
Development
SG1
Discovery
SG2
Options
Development
08 May 08 12 Aug 08
SG3
Solution
Development
SG4
Implemen
tation
Support
SG5
• Contact Matt Ashford for more info or if you want to get involved
Systems & Software Collaboration / Research
Government
MoD
DoD
DoD
DE&S
DMO
AT&L
D-SET
HENG
DSSE
SISAIG
US/UK/AUS Software-Intensive Systems
Acquisition Improvement Group (SISAIG)
• Established circa 2003
• Facilitates information sharing and collaboration across
US/UK/AUS Governments
• Leverage each Nations’ initiatives, R&D, etc
• Aus Sponsor: Ms Shireane McKinnie
• Aus Lead: Mr David Marshall
• Current Work Streams:
– Systems Assurance
– Systems of Systems (SoS) / Network-Centric Warfare
(NCW)
– Software Estimation & Earned Value Management
(EVM)
Systems & Software Collaboration / Research
Government
Software
Research
MoD
DoD
DoD
DE&S
DMO
AT&L
D-SET
HENG
DSSE
SISAIG
DMO Systems and Software Engineering
Independent Advisory Panel
• Established by DMO Head Engineering (Shireane
McKinnie) to:
– Provide strategic guidance on future trends and
challenges for the DMO
– Independent expert advice on emerging Systems &
Software Engineering issues
– Identification of key developmental / emerging
technologies
– Facilitate technology transfer to the DMO
• E.g. NICTA, ACCS, CSSE, SEI, UK SSEI etc
• Linkages to academic and research programs
• 5 “Core” Members + 2 “Consultative” members
DMO SESW Independent Advisory Panel –
Core Members
Prof. Geoff Dromey
Foundation Professor of Software Engineering,
Griffith University
Director, Software Quality Institute
ARC Centre for Complex Systems
Dr Clive Boughton
Senior Lecturer,
Australian National University
Prof. Ross Jeffery
Professor of Software Engineering,
UNSW
Program Leader, Empirical Software Engineering
Research Program
National ICT Australia (NICTA)
Prof. Peter Lindsay
Boeing Professor of Systems Engineering,
Uni of Queensland
ARC Centre for Complex Systems
Prof. Stephen Cook
Director, Defence & Systems Institute
Director of Centre of Expertise in Systems Integration
University of South Australia
DMO SE and SW Independent Advisory
Panel – Consultative Members
Prof. Barry Boehm
TRW Professor of Software Engineering, University of Southern California
(USC)
Director, USC Center for Systems and Software Engineering.
Prof. John McDermid
Professor of Software Engineering, University of York
Leader of the High Integrity Systems Engineering Group (HISE)
Member of the UK MoD Defence Scientific Advisory Council (DSAC)
SEI Australia Initiative
• Initiative to establish an SEI research facility
in Australia
• Co-funded by SA Govt and CoA (through
SADI)
• Outcome of failed negotiations was a
recommendation to explore alternate
strategies to establish indigenous capability
SSE Research Proposal
•
•
Premise:
– There is a need for greater sponsorship and coordination of
Systems & Software Engineering research to improve the
acquisition, development and sustainment of Defence systems.
– SEI Australia & SESW Panel shows there is a willingness to
support such a concept
Consultative workshop conducted 14/15 Aug 08 in Canberra
– To elicit information to help develop a proposal to establish a cost
effective, indigenous Defence-related systems and software
engineering research capability.
– Key questions:
• What are Defence’s research interests?
• What research capability is already available (Domestic /
International)?
• What models / contract mechanisms can we use?
– What are their strengths & weaknesses?
– What commercial aspects do we need to address?
Workshop
Attendees
Barry Boehm
 TRW Professor of Software Engineering, University of Southern California (USC)
 Director, USC Center for Systems and Software Engineering (CSSE)
John McDermid
 Professor of Software Engineering, University of York
 Leader of the High Integrity Systems Engineering Group (HISE)
 Member of the UK MoD Defence Scientific Advisory Council (DSAC)
 Technical Director, Software Systems Engineering Initiative (SSEI)
Rich Turner
 Distinguished Service Professor, Stevens Institute of Technology
 Visiting Scientist, Software Engineering Institute (SEI)
Geoff Dromey
 Foundation Professor of Software Engineering, Griffith University
 Director, Software Quality Institute
 ARC Centre for Complex Systems
Ross Jeffery
 Professor of Software Engineering, University of NSW (UNSW)
 Program Leader, Empirical Software Engineering Research Program, National ICT Australia
(NICTA)
Stephen Cook
 Director, Defence & Systems Institute (DSI)
 Director, Centre of Excellence in Defence and Industry Systems Capability (CEDISC)
 University of South Australia
Clive Boughton
 Senior Lecturer, Australian National University (ANU)
Gary Millar
 Lecturer, University of NSW @ Australian Defence Force Academy (ADFA)
David Marshall
 Director General COMMS Branch, DMO/ESD
 Deputy DMO Head Engineer
Adrian Pitman
 Deputy Director HFMOD, DMO/ESD
Matt Ashford
 Director Software Engineering, DMO/ESD
Thea Huber
 Director Strategy & Standardisation, DMO/ESD
Shari Soutberg
 Director Systems Engineering DMO/ESD
Stuart Garrett
 Project Management Coach DMO/DCEO
Thea Clark
 Head Future Force Integration, DSTO/Joint Ops
Todd Mansell
 Research Leader Surface Ship Ops, DSTO / Maritme Ops
Clive Wormsley
 DSTO Representative, RPDE
Outcomes of Workshop
• Summarise key outputs of workshop and way ahead.
Questions?
Matt Ashford
A/Director Software Engineering
Electronic Systems Division (ESD)
R3-3-121
Russell Offices
Canberra ACT 2600
Ph: (02) 6266 7054
[email protected]
Don’t forget
the survey!