incose uk spring symposium 2001 pgr systems engineering - in retrospect and in prospect peter robson...
TRANSCRIPT
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
Systems Engineering - In Retrospect and In
Prospect
Peter RobsonSystems Engineering Consultant
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
Overview of Paper
• Introduction
• Generic Systems Engineering
• Systems Engineering Maturity – Performance
– Methodology
– Focus
– Life Cycle
• In Retrospect
• In Prospect
• Conclusions
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
Disclaimers
• My views only!
• Examples come primarily from defence background
– but read across / translate to other sectors
• Examples are mostly end customer – prime supplier
– but don’t forget that prime suppliers are customers down
the supply chain
• Not an analysis of thirty years of systems theory
– but subjective identification of key issues
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
Introduction
• Resurgence of national interest in systems engineering during the last decade
• DTI’s Foresight Systems Engineering Working Party
• MOD’s initiative on ‘Smart’ Procurement’ • Importance of systems engineering to other
sectors of industry • Signs that the resurgence has not taken
sufficient root• Systems engineering is a shared interest
throughout the whole supply-and-use chain for a project or product
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
Thesis
“Complementary systems engineering capabilities are needed
within and between the various enterprises in the supply chain, in particular between customers and
suppliers”
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
Generic Systems Engineering
• What is a System? – “A system is an interacting combination of elements,
viewed in relation to function” – the human element– emergent properties
• What is Systems Engineering? – “Systems engineering is the art and science of creating
complex systems” (Hitchins)
– Discovery, a specialist type that involves significant analysis, particularly of the problem space (Sheard)
– Program Systems Engineering, a co-ordination or generalist type that emphasises the solution space and technical and human interfaces (Sheard)
– Approach, a process type that can (and should) be performed by any engineer (Sheard)
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
Generic Systems Engineering
• The Human Foundation of Systems Engineering– Propositions about Humans and Problems– Propositions about Knowledge – Propositions about Systems Engineering
• Systems Hierarchies
Constituents of a system
Systems engineering hierarchy
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
Structure based on SE Maturity Model
based on Brian Mar’s “Systems Engineering Maturity Model”NCOSE 1993
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
Performance
Have we just promised to provide
SE (Level 1) or have we demonstrated
it (Level 2) or have we made it
repeatable (Level 3) or have we already
improved our (repeatable) systems engineering process (Level 4)?
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
The Best of Cost-Plus :The heady days of Ptarmigan
• Succinct top-level User requirement• Strong MOD/SRDE/Industry team built-up
through earlier studies and demonstrator activities
• Green-field approach to management and technical processes (top-down)
• Time to think first then act• Trade-offs ‘round the table’ - an integrated
project team!• How do we recapture the best of this?
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
The Worst of Fixed-Price
• The requirement is assumed complete and fixed• Implementation defined as part of the
requirement• Lip-service paid to whole-life costs• Jump off the cliff if you want to win
– Don’t bother to explain the potential problems and how they’ve been allowed for in the price
– The customer won’t be impressed, he’s looking for someone that doesn’t have problems
• Sort it out after you’ve won– You’re bound to make a loss reworking and retesting– The customer will slag you off because of late delivery– Next time, someone else can jump!
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
Becoming Smarter?
• Match the chosen procurement approach to the feasibility of the systems engineering involved – A cry for “faster, cheaper, better” in accordance with
the principles of systems engineering is unlikely to be successful unless complemented by other changes
• Recognise precedented ‘project systems’ should be delivering the ‘unprecedented systems’
• Complete the bridges between – systems engineering and project management– customer and supplier
• Customers must have a full systems engineering capability– education and training in SE process, methods and
tools
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
Laws & Principles
• Foresight SEWP Report
– Build the right system
– Build the system right
• EIA/IS 632 Standard
– The right things should be done
– The right things should be done at the right time
– The right things should be done by the right people
– The right things should be done right the first time
• Simple, even trivial, statements but often
forgotten and difficult to achieve
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
Methodology
Do we think SE is just documents (Level 1)
or about producing a self-consistent set of
requirements, functions and architectures (Level 2) or are we able to simulate our systems dynamically before trying to build them (Level 3) or have we done all this and
have now optimised our approach (Level 4)?
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
Methodology
• Level 1 - documents, documents and more documents• Level 2 - process together with recognition that SE is
about ‘joined up information’ – base on needs design analysis not requirements
analysis design
– ‘process’ not ‘lifecycle’
– but more than ‘process’, more than ‘information’ - emphasis on design or architecting
– systems engineering plans
– systems engineering tools
– ‘tribalism’ becomes more visible
• Level 3 - dynamic interactions between system elements – discover emergent properties sooner!
• Becoming smarter - does concurrency help? – handle increased concurrency in the ‘project production
system’ with care!
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
Focus
Are we doing SE on our products only for
those customers that have demanded it (Level 1)
or do we apply SE to the process as well but again only at customer demand
(Level 2) or are we mature enough to apply SE to all our products
because we believe in SE as a contractor (Level 3) or do we apply it to our
processes as well (Level 4)?
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
Focus
• Dimension reflects a contractor’s motivation or enlightened self-interest
• Focus metric is defined as two state pairs – is systems engineering driven by the customer or the
producer (contractor)?– is systems engineering used to develop the product or
the process?
• Level 1 harks back to the days when the contract would stipulate the applicable standards
• Relates to the role of systems engineering standards and maturity models
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
External standards - just what are they for?
• Provide an external and believable reference– rather like a consultant, only impersonal and cheaper!
• Level the playing field for procurers and suppliers
• Capture of best practice
• Avoid inappropriate practice– e.g. misapplication of software-based systems approaches to
levels of systems well above software
• Establish a common language– eliminate the need for the SE Babel Fish!
• Aid to the conduct of Program Systems Engineering– not so useful for mandating that “the fundamental concepts
of generic systems engineering shall be followed”
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
Key Issues for SE Standards
• Essence of SE v detailed systematic engineering• Needs Design Analysis rather than
Requirements Analysis Design?• Process v Life Cycles?• Synthesis v Decomposition?• Hard (Products) or Hard Soft (Services)?• SE component of SW standards?• Project management - in or out?• Humans - in or out?• Software engineering - in or out?
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
Capability Maturity Models - do they help?
• EIA 731 SECM – can really help an organisation assess its capability to
carry out systems engineering – not a standard for systems engineering, or for a
systems engineering process – encourages systematic engineering in a more capable
way
• EIA 731 status and intended use
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
EIA 731 status and intended use
“This Interim Standard is intended solely to be used for self-development, self-improvement, and self-appraisal. Organizations should not apply this Interim Standard to suppliers as a means of source selection or as a
means of qualification to be a supplier.”
EIA 731 Part 1 Section 2.3
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
Capability Maturity Models - do they help? • EIA 731 SECM
– can really help an organisation assess its capability to carry out systems engineering
– not a standard for systems engineering, or for a systems engineering process
– encourages systematic engineering in a more capable way
• EIA 731 status and intended use– but customers using SECM in source selection to help give
confidence in the outcome of a resultant contract
• Pursuit of ‘Focus maturity’ – mixture of self-assessment and third-party assessment on an
annual cycle
• Customers need to apply SECM to themselves – SE is not done only by suppliers
– SE is a critical capability for customers who want their projects to succeed
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
The tyranny of requirements
• The apparent inability of customers to ‘focus’ on the wider aspects of systems engineering is a major cause of project failure
• Customers must aspire to more than ‘systematic requirements engineering’ – even though this contains much good practice – e.g. establish a minimal and abstracted set of project
requirements, fully traceable and verifiable and which do not constrain the creativity of both customer and supplier
• Over-emphasis on functional requirements at the expense of non-functional requirements – desired ‘emergent properties’ often relate to NFRs
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
Requirements and Options
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
The tyranny of requirements- quote from a past INCOSE President
“If someone is trying to contract for a system, and they can properly identify all the necessary ‘requirements’, then it makes sense to do so. The preference then usually becomes: “meet the requirements, and pick the lowest cost.” But the reality is, in every complex system I've seen, “most ‘requirements’ aren't.” Given the right combination, nearly any ‘requirement’ will be relaxed to obtain some other gain. The real preferences are hidden behind the ‘requirements’ in some operational analysis space. This is the ‘real’ reason behind the recent Acquisition Reform initiatives to contract based on performance specifications - but even these initiatives don't go far enough. As soon as someone lists a set of ‘requirements’ without indicating what makes one system better than another, they've lost the information for comparison.”
Eric Honour
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
Life Cycle
Do we apply SE only during the concept
phase of projects (Level 1) or throughout the
development phase (Level 2)
or throughout production and delivery (Level 3)
or does SE form the basis for all project activities, including operation, modification and
decommissioning (Level 4)?
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
Life Cycle
• Lifecycles are well appreciated – but the ordering of the four levels of maturity is telling
• Processes associated with all stages of the lifecycle are active throughout the lifecycle
• Stakeholders associated with all stages of the lifecycle need to be active throughout the lifecycle as well – Sponsor, User, Procurer, Technical Adviser, Supplier,
Operator, Maintainer, ….
• Exclusion of suppliers during the formative work on a project significantly increases the risk to the project
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
SE Maturity of UK Industry
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
The ‘People Problem’
• Common factor many problems is ‘people’ rather than technical issues – particularly true for Program Systems Engineering – SE is not the responsibility of a single ‘tribe’ of
Systems Engineers within industry– Project management must use SE principles to design
projects – Software engineers (and other engineering disciplines)
must adopt a SE approach, which complements that on the overall project
– Customers must carry out SE activities in cooperation with their suppliers
• So – what are the prospects for Systems Engineering?
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
In Prospect
• Overview
• Barriers to SE excellence
• SE research objectives
• Management of SE research projects/programmes
• Barriers in a SE research programme
• Conclusions
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
In Prospect - Overview
• SE has had many major successes
– but it clear that much remains to be done
• What are its prospects for the coming decade?
• What aspects need to be tackled?
• The ‘SERF’ Report
– Boardman, J, Tully, C & Rose, M; ‘Systems
Engineering: A Research Framework’; Report published
by De Montfort University, 1997
• Update of personal input to the study
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
Barriers to SE excellence
• Lack of understanding of the potential of SE within industry • Ill-considered systems procurements• Poor conceptual understanding of SE
– i.e. regarded only as a engineering specialisation, domain by domain
• Lack of training, both on and off job• Lack of recognised qualifications• Reluctance to apply SE to other than deliverable engineering
projects– i.e. not to project or business processes or to service-providing
systems
• Too close association of SE with software engineering– leading to demarcation issues which give impression that SE only
applies to computer-based systems
• Over-reliance on SECM and (potentially) CMMI models for SE capability improvement
• Over-specialisation in UK educational system (leading to fragmentation of engineering disciplines)
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
Barriers to SE excellence
• Lack of understanding of the potential of SE within industry • Ill-considered systems procurements• Poor conceptual understanding of SE• Lack of training, both on and off job• Lack of recognised qualifications• Reluctance to apply SE to other than deliverable engineering
projects• Too close association of SE with software engineering• Over-reliance on SECM and (potentially) CMMI models for SE
capability improvement• Over-specialisation in UK educational system (leading to
fragmentation of engineering disciplines)
The ‘People Problem’ is very apparent in the above list. For systems engineering to take its proper place in the
future, these barriers need to be removed or steps taken to overcome them.
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
Systems Engineering Research Objectives
• Generic system architectures for complex systems• Taxonomies for systems• Approaches to requirements elicitation and subsequent cost/benefit
analysis to facilitate trade-offs• Generic SE process with per-domain interpretations • Techniques for information-based SE
– e.g. to allow greater abstraction than process-based SE
• Interaction of SE with project management • Application of SE to project design
– i.e. to the system for producing the system
• Methods for system behavioural and performance modelling• SE tool integration & interfacing• Assessment of quality of a system design and appropriate metrics• SE processes tailored to evolving systems and to ‘systems of
systems’ rather than just to unprecedented systems • SE applied to human aspects of systems and to human
organisations in general
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
Systems Engineering Research Objectives
• Generic system architectures for complex systems
• Taxonomies for systems
• Approaches to requirements elicitation and subsequent cost/benefit analysis to facilitate trade-offs
• Generic SE process with per-domain interpretations
• Techniques for information-based SE
• Interaction of SE with project management
• Application of SE to project design
• Methods for system behavioural and performance modelling
• SE tool integration & interfacing
• Assessment of quality of a system design and appropriate metrics
• SE processes tailored to evolving systems and to ‘systems of systems’ rather than just to unprecedented systems
• SE applied to human aspects of systems and to human organisations in general
Work on many topics is on-going; no claims for originality! The need for more work on and a better understanding of ‘soft’ systems engineering is apparent, together with
techniques to facilitate ‘read across’ from one application domain to another.
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
Management of SE research projects/ programmes
• National Systems Engineering Research Agenda supported by Government, Academia and Industry (not tied to specific engineering disciplines or domains)
• Evolution of INCOSE UK Chapter into THE recognised body for Systems Engineering (not tied to specific engineering disciplines or domains
• Network of excellence for systems engineering, provided in association with INCOSE
The DTI’s System Integration Initiative was a major achievement of the Foresight work but it has not been adequately consolidated by the Systems Engineering National Advisory Committee or networks of excellence.
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
Barriers in a systems engineering research programme
• Emphasis on discipline boundaries in academia and engineering institutions
• Lack of formally trained and industrially experienced academic staff
• Unwillingness/inability of industry to release experienced SE staff to help overcome academia's shortfalls
• INCOSE UK Chapter may not be able to evolve beyond group of "enthusiasts“– hence be unable to exercise influence over the national
agenda
These barriers need to be tackled to assure the future of systems engineering.
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
INC
OS
E U
K S
pri
ng
Sy
mp
os
ium
20
01
PGR
Conclusions
“In some ways, the whole paper is a conclusion, or rather an interim conclusion. Perhaps the point that stands out above all others is the extent of the ‘People Problem’; this is far more likely to hold back the advancement of systems engineering than a lack of systems theory.”