Key Aspects Of Rup And How To Use It

A Brief Overview of RUPmoderate Probability of Requirements/Design Change
RUP is the abbreviation for "Rational Unified Process"expected - moderate
- a systems development methodology devised by3. Construction - in traditional terms, this is where the
Rational Unified Corporation and now owned by IBM.bulk of the system is built. The exit gateway is
The author has no connection with any of theseknown as the Initial Operational Capability Milestone.
organisations, but has used the process framework inIn other words, at the end of this phase it is now 'on
major development projects.the runway'.
The process was developed as a result of workDegree of Requirements/Design Change expected -
done by Booch, Jacobson and Rumbaughlow Probability of Requirements/Design Change
(affectionately known as the Three Amigos), whoexpected - moderate
were in large part the designers of UML ("Universal4. Transition - the system moves into production, it is
Modelling Language"). The RUP concept was basedavailable in beta form to the users and training is
on an analysis of what really happens in developmentunderway. It is reviewed against the quality criteria
projects and why many projects fail (typically theestablished during the Inception Phase. The exit
'failure rate' is 30%).gateway is called the Product Release Milestone.
It fits into the 'Agile' project management spectrumWhy Should a Project Use RUP?
at the top end, after XP, Scrum and DSDM on aThe main reasons are:
scale of complexity and team size.1. The project risk profile is front-loaded.
By designing the project process in a way which2.The technically risky aspects of the system are
ensures that the riskiest items are addressedaddressed and proven early in the project. If not, the
according to highest risk first, then the overall projectproject is redesigned or cancelled before significant
risk (as initially perceived) is not reduced at first, butresources are committed.
the risk of a large project gaining momentum andEach formal phase has its own inbuilt inception,
then collapsing spectacularly at a late stage iselaboration, construction and transition phases. After
aggressively addressed. This means that the risk ofall, the project manager has to 'deliver the
failure should clearly taper off as a projectmilestones'. This gives the project a 'fractal' structure,
progresses - this is quite different to what happens ineven down to lowest level programming tasks.
many 'traditional' projects. Of course, it will not3. Testing is a pervasive of the framework process.
protect against the risk of a business paradigm shiftTesting effort is required throughout, with proofs
that has not been foreseen.confirmations/reviews a constant theme, so that
In summary, RUP identifies nine project disciplines:sixproblems and issues are flushed out as early as
'engineering disciplines': Business Modeling,possible, both from non software products and other
Requirements (capture, management), Analysis andproject products such as Use Cases, budgets, plans
Design, Implementation, Testing, Deploymentandand infrastructure behaviours.
three 'supporting disciplines': Configuration and Change4. Because it is iterative, a project can be designed
Management, Project Management, Environmentso that a useful prototype may be delivered early
Management.(unlike a waterfall approach), with the attendant
These are supported by software toolsets (forbenefit of flushing out issues for addressing in later
example UML modelling tools, automated testing anditerations.
test management tools and so on). Iterative working5. It is ideal for larger projects which have significant
is an essential component of the process structure,technical risks or are 'ground-breaking' in other ways.
with artefacts (in Prince terms these would be called6. The iterative buildout and incremental delivery
products) being continually refined and retestedapproach lends itself to situations where a business
(remember that even a test plan should be checkedarea is undergoing rapid change, so that there is early
against its defined standards, as it is itself a projectdelivery of value, and the shape of the system can
artefact).be coupled to the rate of change; this however will
A project is defined in terms of four phases:also require a suitably flexible underlying systems
1. Inception - this is the high level design of thearchitecture.
project itself, including governance, business case,How Should it be Used?
budgets, risks, plans and, often, assessment of anIt requires a suitably experienced project manager to
architectural prototype. The exit gateway is calledmake it work effectively. It can be applied with a
the Lifecycle Objective Milestone - that is, what thelight touch or a heavy touch - the experienced
project is seeking to achieve over the completeproject manager will be able to apply a 'contingent'
lifecycle (including realisation of the benefits).approach and adjust the intensity of the process
Degree of Requirements/Design Changes expected -implementation according to the risk profile, team
high Probability of Requirements/Design Changeexperience and so on. Indeed, the ability to
expected - highcustomise the process is key attribute of a suitable
2. Elaboration - this phase involves extension of theproject manager, as the process lends itself very well
teams, design products and prototype buildout,to customisation.
enhancement of project processes (eg testing)The Technical Architect, lead analysts, lead designers
infrastructure, and so on, with a formal exit gatewayand test team need to be well versed in the iterative
known as the Lifecycle Architecture Milestone, theapproach.
passing of which confirms that an executableTo be used properly, it will require investment in
architecture has been demonstrated which 'realises', -software toolsets and infrastructure. Given this, the
that is physically delivers - the architecturally riskyminimum project size at which the process
Use Cases and how their associated risks areframework becomes practical is probably in the
mitigated. It also requires that 80% of Use Casesregion of 4,000 - 5,000 man days, unless of course
have been identified and designed, prioritisedan organisation is large enough to share the overhead
according to risk, together with rework of theacross other projects.
Business Case and Risks. There are other importantThe iterative approach is not always easy to grasp
'tangibles' required at this time too, including theby people who are used to a waterfall structure. It is
software architecture model and the developmentimportant that the process is well-understood at the
plan. At this point the project will move into a phasehighest levels of project governance. However, early
where the risk profile is raised (relatively) as changesbenefits delivery is nearly always of interest at a
will be more difficult to accommodate.governance level!
Degree of Requirements/Design Change expected -