Funded by the: JISC Flexible Service Delivery programme.
Lead Institution: King's College London.
Key Words: Enterprise Architecture (EA).
Our Enterprise Architecture (EA) journey
a story of impact and value
Background and institutional context
Institution: King's College London
King’s College London has grown as an amalgamation of previously autonomous institutions. The College is spread over many separate sites across central London. There is a history of departments managing their own IT systems. As a consequence, the legacy information infrastructure is fragmented, and can be difficult to maintain and frustrating for users. Factors such as space, the listed nature of many buildings, the reliability of power supplies, as well as difficulties of staff recruitment in the central London location add to the integration challenges.
A key priority, incorporated into the College’s Strategic Plan, is the development of an e-environment enabling greater connectivity and providing integrated access to services for supporting research, teaching and administration. The framework for implementing this strategy is provided by the College's Connected Campus programme (2007-2010) and its successor, Collaborative Campus that implement IT provision to support the College’s 10-year Strategic Plan 2006-2016.
The motivation for the work on Enterprise Architecture (EA) is to develop an architectural vision, which can be shared across different teams and business areas to improve both strategic alignment and operational efficiency.
By developing a shared architecture, our aim is to enable interoperability between systems, sharing of resources and enable 'joined up' services for the user. In our experience, shared understanding of the goals and objectives results in solutions that make better use of the available resources and provide a higher quality end user experience. Work on EA is being carried out in parallel to other initiatives such as improvement of software development processes that aim to improve agility and efficiency. At senior levels in the institution there is already recognition of the need for an architecture to describe our current assets, in order to assist with planning and strategic decision making. We are promoting EA tools and methods as a way of addressing this need.
Objective/Overall vision for EA practice
To date
|
Looking ahead
(next 12 months post June 2011)
|
EA provides a mechanism to develop a shared architectural vision that is aligned with institutional or departmental IT strategies, as well as providing a framework for planning and execution of the projects that implement the strategy.
By aligning our projects with a common architectural vision, we aim to improve efficiency, enable sharing and reuse and reduce duplication.
We have successfully applied frameworks such as PRINCE2 in the College by reducing complex and generic principles to manageable processes and guidelines. We have been taking a similar approach with EA.
It is important in our experience to provide existing staff with a set of practical EA techniques rather than creating complex and unwieldy bureaucracy.
|
We have expended considerable effort in understanding and applying EA techniques in selected areas, and training staff. The emphasis for the next 12 months will be explaining and encouraging adoption of the practical EA methods and tools.
Progress in EA is highly dependent on having champions in the institution who are able to drive such initiatives forward. In particular, short term deadlines often mean that longer term strategic goals often have lower priority.
Recent restructuring of IT and library services provide both challenges and opportunities for initiatives such as EA.
|
Getting value from JISC resources
To date
|
Looking ahead
(next 12 months post June 2011)
|
The FSD programme has enabled us to explore EA frameworks such as The Open Group Architecture Framework (TOGAF), share experiences with other institutions implementing EA who are at different levels of maturity, as well as learn specific EA tools and techniques such as ArchiMate modelling and Service Oriented Architecture.
|
The emphasis in the next 12 months will be on increasing knowledge of EA tools and techniques, widening our use of EA where there are business benefits to be realised, as well as sharing experiences and learning from other institutions.
Practical examples of successful adoption of EA as well as sharing of successes and failures are useful as learning experiences.
|
Progress and achievements
We have completed the College’s Enterprise Architecture Project 2 that modelled shared services for research and research administration. This has enabled us to better align the development of research administration and repository systems around common strategic goals and Enterprise architectural components.
King's College London have adopted the ArchiMate modelling language and Archi modelling tool across a large number of projects. This has facilitated sharing and reuse of architectural and software artefacts, which has reduced development times by enabling reuse.
By running an EA workshop at King’s College London in collaboration with JISC, we were able to train five staff across different departments in EA principles and modelling techniques. This has helped to increase the number of IT staff who are 'EA literate'.
Business benefits (impact) and opportunities (3xE's)
Efficiency: There are potential benefits to be gained in improving efficiency and lowering costs. The cost savings relate particularly to identifying opportunities for aligning projects across different teams, defining common business processes so we can deploy one system instead of several, identifying common application components, saving on procurement or development costs, and sharing infrastructure. We have applied the EA approach to research management and repository systems. This has enabled us to identify and define common business processes, which in turn have led to simplification of our systems and reduced development effort.
Effectiveness: Users appreciate the benefits of simplification, resulting from aligning business processes across different systems, and the breaking down of silos so that several systems can interoperate resulting in an integrated user experience. In developments such as the JISC funded CLIF project, we have integrated SharePoint, Sakai and Fedora repository systems around a common architecture to provide services to users that enable integrated workflows across highly heterogeneous systems. These realise practical usage scenarios such as preparation of exam materials, research collaboration and development of administrative policies.
Enablement: The architectural models we have developed so far enable far more rapid communication and development of a shared understanding. This has increased our agility in being able to respond to new requirements.
Key learning points: do's and don'ts, and advice to those who follow
In order to achieve adoption of EA at a senior level, it is necessary to articulate clearly the purpose of using EA techniques and the potential benefits. In order to provide more concrete evidence of the benefits, a 'bottom up' approach is useful. Providing architectural alignment of a specific business area can act as a proof-of-concept as well as a catalyst for adoption.
A strong perception exists that EA can become an 'ivory tower' where architects work in isolation and become divorced from the operational staff. The principle of 'just enough EA' should be adopted to keep the effort required to implement EA in proportion with the size of the IT department and the realisable benefits.
Language is a barrier to implementing EA. EA requires a basic common understanding of principles and terminology in order to enable cross-organisational communication about architecture.