Most relevant topics

Thursday, April 30, 2009

Future Trends in Enterprise Architecture

Deepesh Joseph
Date: 08-01-2008

Enterprise Architecture (EA) has been in practice since last 3 decades and it is considered as the core IM tool and key enabler of Enterprise business success through effective alignment of business and IT. EA is an interesting concept and is subjected to constant research and development to cater to varying needs of complex and evolving business models. The objective of this paper is to discuss future trends in EA development by providing some background on role of EA in an enterprise and current trends in EA development. The paper also attempts to present a draft concept about personal enterprise architecture (PEA), which is solely conceptualized by the author of this paper.


1. Introduction

Enterprise Architecture (EA) has been in practice since last 3 decades and it is considered as the key enabler of Enterprise business success through effective alignment of business and IT. EA is an interesting concept and is subjected to constant research and development to cater to varying needs of complex and evolving business models. The objective of this paper is to discuss future trends in EA development by providing some background on role of EA in an enterprise and current trends in EA development. The paper also attempts to present a draft concept about personal enterprise architecture (PEA), which is solely conceptualized by the author of this paper.

2. EA and its role in enabling enterprise building and maintenance

EA is a tool that enables structured setup and maintenance of an enterprise by ensuring constant success through careful alignment of business and IT strategies. Even though the concept of EA was materialized on paper during the late eighties, we should say that EA has long before evolved with its roots in Enterprise engineering and Information systems development. EA existed in some form (even when man first set up mainframe systems), but was not leveraged to actually visualize how structured an EA framework should be. Zachmans' and John Spewak's work on Information Systems architecture and EA planning method, laid down the basic framework and understanding of how EA can be conceptualized, established and used towards enterprise success.

The core idea behind establishing and maintaining EA is to incorporate controls within an enterprise setting in the form of EA framework, methodology and program plan which ensures that enterprise IT strategy is aligned with business strategy and that there is actual documentation of current and future states of the enterprise (in terms of IT and other resources) for taking informed decisions. Even though EA has its existence in the form of intangible framework and methodology, it gets life when exercised through various activities such as setting up of customized EA framework, identification and modeling of business processes, creation of EA artifacts (current and future) at various levels of the framework, identification and development of architecture segments, creation and updation of EA program plan, setting up and constant updation of IT strategy, Information architecture, systems, application and network infrastructure. Throughout this process of EA development and usage, the enterprise is equipped with necessary documentation and management controls to ensure success in vibrant business and technology environment.

3. Current Trends in EA development

There exists various frameworks, concepts and methodologies based on which current EA conceptualization, establishment and maintenance is carried out. Complex business environments leads to constant rethinking on how better can we do EA in order to maximize benefits and minimize costs. There exists no one-size-fits-all EA approach to cater to the various enterprise business models. Even the frameworks that offer wide range of flexibility has to be highly customized to particular business models. This situation has lead to constant development in the field of EA, creating specific frameworks and methodologies for specific needs such as Federal Enterprise Architecture (FEA), DODAF (Department of Defense Architecture Framework), TEAF (Treasury Architecture Framework, Extreme EA framework, Synthesized Architecture framework for Agile and Virtual Enterprises, specific e-Governement EA frameworks, Network based EA, SOA based EA and so on.

Recent developments in EA frameworks and methodology such as TOGAF (The Open Group Architecture Framework) and EA3 frameworks, has been a major step towards setting up common approach that is applicable to various public and private sectors. But, even then, until these generic models are proven successful for all possible business models and as complex business models evolve along with constant innovation in IT, the above trend in EA development will continue.

4. Future trends in EA development

EA is constantly evolving with developments in business models, architecture concepts and new technology. There are several areas of development that foresees future trends in EA, but we are restricting our discussion on major developments such as evolution of complex Virtualized Enterprises, development in Service Oriented Architecture (SOA), developments in Network based view and Information Visualization.

4.1 Evolution of Highly Virtualized Enterprises

Constantly changing business environment and enhancements in technology (eg: Collaborative Networks) drive enterprises to closely link their internal operations to external value chains including suppliers, customers and other entities. This has lead to increasing trend in creation of Virtual enterprises (VE), which is really a collaborative network of value chains. Two major capabilities that help VE being successful is their agility and interoperability. Agility is their ability to modify enterprise models according to vibrant environments. To support interoperability, there should be support for seamless communication between collaborating business entities. It is through effective EA that participating business entities achieve agility and interoperability.

Current EA frameworks and/or methodology is too generic to be utilized for specific types of business processes of participating business entities and does not fully support the configuration of VE. Research is undergone in this area to arrive at most suited frameworks to design and configure agile and interoperable VE. One of the major areas of concern is inter-relating the relevant elements of the value chain and allow seamless collaboration and integration.





Figure 1. Systematic modeling framework for Agile Virtual Enterprises
(Courtesy - Kim T.Y, Lee J.S, Infosys labs, Dept. of Industrial & Management Engineering, POSTECH)


One example of current development of framework is the Systematic Modeling framework (Figure 1) that integrates various aspects such as Zachman's framework, MDA (model driven architecture) based meta-modeling, options for model transformations to support changing views, modeling languages that supports well defined ontology and SOA based architecture to support loosely coupled systems. This systematic modeling framework, as introduced by Tae-Young Kim et. al, focuses on developing a synthesized architecture which provides the framework for modeling the entire VE. The various steps outlined in the systematic modeling framework towards creation and modeling of effective VE are EA design phase, Design of domain specific language (DSL) and domain integration language (DIL), Modeling the component independent models (CIM) and platform independent models (PIM) and deployment phase which involves implementation of platform specific models (PSM).

Systematic modeling framework as briefed above is a major attempt on building robust EA for VE. Research is still going on how we can automate and integrate the various modeling and design phases using standard practices that would the internationally recognized. Another area of concern is testing the framework for large and complex business models to arrive at more concrete results. The drive towards adoption of VE and current research would prompt further EA developments in the future.


4.2 Development in SOA

Service Oriented Architecture (SOA) is one of the system design models which has and will continue to influence the way EA is successfully implemented. SOA is an enabler of EA which aids enterprises to effectively provide their software and other information resources as commercially available, loosely coupled business services. SOA is currently a hot topic that drives EA deployments since enterprises strive to be highly agile and service orientation is being accepted as the best solution for an agile enterprise. SOA ultimately leads to decoupled business services that could be used and reused to satisfy the varied business requirements.



Figure 2. Simple SOA model

A simple SOA model (Figure 2) contains three distinct entities – organization, business processes and services. Ideally, SOA design model is particularly suitable when future EA models are designed and implemented. This is because, the enterprise need to transform from the "as-is" state to the future state with targeted improvements to align IT strategy with business strategy and SOA is applied when the rethinking process starts to plan for the future state. The beauty of SOA driven future EA state is that it still follows the core EA concept of strategy alignments and focuses on how the identified business processes may be designed and coupled using common services.

SOA components are usually registered as web services (independent programs that perform specific business logic/data centric operation) that are invoked and communicates with each other over technology neutral medium such as XML (Extensible Mark up Language), SOAP (Simple Object Access Protocol) and WSDL (Web Services Description Language). SOA conceptualization looks at the "as-is" EA models and tries to model loosely coupled, independent web-services which collectively define the business processes as identified in the refined business strategy. SOA provides ways to implement (register) these services and define methods (service request/response and intra-service communication) by which various services interact to achieve specific or coupled business services.

The main concept that drives the increased usage of SOA in EA implementation is its simplicity and clear ways to model and transform business processes (which is reflection of business strategy) into technology based services (reflects what IT strategy is). This concept redefines the traditional IT implementation model to say that IT requirements are nothing but business service requirements to achieve business goals and objectives. SOA implementation implicitly looks at the service level agreement (SLA) to incorporate the service level performances and thus helps to achieve the ultimate goal of achieving operational and thus business service performance by improving communication across service interfaces. Another factor that drives the usage of SOA is its ability to easily adapt EA to changing business needs (by leveraging the loosely coupled and technology neutral aspects). We may see SOA extending itself, thus taking a major role in EA conceptualization and implementation in the next 5-10 years.

4.3 Developments in Network based view

Network-based analysis techniques have been in use since quite while aiding in representing systems as networks based on the inter-dependency of its individual components. The idea behind utilizing network based techniques is to enable visualization of the system as nodes and dependencies (links). Similar concepts have been recently applied to EA development which has far reaching implications in the way EA would be visualized.

The driving factor behind applying network based techniques in EA development is that dependencies occur during the design, production and use of enterprise components in order to align the task structure according to business strategy. This approach is in line with Zachman's original framework since it bases its core on the insight that the overall EA requires various views of the system and each view has its own architectural components (artifacts) that convey significant design decisions.

The network based EA model attempts to represent the overall EA as a network model with nodes and links where each node represents the Information system component and links represent the dependency between them (Figure 3).





Figure 3. Example Network based EA model (Courtesy-Case study by Iyer B. et. al, from Saha P.)

The represented model is capable to learn and adjust to the changing weights added or reduced, so that it could visualize a flexible, robust and adaptable EA model. This has immense possibilities for opening new avenues for the executives to take informed decisions based on visual models when they anticipate changes in business strategy. The model also serves as a powerful tool to the architect to derive relevant metrics and also as a valuable basis for communication among the architects, development managers, business managers, and senior management. By creating renderings of the information system at different points in time, stakeholders can see how the system evolves over time. Each instance show the abstract view of the entire organization and its information flows. Each node represent the LOB manager’s systems that organization uses everyday and links show manager’s dependencies with everyone else. EA conceptualization and implementation based on network based view is constantly evolving with these striking benefits that the approach offers. Focus on enterprise agility and componentization of business services and the need to visually represent the components and their dependencies in order to predict and analyze changes in EA, would be another driving factor behind adoption of this approach.

4.4 Holographic Information Visualization (IV)

Information Visualization is one major technological breakthrough which allows us to visualize the hidden organization, structure and meaning contained in information. There are two major areas of information visualization according to the way the information attributes are visualized. One approach to information visualization is to represent stored information as a 2D/3D model so that various correlations and patterns can be modeled and interpreted. Another approach is to visualize the actual information attributes as a virtual model in physical space. We are interested in this particular approach of IV and shall see how it will lead to developments in the field of EA.

The most immediate development would be creation of modeling frameworks, meta- models and languages that support visualization (or transform existing artifacts into visual information models) of EA models in visual space. The core concept of business-technology alignment would still remain the basic idea behind establishing and maintaining EA, only difference would be the way frameworks and models will be modeled and visualized. This is most likely to happen in next 10-15 years.


5. Personal Enterprise Architecture – A draft concept

The idea behind creating a draft concept about Personal Enterprise Architecture (PEA) is to introduce the EA concept to common man. PEA is an original work of author of this paper and is targeted at audience who are common individuals or small interest groups who would like to reap benefits of EA within their Information Management (IM) endeavors. The idea evolved as a result of knowledge gained through the EA course work and realization that EA when introduced to common man can create more structured and goal oriented Information Management mediums. This concept is envisioned to conceptualize and become a part of common life in the future, thus relating the concept to future EA trend.

What is PEA?

PEA or Personal Enterprise Architecture is conceptualized as the blueprint of Personal Goal Oriented IM Endeavors (PGOIME). The concept of PEA is most relevant when individuals are empowered by IM solutions and tools which equips them to function as individual virtual enterprises (with personal vision, mission, goals and objectives). The PEA does not deviate from the parent concepts that has been in practice within EA. When EA concentrates on the enterprise as a core business entity and its business-IT alignment, PEA blends into PGOIME of common man. To introduce the validity of the PEA concept, it would be interesting to see a few examples of PGOIME. Following are two examples of PGOIME -


E1. Distance learning student whose goal is to successfully complete the MS program, online

In this case, the vision of the student is to complete course work successfully and in least amount of time. He then sets the goals and objectives to accomplish the vision. The goals and objectives to successfully complete the course work requires him to perform certain activities to complete the distance learning. These activities are then needed to be supported by an IM system through which he realizes the goal of performing the online learning. The IM system established, thus supports the goal oriented personal endeavor, i.e student performing distance learning. PEA, in this case, is relevant in the way the initial vision is mapped to goals & objectives, which are then mapped to the IM system that student utilizes to realize success.

E2. Independent contractor who runs home office

Example of Independent contractor is a more realistic scenario where PEA can be applied. Vision might be to become premium provider of EA consultancy services across the globe. The goals and objectives to achieve the vision requires him to perform certain activities and collaborations to accomplish day to day activities. These activities are then needed to be supported by an IM system through which he realizes the goal of performing successful independent consulting. The IM system established, thus supports the goal oriented personal endeavor, i.e contractor performing independent consulting from his home office. PEA, in this case is relevant in the way the initial vision is mapped to goals & objectives, which are then mapped to the IM system that contractor utilizes to realize success.

PEA Framework

PEA in its simplicity is based on original Zachman’s framework. Basing the above simple IM needs on to Zachman’s framework, it is evident that it can be easily mapped to the various perspective rows (all views share the common individual with differing needs starting from owning the vision to actually implementing the IM solution). The intersection of perspectives and interrogative columns would represent the artifacts that the individual would have to have in place in order to apply and maintain PEA. Conventional means of documentation can be applied to arrive at “as-is” and future states of PEA. The “as-is” state would document how the current PGOIME is conceptualized. The future state would document how PGOIME would be achieved given predicted changes in personal goals and changes in technology. PEA does not propose extensive transformation plan, but suggests documentation of what activities, resources are required to be performed, acquired and integrated to reach the future state.

PEA - Concept Diagram

Figure 4 personifies the PEA concept by relating the PEA development to the above conceptual framework. As shown in the diagram, personal vision is set at the top (owner‘s perspective). Personal vision leads to identification of personal goals and objectives to accomplish the vision (Planner‘s view). Based on personal goals and objectives, demand issues are identified, which are essentially activities required to accomplish goals and objectives. The activities are then mapped to specifications for the required IM solution and IM architecture is conceptualized (Designer‘s view).



Figure 4. PEA concept diagram

Conceptualized IM solution is then implemented as physical setup (Builder‘s view). EA artifacts are created for each view that captures the “as-is” state for the PEA. Future states are also documented by predicting vibrant goals and technology innovation. Drivers of change requires re-thinking on the personal vision and the future state is used as the benchmark to attain the required state.

Advantages of PEA

The main advantage of PEA is that it reinforces human goal oriented nature and acts as a documented aid that describes his preparedness to maintain the IM solution in response to changing personal goals and technology advancements. Representing the “as-is” state serves as the visible model of his thinking process, activities and IT solution. This improves sense of ownership for the IM solution that he maintain to realize his personal goals and the future state documentation & transformation plan aids as an excellent aid in risk management when changes are inevitable.

6.Conclusion

EA is an effective tool for enterprise building and leading it to continuous success. The critical factor behind EA is that it enables business and IT alignment. From the draft concept of PEA, we found that EA can even be applied in day to day IM needs of common man. The current drive towards agile and complex virtual enterprises would be the major area of EA research and there would be major work towards globally accepted EA framework. Information visualization within machine and physical space will redefine the models and language used to represent EA. Many new EA frameworks are likely to evolve and creation of more standardized frameworks that fits all complex enterprise models will take much time and would mostly end up with extended EA frameworks based on SOA.



References

1. Bernard S. A. (2005).An Introduction to Enterprise Architecture (2nd ed.)
2. Saha P. (2007) Handbook of Enterprise Architecture Systems in Practice (Section 1, 3, 5). IGI global.
3. Kim T. Y. , Lee J.S. Enterprise Architecture Framework based on MDA
to Support Virtual Enterprise. Retrieved Nov 16, 2008, from http://www.loa-cnr.it/Guizzardi/VORTE05/VORTE2005.pdf
4. Finkelstein C. (2007, April 1). Introduction to Enterprise Architecture. Retrieved Nov 20, 2008, from http://www.tdan.com/view-articles/4430
5. Yee l., Hsu C. (1997, June 24). A Visualization Architecture for Enterprise Information. Retrieved Nov 20, 2008, from http://www.codata.org/archives/1997/1997Vis/pp6.htm
6. Varhol P. (2007, February 1). The Role of Web Services in Enterprise Architecture. Retrieved Nov 20, 2008, from http://adtmag.com/reports/article.aspx?editorialsid=731
7. Rousselin T., Caumont H. Steps for an Enterprise Architecture in the Geospatial Intelligence Gathering Domain. Retrieved Nov 24, 2008, from http://earth.esa.int/rtd/Events/ESA-EUSC_2008/Presentation/Pr24_Caumont.ppt
8. Harmon P. (2003, January). Business Process Trends. Retrieved Nov 24, 2008, from http://www.bptrends.com/publicationfiles/Enterprise%20Architecture%20Whitepaper-1-23-03.pdf
9. Bass T., Mabry R. (2004). Enterprise Architecture Reference Models:
A Shared Vision for Service-Oriented Architectures. Retrieved Nov 27, 2008, from http://www.enterprise-architecture.info/Images/Defence%20C4ISR/enterprise_architecture_reference_models_v0_8.pdf
10. Radhakrishnan S. (2006, April 6). Enterprise architecture and the role of technology. Retrieved Nov 27, 2008, from http://www.ibm.com/developerworks/library/ar-eatech/index.html?S_TACT=105AGX20&S_CMP=EDU
11. RedHat. EA for State and Local Government. Retrieved Nov 27, 2008, from http://www.redhat.com/pdf/gov/RHinPublicSector.pdf
12. Iyer B. Gottleb R. The Four-Domain Architecture: An approach to support enterprise
architecture design. Retrieved Dec 1, 2008, from http://www.research.ibm.com/journal/sj/433/iyer.pdf
13. IBM, Rational Software. (2004 June). From Inception to Implementation: Delivering Business Value Through Enterprise Architecture. Retrieved Dec 3, 2008, from http://www3.software.ibm.com/ibmdl/pub/software/rational/web/whitepapers/G507-1006-00_EntArch4.pdf
14. Ruest D. , Ruest N. (2006, June 6). Let's talk: Build your enterprise architecture using communication and the right framework. Retrieved Dec 4, 2008, from http://www.ibm.com/developerworks/library/ar-eaframe/index.html?S_TACT=105AGX20&S_CMP=EDU
15. Delegata. (2005, September 9). Service Oriented Architecture: Practical Enterprise Architecture. Retrieved Dec 4, 2008, from http://whitepapers.zdnet.co.uk/0,1000000651,260256626p,00.htm
16. Information Builders. Leveraging Your Data Architecture for Enterprise Business Intelligence. Retrieved Dec 4, 2008, from http://www.informationbuilders.com/products/whitepapers/pdf/DataArch_WP.pdf


(c) Deepesh Joseph, 2001

Related websites -- http://www.getallarticles.com

No comments:

Post a Comment