Presenting A Method for Benchmarking Application in the Enterprise Architecture Planning Process Based on Federal Enterprise Architecture Framework Abbas. Akkasi
Mir ali. seyyedi
Computer Engineering Department Islamic Azad University, Science and Research Branch Tehran, Iran
[email protected]
Computer Engineering Department Islamic Azad University, Tehran-South Branch Tehran, Iran
[email protected]
Fereydon. Shams Computer Engineering Department Shahid Beheshti University Tehran, Iran
[email protected] Abstract— One of the main challenges of the enterprise architecture planning process is its time consuming and to some extend having unrealistic results from this process under heading target architecture products. Getting best practices in this area can be to a large extent effective in speed up and quality enhanced of the results of enterprise architecture planning. Utilization of best practices in most methodologies and the enterprise architecture planning process guidelines namely EAP Methodology presented by Steven Spewak [14] also BSP Methodology produced by IBM [15], have been recommended. However there have been no presentation of any process or a specific method which would lead to benchmarking at enterprise architectural planning level. In this paper, a systematic and documented approach to employ benchmarking in the enterprise architecture planning process is being presented which can be used to assess the equally successful enterprises as best practices in target architecture documentation or by building a transition plan , utilize the enterprise architecture planning process. No doubt in order to have a basic and specific framework and also because of its vast application in governmental and nongovernmental organizations, federal enterprise architect reference models are utilized, though other frameworks and their presented reference models can also be used. Results obtained from proposed approach are indicative of reduced enterprise architecture planning process time especially the target architecture documentation, also risks reduction in this process and increased reliability in production. Keywords- Terminology – benchmarking; Best Practice; Enterprise Architecture Planning Process; Architecture Reference Models; Federal Enterprise Architecture Framework
I. INTRODUCTION Currently, enterprise architecture is considered as one of the important key in enterprise mission [2]. In order to reduce costs and various risks, to enhance effectiveness, profitability and efficiency, the chief officers are required to document a comprehensive information technology programs focused on enterprise architecture. Currently the majority of enterprises particularly large ones are in difficulty in getting enterprise architecture process especially enterprise architecture planning process done. Enterprise architecture planning process is a very slow and tedious procedure and hence leads to misrepresentation of enterprise architecture process and even to its failure [7]. Utilizing benchmarking from the best practice in the enterprise architecture planning process can to a large extent speed up the process. From a different perspective the value of the benchmarking in enterprise architecture process become evident when we know that by using benchmarking we can get to know the structure of other enterprises as progressive or successful organizations or to examine the current benchmark able applications in the above mentioned organizations. Also the techniques and procedures in other organizations based on benchmark able IT are being examined and finally the current projects in other organizations which are related with information technology which are benchmark able are being studied[5]. Various methodologies which are being presented for benchmarking are all centered around process and there is no methodology or approach for benchmarking has ever presented at enterprise architecture level. In this paper effort has been made to present a documented method to use benchmarking in the enterprise architecture planning process so that the current risks in target documented architecture and in the transition plan will be reduced. In the 2nd section a short review on the enterprise architecture planning process as one of
the enterprise architecture process phases is made. In the 3rd section, the federal enterprise architecture framework and its reference models are discussed and in the 4th section the benchmarking process is introduced. The benchmarking method in the enterprise architecture planning process, in the 5th section and in the 6th section, the outstanding challenges in this area is expressed. Finally the conclusion is presented in the 7th section. It must be said that in order to prove the validity of this work, the results of this study have been implemented on several enterprise architecture projects as case studies which in every one of them the outcome agrees with expectations. II.
ENTERPRISE ARCHITECTURE PLANNING PROCESS
Enterprise architecture planning process is part of enterprise architecture process that is shown in Fig. 1. The enterprise architecture planning is a process used to define the essential architecture (business architecture, data architecture, application architecture and technology architecture) and planning for the implementation of the above architecture and its objective is to bring about state where information can be used more effectively to support the enterprise mission [2]. The enterprise architecture planning is a process which in the end results in missions and architecture definition of an organization [3]. As it can be seen in fig. 1 this process consists of four elements of encouraging the organization's top management, to create baseline architecture, document an architecture for the target state and the implementation of transition plans. Of course, in this paper because encouraging the top management is considered to be part of management system, is not discussed here.
is of prime importance as it presents suitable standard for future progress evaluation. On this basis it can be concluded that what changes have taken place in the state [2, 3]. At this stage of enterprise architecture planning process no need is felt for benchmarking as these descriptions and modeling must be based on the present and current facts of the enterprise. B. Develop target enterprise architecture The target architecture must provide a vision of the future operation and its technological support. At this stage of enterprise architecture planning process making use of the experiences of successful organizations or put it differently, benchmarking the successful experiences can help effectively to reliability, agility and being close to reality of the documentation of target architecture. In next sections, after having familiarized with the benchmarking process we have presented a method which would be used to document the target state of architecture. C. Implementing the Transition Plan (develop sequencing plan) The changes that must be carried out in order to transfer the present state to the target one cannot be happened at once. To transfer from the present state to the target one, having a transfer plan is the best. The transfer plan must include a step by step process to transfer from present state to the target one. The particulars of the transfer plan are that while it includes the present state, it also embraces the development plan. The above mentioned plans are either currently being implemented or will be implemented in the future [1, 2]. This plan is mainly implemented on the basis of identified architectural requirements of the target state [2]. In this part of enterprise architecture planning process, it seems that it will be possible to benchmark the successful experiences. Since in the enterprise architecture process the organization is studied in four layers of business, data, application and technology, or in other word, the present state and the target one are implemented in four layers and the transfer plan must be used to transfer from the above four layers architecture in the present state to the four layers architecture in the target state [6]. In benchmarking, these four layers must be born in mind as it will be referred to in the sections to come, for this purpose, the reference model offered by architecture frameworks will be utilized. Because the federal architecture framework is more popular in governmental and nongovernmental organizations , in this paper, this framework is considered to be the reference framework.
Figure 1. Enterprise Architecture Process
III. FEDERAL ENTERPRISE ARCHITECTURE FRAMEWORK(FEAF) AND REFERENCE MODELS
A. develop baseline enterprise architecture The first logical step towards the enterprise architecture planning is to describe the present state of the enterprise. This
In order to have a well defined enterprise architecture it must be somehow possible to bring the complexity of the enterprise under the control and to determine what views or models are required to do the enterprise architecture. In this way it would be possible to bring the enterprise architecture under the control and manage large volume of information hidden in enterprise architecture. This is how John A.
Zachman, the creator of Zackman framework defines the architecture framework [13]:
Benchmarking, Process benchmarking [11].
Benchmarking
“ the enterprise architecture framework is a logical structure used to categorize and organize the various descriptions of an organization which are important for the management and the development of her systems” . Various types of architecture frameworks have been created depending on their application namely Zachman, FEAF, TOGAF, …which in this section we generally introduce federal framework.The goal of FEAF is to ease , develop common processes and common information between federal agencies and other governmental agencies. This framework is suitable for governmental application and profit making and nonprofit making organizations. Federal framework is depicted in fig. 2. As can be seen, various parts of this framework are [6, 16]: architecture drivers, strategic direction, current architecture, target architecture, transitional processes, architectural segments, architecture models and standards.
From the point of view that who is being benchmarked, benchmarking is divided to four kinds: Internal Benchmarking, Competitive Benchmarking, Non Competitive Benchmarking and Best Practice Benchmarking [9, 10, 11]. What has been considered in this paper so far as benchmarking, was the benchmarking of the best practice type, hence it is necessary to fully understand the meaning of the best practice. There are many definitions of best practices; some of them are as follows [9]: •
Something which is effective.
•
Something which works perfect.
•
Something which effectiveness.
repeatedly
and
strategic
demonstrate
Best practices can be identified outside or inside the organization. V.
THE BENCHMARKING METHOD IN ENTERPRISE ARCHITECTURAL PLANNING PROCESS What this paper has been focusing on is the utilization of benchmarking using the best practices in the enterprise architecture planning process. As has been described in sections 2-2 and 3-2, the benchmarking may be used in two sections of target architecture and transition plan from enterprise architecture planning process. In this section the way they were presented in two above sections are being studied.
Figure 2. Federal architecture framework
It can be said that reference models are a list of best practices of other organizations and makes the organizations consistent with their own architecture drivers. In fact the main aim of reference models FEAF, is to standardize the organization's components, to make better use of mission goals and to enhance the effectiveness [16]. Since the enterprise architecture study the organization in layer style, for every layer a reference model in considered. Reference models that are provided by federal framework are: Performance Reference Model (PRM), Business Reference Model (BRM), Service Reference Model (SRM), Data Reference Model (DRM) and Technology Reference Model (TRM) which can be referred to references [6, 7, 16] for further study. IV.
BENCHMARKING
Benchmarking is a continuous evaluation process and the comparison between the home business processes and the corresponding processes in the progressive organizations to obtain information which will help organizations to identify the improvements and put them into action [8]. Benchmarking can be divided into various kinds depending on what or who is to be benchmarked. Based on what is to be benchmarked, the benchmarking process is divided into three types; Functional
A. Benchmarking in target architecture from enterprise architecture planning process As has been mentioned in the previous sections because the target architecture is being developed in four layers (business, data, application and technology), the benchmarking process must observe these four layers in development of the target architecture. In order to achieve this aim, the reference models which have been offered by ّFEAF, must be utilized which has been covered in section 3. The benchmarking model that has been presented is depicted in fig. 3 which every stage will be explained along. Since benchmarking is one of the sustained improvement approaches [8], it must be carried out gradually and continuously and the model may be repetitive however it results in evolution. 1) Identifying Benchmark able Domain In this stage, those sections and domains of the four layer enterprise architecture that have added values for benchmarking are selected. We can use the reference model when we want to search for these domains. In order to choose the benchmark able domains we use taxonomy approach [5, 12].
TABLE II.
NORMALIZED DATA TABLE
STANDARD
A
B
C
CASE BA1
1.21
-0.45
0.17
BA2
0.72
2
1
BA3
-1.21
1.35
1.54
BA4
-0.75
-1.35
-0.51
1.21
1.35
1.54
DO (
DO
j
is the positive ideal value for each standard) j
4) Determination of compound distance between options basis on formula [2]
Figure 3. the Benchmarking Model of the Target Architecture
Since the federal framework reference models are hierarchical we are facing a multilevel priority with various standards. In order to prioritize the leaves of hierarchy of every reference model, we calculate the priority of every node and then to calculate the priority of leaves, take the average of each hierarchical node and its parents to calculate the priority of leaves. As an example, in this section, we demonstrate how taxonomy approach can be used to calculate the first level priority of the business reference model, i.e. business domains for three hypothetical standard A,B,C . In the following section, each of the eight stages approach is being applied: 1) Having unknown options with respect to the aim of the subject in question and determination of the various indexes for selecting the options (supposing we have four options ;BA1, BA2, BA3, BA4 where we want to measure their priorities on the basis of three criteria A,B,C). The value of each option with respect to every criterion is represented by a figure ranging from zero to five. 2) Forming data matrix and calculate the mean and standard deviation.
Dab =
m
∑ (z
− zbj )
2
aj
j ==1
(2)
5) Shortest distance determination (in the above table, the last column represents the shortest distance) 6) Option restriction (option homogenous) At this stage, the upper and lower boundaries of the distance are obtained and the values outside this range are removed from data table and go through the process once again. To determine the upper and lower boundaries, the following formula is also being used: (Upper boundary)
or+ = d r + 2δ d r
(3)
(Lower boundary)
3) Resulted data normalization matrix based on formula [1]
Z IJ =
TABLE I.
X IJ − X I
δI
or− = d r − 2δ d r `
(1)
Which in this example the upper boundary becomes 2.81 and the lower 0.29 . We do not need to remove any value as the distances are within the required range.
DATA TABLE
STANDARD CASE BA1
5
1
3
BA2
4
2
1
BA3
0
3
5
BA4
1
0
2
2.5
1.5
2.75
2.06
1.11
1.46
δJ
(4)
A
B
C
TABLE III.
DISTANCE BETWEEN OPTIONS
BA1
BA2
BA3
BA4
(shortest gap)
0 1.56 2.44 2.23 1.55
1.56 0 3.31 2.4 1.81
2.44 3.31 0 3.72 2.36
2.23 2.4 3.72 0 2.08
1.56 1.56 2.44 2.4 1.55
0.47
0.44
0.41
1.67
0.63
dr BA1 BA2 BA3 BA4
dr δd r
3) The study of the Best Practice and Current Informal Organizational State The domains which have been selected in the first part are being examined in organizations that are chosen as a best practice. Also the informal organizational state must be examined in the same domains. 4) Gap Analysis After having studied the best practices and the informal organizational state, the distance between them must be analyzed. Various approaches of gap analysis mentioned in [4] can be used to carry out the task.
Figure 4. Benchmarking partner selection process
7) Determination of pattern or exemplar. (At this stage the pattern of every option is calculated according to formula [5])
Cio =
2
m
∑ (z
ij
j =1
− DO J )
(5)
8) Sort or categorize the spreading out of options .(at this stage using the formula [6])
Fi =
Cio , CO = cio + 2δCIO CO
(6)
Calculate the priority of each option which in this example is as follows (has shown in table IV): This way the priority of each node is calculated and finally in order of priority, the benchmark able domains will be specified. 2) Identifying Best Practice Candidate and Benchmarking Partner In this section , with regards to criteria and benchmarking domains that we have chosen in the previous stage, a number of progressive organization are chosen as best practices candidate [9]. The process shown in figure 4 illustrates clearly this stage of benchmarking. TABLE IV. Cio
OPTIONS PERIORITY
F
priority i
BA1
2.26
0.59
1
BA2
2.91
0.76
3
BA3 BA4
2.42 3.9
0.63 1.02
2 4
5) Feasibility Study After having studied the gap analysis, it is the time to investigating whether it is possible to transfer the best practices of the progressive organization to enterprise architecture planning process , related to informal organization or not? This stage is of prime importance because of the limitations and constraints that similar organizations encounter, It is not always possible to transfer other's experiences to another organization. This stage must be considered from three view points of technical, operational and economical. 6) Planning to Apply Providing that the results of feasibility study prove positive, the transfer and customization of the good practices of the successful organization to informal organization must be planned [8, 9]. 7) Acceptance and Reviews At this stage, the formal acceptance of the top management in order to make changes will be sought. The project team must not wait till the project is completed rather they have to assess the progress of the project during the benchmarking period [11]. Making use of analysis SWOT during the information preparation for review is extremely useful. B. Benchmarking for Transition Plan Creation Benchmarking the transition plan of benchmarking process kind is considered to be the best practices, because it examines and benchmark the ways and means of reaching the aim in progressive organization. Certainly the work requires to benchmark for the transfer plan is less than the work required for target architecture documentation, since in this case it is not required to examine the other reference models. VI.
CHALLENGES
Despite the presented approach for benchmarking in enterprise architecture planning process, there are challenges and problems which deserve considerations and study. Two instances of these challenges are: •
Assessing the best practices and customizing their experiences for transferring to informal organization is not an easy task.
•
In benchmarking the best practices, normally the current state of the progressive organizations is taken as a desired state for the informal organization.
In any case, although answers has been found for some of the above questions [11], however effort must be made to find techniques to sort out these challenges and other problems.
REFERENCES [1] [2]
VII. CONCLUSION In this paper, the enterprise architecture planning process and its various aspect apart from the top management attraction feature explored and it was pointed out that for documentation of the target architecture and to build the transition plan using the current approaches, long time will be lost and it increases the possibilities of obtaining the unrealistic results. In order to solve this problem a benchmarking approach from the best practices or another word, the progressive organizations similar or dissimilar in business kind, has been introduced in the enterprise architecture planning process which when utilize their experiences, will result in speedy enterprise architecture planning and in the end increases the agility of enterprise architecture process. Also since enterprise architecture planning process consider the organization in four layers(business, data, application and technology), it was shown that benchmarking process from enterprise architecture using referential models which introduces the architecture frameworks , in this phase, will cover all four layers. In the closing stage, the existing challenges and obstacles on the way of benchmarking process in the enterprise architecture planning have been expressed. As a general conclusion it can be pointed to this fact that utilization of the benchmarking of best practice in enterprise architecture planning process, can increase the speed of carrying out the task, also enhances the accuracy and concentration on the key domains and in the end raise the agility of the enterprise architecture process, also will lead to existing risks reduction in the target architecture documentation and creation of transition plan in this process.
[3] [4]
[5] [6] [7] [8] [9] [10] [11] [12] [13] [14]
[15] [16]
"Enterprise information technology architecture faramework:business drivers and architecture principles", US department of education,1998 "Department of energy enterprise architecture , version 1.01",US . Department of energy,2000 "NACSIO Enterprise Architecture". December emmd. Version c,d. National Association of state, chief information officer. Schekkerman, “Trends in Enterprise Architecture: How are Organizations Progressing?”Report of the Third Measurement, http://www.enterprise-architecture.info, J. 2005 Hwagng c.l and k. yoon,"multiple attribiute-decision making",springer verlag,1985 "Chief information officer council (USA), a practical Guide to federal enterprise architecture, version 1.0", February 2001. Executive office of the president of the USA, "FY07budget formulation FEAconsolidated refernce model document", May, 2005. Jim Highsmith , "Agile Project Management: Creating Innovative Products " ,April 06, 2004 Bogan, CE and English," benchmarking for best practices: winning through innovation adaption", McGraw Hill, 1994. Coding, s," best practice benchmarking: the managent guide to successful implementation", industrial Newsletter Ltd, 1992. Mohamed zairi, "Effective mamagemnt of benchmarking projects", planta tree,1998. saaty t.,"the analytic hierarchy process" ,mcgrw – hill ,new york,1980. Zachman, John A., "A Framework for Information Systems Architecture", IBM, Systems Journal, IBM Publication, 1987 Spewak, Steven H. with Steven C. Hill. "Enterprise Architecture Planning, Developing a Blueprint for Data, Applications and Technology", John Wiley & Sons, 1992 “Architecture Alignment and Assessment Guide”, U.S. Federal Architecture, Working Group, October 2000. "The FEA program management office (PMO) published the OMB reference models". Official version of the reference model can be found on the FEAPMO web site :http://www.feapmo.gov