Abstract
A company started loosing its market share due to a prolonged period of the system development life cycle (SDLC). We ensure that the time-to-market is reduced in order to increase a company’s competitiveness as we also work on minimizing the requirement engineering (RE) phase of the SDLC while holding the current quality levels of product high. In solving this problem, action research was utilized to a large extent as the preferred method of inquiry. Action research is divided into five types; dialogical, participatory, canonical and dual imperative. In this study, we combined the joint application design (JAD), group decision support system (GDSS), computer aided software engineering (CASE), and unified modeling language (UML) (Argyris, 1985). At the close of this case, productivity had been successfully increased by 250% whereas the SDLC had been reduced from 14 days to 4 days.
Many a times, information systems (IS) are late due to the short time required for its implementation or the problems that may arise during the implementation life cycle. During systems development, requirement engineering has been identified to be the most challenging step. Therefore this is the critical phase which accounts for the successful reduction of the product-to-market time through decreasing requirement engineering (Baskerville & Pries-Heje, 1999). For as long as a decade ago, engineering requirement has been a challenge. Estimations have indicated that the reason for project failure is requirements which accounts for almost 40-85% failures recorded. It is therefore important to know how to develop an information system that is successful by spending limited time on RE. This shall be the research question for this project.
Action research, is a brain child idea of two research institutions which developed it independently for the study of social psychology, these were Kurt Lewin form University of Michigan and researchers at Tavistock Institute. Evils of social illness were studied through this approach which was the basis for treating these illnesses. There are six steps related to action research which include analyzing, searching for facts, conceptualization, coming up with plans, action for curing implementation and finally the evaluation of the solution thus arrived at (Baskerville & Pries-Heje, 1999). Over the years, modifications have been done to these steps which may increase the number of steps.
`Various disciplines use action research in solving different problems that they face from social sciences through to IS. The ISD-information system development is a practical approach that can be related to real world challenges and use to solve problems (Argyris, 1985). Action research was the most relevant approach for this project given its high rate of success in solving practical problems. Action research has various forms such as collaborative practice research, conical action research, participatory action research, dual imperative action research and dialogical action research. For this research, the dual imperative action research was preferred. This came in handy in the bid to dispel concerns on whether there was an obligation or not to conduct the research.
In analyzing requirement engineering, IS projects are known to deliver a 42-67% of their inventive requirements. A further 48% of the developments of IT relates to requirement engineering. Success of the SDLC is directly connected to the of requirement engineering. The difficulty of RE is the unease in the determining what the system is required to do.
The formulation of the research and solving process followed the six steps stated in the dual imperative action research. Late projects at Evans Data Corporation were at an alarming rate of 45% and we had to consider what could be done in order to reduce this percentage significantly. This was due to the long time that the SDLC took in carrying out the projects. Methods employed in the company included the Agile, Extreme Programming and prototyping which spent very limited time in the RE thereby resulting in many errors and poor quality products. Moreover, these methods are not viable for large project and therefore it was not preferred in this conducting project. In lieu of this, the most appropriate way of solving this problem was through the use of JAD.
The research was conducted at the Computer Aided Design business plant. This section is liable for developing production plans, designing of motherboards for computers and RE. It’s organized into management teams that are given various tasks to perform. This forms the primary connection between design and actual manufacturing. The production teams consisted of highly experienced members who had been in the business for close to twenty five years.
The activities of teams in this unit were to be analyzed more in order to to identify the problems as agreed by the management. Prior to the study, there were conflicting views which required clarification of the unit operations. The team was qualified although the members were not conversant with its broader activities. The workshop was run by JAD facilitators. Business activities were identified through a group support tool. A three steps approach was used as a model: pre-workshop, workshop and the post workshop. The prevailing workshop conditions were studied in this step. The stakeholders in each production unit were identified and the workshop expectations reviewed. The pre-workshop had five steps that were followed; process owner were interviewed to find out their need and objectives, Re models were selected and applied appropriately, administrative tasks for example signing of contracts, educating participants and finally workshop agenda were formulated to perform the case study. JAD were used for gathering information.
This project was a success as evident by the reduction of RE phase from 14 days to roughly four days in the product time-to-market. The accomplishment of this did not compromise on the quality as stated in the research objectives requirement. A productivity increase of 250% was achieved through the enhancement of coordination, merging of some department and improvement of the team communication models.
Through interviews, discussions, analyzing problems and systems more compelling facts were unearthed. These were incorporated in the model for making the necessary changes. After implementation of the suggested changes, data was collected to indentify the new performance of the product time-to-market. For the determination of the levels of satisfaction, customers and managers were interviewed and the results were positive. Six issue we discovered to be derailing the production process which included a lack of common understanding, efforts duplication, lack of team and individual coordination, inconsistencies in product description, poor communication among teams and finally high unacceptable re-work levels.
The participants faced challenges in making distinctions between the physical and logical views in the business. Previously they had never viewed the business in a logical way therefore it was easier to represent their work in a techno-specific way.
It will be interesting in future to test the other approaches of action research in undertaking RE. Although the dual imperative approach was successful in doing this research, it gives an open end on whether the choice of other approaches would have yielded similar results. Shortening Requirements Engineering is important and applicable to almost all engineering firms in trouble-shooting of the possible causes of delay. It is important to note that there are many models that can be specific to particular industries study. Research models that encompass all the parameters affecting a particular situation should be keenly studied and incorporated in the research. In this way, time wasted between the start of production to the point of product entering the market as a finished product is reduced. Quality is also maintained at high levels through upholding recommendations from research.
References
Argyris, C. (1985). Action Science: Concept, Methods and Skills for Research and Intervention. San Francisco: Jossey-Bass publishers.
Baskerville, R. & Pries-Heje, J. (1999). Grounded Action Research: A Method for Understanding IT in Practice, Accounting Management and Information Technology, Vol.9, 1-23.
Hooks, I & Farry, k. (2001). Customer Centered Products: Creating Successful Products through Smart Requirements Management. New York: AMACOM.
Schrodl, H & Wind, S. (2001). Requirements Engineering for Cloud Computing. Journal of Communications and Computers, Vol. 8, 707-715.
