A Generic Agent Framework to Support the Various Software Project Management Processes

Introduction Software Project Management (SPM) has become a critical task in many organisations. Managing software projects is a complex task, further complicated by a continued increase in the size and complexity of the software-intensive system. In the 1980's SPM methodologies primarily focused on providing schedule and resource data to management (Schwalbe, 2006.) However, present-day SPM activities involve much more. With the advent of the Internet, improvement of computer hardware, software, and networks, global interdisciplinary work teams have changed the working environment addressed by SPM. Global networking capabilities have become more pervasive with the result that cost effective computing resources will continue to play a major role in improving organisational operations. SPM involves the management of all issues involved in the development of a software project, namely scope and objective identification, evaluation, planning, project development approaches, software effort and cost estimation, activity planning, monitoring and control, risk management, resource allocation and control, as well as managing contracts, teams of people and quality. Since publication of the 1995 report of The Standish Group, this same organisation studied 13,522 projects in a follow-up survey, aptly dubbed EXTREME CHAOS (The Standish Group, 2000). This study determined that 23 percent of the surveyed projects failed, 49 percent did not meet the requirements and only 28 percent succeeded. In March 2003 the group reports that success rates increased to a third of all projects, but time overruns increased to the 82nd percentile, whilst only 52 percent of required and specified functions and features were included in the final product (The Standish Group, 2003). Many software projects still failed to comply with the triple constraints of scope, time and cost (Oghma: Open Source, 2003). These triple constraints refer to the fact that the failure of software projects can mostly be attributed to projects not delivered on time and that it does not meet the expectations of the client (scope), and as a result have cost overrun implications. As previously mentioned, the SPM environment is continuously changing due to globalisation and advances in computing technology. This implies that the traditional single project, commonly executed at a single location, has evolved into distributed, collaborative projects. The focus in SPM processes has clearly shifted from the position that it held two decades ago. Consequently, the size, complexity and strategic importance of information systems currently being developed require stringent measures to ensure that software projects do not fail. As organisations continue to invest time and resources in strategically important software projects, managing the risk associated with the project becomes a critical area of concern. Software agent technology offers a promising solution in order to address SPM problems in a distributed environment. According to this technology, software agents are used to support the development of SPM systems in which data, control, expertise, or resources are distributed. Software agent technology provides a natural metaphor for support in a distributed team environment, where software agents can support the project manager and team members to monitor and coordinate tasks, apply quality control measures, validation and verification, as well as change control. Agent technology has distinct advantages over client/server technology as distributed system instantiation. SPM skills, especially in the distributed computing environment, are greatly in demand. Moreover, there is a need for technologies and systems to support management of related aspects of software projects in such environments. Our research is therefore aimed at software practitioners and software developers, but will also be beneficial to researchers working in the field of SPM. …