کنترل اختلالات خارجی برای مدیریت پروژه نرم افزار

Software project management processes are examined using the Law of Requisite variety to illustrate that a different control approach is required to prevent failures. Causes (disturbances) of project failures in the literature are identified to illustrate how a control system might be designed. A more inclusive use of project personnel and a much more proactive role by buyer executives to provide observers and control solutions is needed. Failure of the London Ambulance Service computerisation is used as an example. Software failures can only be reduced if these disturbances are observed and controlled as early as possible in the project formation.

Software projects still have a low success rate in terms of reliability, meeting due dates and working within assigned budgets [1] and [2]. Factors, which determine successful project management [3], may be related to technical production processes, time scheduling in a dynamic environment and individual differences in project managers, members and team processes. Jones [4] has estimated that such projects only have a success rate of 65%. Houston et al. [5] examined several models of software development and confirmed that one of the most important factors in overall project success was the error rate in code production and its subsequent detection.

Causes of Software Project Failures identified by Smith can be used to illustrate the principle that it is possible to devise a regulatory scheme to eliminate the effects of these causes. The Law of variety tells us that we need to monitor all the disturbances that will cause significant changes and provide at least a different regulation for each of them to prevent their effect on project goals. This requires a stable internal project structure, which in turn will require a good internal model of how the system behaves such as a Systems Dynamics model as in [6]. Many of the sources of disturbances identified are internal to the vendor organisation and can be dealt with by the normal project management structures, provided they are effective. Major causes of disaster such as the buyer changing their requirements or their business model changing due to take-over need to be detected well in advance of the effects being felt by the project team if a remedy is to be found. This means an observer in the project team being tasked to look out for signs of these events. Action will be required at the highest level in the vendor to cope with such events. It is suggested here that the team adopt a role like an organism to detect threats to its’ achievement of the required project goals. The whole team must be the source of ideas to provide the necessary variety of regulation. A major difficulty in implementation is the lack of suitable measures of successful projects. The example of the London Ambulance Service is examined and shown to be a difficult system to deal with but can be adequately solved if radical negotiation had been pursued at the start. A data bank of information in each vendor on successful project parameters is needed as well as much better training for executives in purchasing organisations.