Abstract :
To be effective, systems engineering requires well-defined processes detailing what activities need to be performed and in what order. It is usual practice to embed these system engineering activities within life-cycle models which frame the big picture for the project and which provide a context within which requirements can be determined and risk managed in a satisfactory manner. However, there are many life-cycle models to choose from; some ´borrowed´ from other organisations, some home-grown, and some taken and adapted from standards-making organisations. Often, once introduced, the life-cycle model tends to become cast in stone and used for all projects within the organisation. This presentation considers whether this ´one size fits all´ approach is necessarily a good thing, since process inflexibility can cause more harm than good, confusing people who have to work within a rigid framework. We argue that greater flexibility surrounding the choice of life-cycle offers a better approach and should therefore be encouraged. We argue that the choice of life-cycle should be predicated on project-specific attributes such as its risk profile and rate of evolution of applicable technology. We go on to discuss a generative approach in which a specific project´s systems engineering life-cycle (and therefore its project structure) is ´composed´ from a few standard life-cycle models. We also examine the hierarchical nature of systems and how the organisational hierarchy needs to mirror the systems hierarchy to prevent undesirable system and project problems.
