Model-Based Self-Managing Systems Engineering

For many years, the vision of smart computing where systems can function and/or manage themselves independently from human intervention has provided numerous theoretical challenges to research communities ranging from intelligent systems and cybernetics to AI communities. These research trends have now been further fuelled by the IBM autonomic computing initiative, where biologically inspired concepts inform the development of systems that can adapt autonomously to their users' requirements and environments. This paper considers the extent to which well-established general systems concepts might be valuable in the design of autonomic systems. The main two approaches considered are Checkland's Soft Systems Methodology (SSM) and Beer's Viable Systems Model (VSM). The paper summarizes the relevant aspects of each approach and demonstrates their potential through the provision of an illustrative case study. Moreover, the paper illustrates how SSM and VSM approaches facilitate autonomic systems engineering by the capture of functional and non-functional application requirements such as lifetime self-management policies and operational tolerances