Problem Frames for Socio-technical Systems: predictability and change John Brier, Lucia Rapanotti, Jon G. Hall Department of Computing, The Open University Walton Hall, Milton Keynes, MK7 6AA, UK {J.Brier, L.Rapanotti, J.G.Hall}@open.ac.uk Abstract Problem Frames are a widely accepted framework for analysing and cataloguing software problems. Their foundation is the widely accepted separation between the machine, the solution which is the object of design, and its context, the environment where the problem exists. Recently, Problem Frames have been shown to be amenable to the analysis of complex socio- technical problems, in which social components, once only part of the problem description, are now the subject of design. To this end, a further separation is required to identify those social elements. In this position paper we proposed a program of research to resolve issues of predictability and change that arise when social and technological components are objects of co-design. Our work is inspired by the Change State Model of Stacey. We argue how this model may apply to Problem Frames in the context of design of complex socio- technical systems, and discuss the implications in terms of both Problem Frames foundation and notation. Key Words: Problem Frames, socio-technical system, Change State Model, problem analysis, predictable/unpredictable, control/flexibility. 1. Introduction Problem Frames were introduced by Jackson [5,6] as a framework for analysing and cataloguing software problems. In their original formulation, they focused on the specification of software solutions. Recently, Hall and Rapanotti [4] have shown how they can be extended to deal with complex socio-technical systems, in which the behaviour of social components of the system are specified together with that of the machine. Such separation raises a number of issues. Among them is characterising the different nature of humans and machines in terms of their inherent flexibility and the control that can be exercised on their behaviour as a result of the design process. From a temporal perspective, control and flexibility also relate to the ability of the socio - technical system as a whole to evolve throughout the development process and in response to changes in the environment. In this paper, we start addressing issues of predictability and change, by considering how the Change State Model (CSM) of Stacey [8] may apply to Problem Frames in the context of co-design of socio-technical components. CSM [8] is a Knowledge Management approach to codifying and categorising the change conditions of a context. Our hypothesis is that, within Problem Frames, it may allow for the identification of the relative degrees of control and flexibility in the interface between environment and machine, and environment and human; and the representation and analysis of the nature of change in a socio-technical system both within the inception to realisation process, and in post-deployment evolution. The paper is structured as follows: Section 2 briefly reviews the background to the paper; Section 3 addresses change in problem frames; Section 4 proposes future work. 2. Background Problem Frames are a concretisation of the ideas of Michael Jackson and others in the separation of machine and its environment’s descriptions [3,10]. This separation is generally accepted as being a useful principle for requirements analysis. The usual representation of the separation of machine and environment descriptions is as the ‘two ellipse’ model, illustrated in Figure 1. In that figure world knowledge W is a description of the relevant environment; R is the statement of requirements; S is the specification that mediates between environment and machine; M is the description of the machine; and P is the program that, on machine M, implements the specification S. The concretisation of this separation in Problem Frames notation is given in Figure 2.