3 Learning by Machine: A Multiple Perspective Approach to Knowledge Work Analysis Jan Aidemark * Johan Persson ** Växjö University, School of Management and Economics, SE-351 95 Växjö, Sweden, *Email: Jan.aidemark@ehv.vxu.se **Email: Johanp@hotmail.com Abstract Knowledge management is a diverse, complex and often contradictory activity ( Newell et al. 2001). This paper investigates a study of operators at a printing company. The operators are characterised as knowledge workers and the possibilities to introduce information based knowledge support for these workers were investigated. To deal with the complexity, a five perspective interpretative framework is suggested. A complex picture emerged; different intertwined and interacting knowledge management themes were found. The outcome of the study is a discussion of the necessity to discuss different and very diverse aspects of knowledge management in an integrated way. Keywords Knowledge management systems, Planning framework, Informal learning, Interpretative case study. 1. INTRODUCTION Knowledge management is a set of procedures and techniques used to get the most from a firm’s knowledge asset (Teece 2000). In practice, it is a diverse, complex and often contradictory activity (Newell et al. 2001). Anything from database technologies and intranet solutions to organisational learning and intellectual capital has been discussed in terms of knowledge management projects. Knowledge management has been described as a necessity for survival (Grant 1996) and as the next management fad (Swan et al. 1999). This complexity, it has been suggested (Davenport and Prusak 1998), could be tackled by selecting one aspect of knowledge management (e.g. organisational learning or intellectual capital etc.) that is known and working in the organisation to build on this aspect. This would facilitate getting things going, setting good examples and educating the organisation about the benefits of knowledge management. However, a knowledge management project that returns benefits for the whole organisation consists of a number of activities that must be interrelated (Kreiner and Mouritsen 2003 p. 237-8). Kreiner and Mouritsen argued that, even if a certain knowledge management effort could lead the project, it must be preceded by planning on a strategic level. This process should produce a picture of the overall aims and an overview of the multitude of the necessary knowledge management sub-projects. To investigate how such a picture could be formulated, an analysis of a case is presented and discussed. The aim of this study is to understand the operators of the printing machines as knowledge workers. We look at the overall possibilities of introducing knowledge management support, trying to pinpoint the difficulties and issues that must be dealt with before a projected is started. The aim of this research is to discuss the problems and possibilities of introducing a knowledge support system in a traditional blue-collar workplace. This is about giving a multi-perspective account of knowledge related aspects of the daily life of operators and to discuss knowledge management solutions and their possible impact on the situation. This has been a favoured theme of information systems research (c.f. Zuboff 1988). The focus has been on moving people’s knowledge to machines. In knowledge management the machine is supposed to aid the building of personal knowledge, either by focussing on personal creativity or by knowledge sharing. A number of studies of the possibilities and problems of supporting operators have been reported, for example (Glassey et al. 2000; Ohashi and Yuki 2002). Both of these studies suggest knowledge bases and focus upon the elicitation and encoding process. A series of special pre-conditions are expressed. For example, Glassey et al. (2000) report from studying bio-industrial processes that variances in processes are solved by the expertise of operators and how they interpret the problem. Further, Ohashi and Yuki (2002) in a study of manufacturing processes, point to problems of complexity and lack of completeness of standard operating procedures. Here the