لطفا منتظر باشید ...
19.2 Analysis of the Requirements and Theoretical Foundations
Boyton, Victor, and Pine (1993) build a framework to explain new production paradigms. In this framework they analyze production as varying on two dimensions: product change and process change. In the mass-production setting, a product and its production process are highly stable. In a research department both product and production process are highly dynamic leading to high costs and low volumes. The question then arises as to whether it is possible to reduce the costs and sell high volumes of customized products. The idea of mass-customization, in which the product changes to fit specific demand and the production is organized around ''loosely coupled networks of modular, flexible processing units''(ibid., p. 49), seems to allow such a production scheme. One of the scenarios by Laubacher, Malone, and MIT Scenario Working Group (1997) proposes a similar structure for future organizations. A network of loosely coupled specialists (in most cases one person firms), who come together to produce a highly customized product (of batch size one) and then reconfigure to meet the challenges of the next project. To support such an organization, an IT support system will thus have to enable people to take flexible building blocks of a production process and reassemble them to fit the specific needs of a particular case.Unfortunately, end-users are usually not trained to reconfigure and reassemble existing processes, a job that is usually performed by business analysts. We therefore need to ''unstick'' the process design knowledge (von Hippel 1996) and make it accessible to end-users by encoding it in building blocks and consistency rules of a design environment. The result would be a type of integrated CAD/CAM[1] tool for business processes. This is consistent with von Hippel's (1996) observations in the ASIC's and the computer telephony industry.The component-based approach contains the problem of how to organize the large number of components in order to make them accessible. Experience in AI has shown that it often makes sense to construct some type of taxonomy of components in which similar components can be found close together, leading to the development of frame inheritance networks and object-type hierarchies (Brachman and Schmolze 1985). Furthermore the usage of template (or prototype) hierarchies, a form of simplified frame inheritance networks, has been observed to be useful in settings with end-user development (MacLean et al. 1990). Thus a template-oriented component hierarchy, which can also hold previously completed cases as templates, seems to be advantageous in helping to solve our problem.[1]Computer Aided Design/Computer Aided Manufacturing.
