Naturally Grouped Assembly Work and New Product Structures

Our latest research has shown that in order to be able to develop and maintain different final assembly systems and to further develop industrial work, the introduction of new types of product structures is a necessity. At the same time these new structures allow resistant relationships to be found as a basis for orientation, to exploit to a greater extent than previously the precise work of the design department in central pre–production and the various final assembly processes. This is possible in the motor vehicle industry since automobiles and trucks display certain generic characteristics. The consequence is that the information systems in use today and the company's transfer surfaces within the internal organization must be changed, which presupposes new types of product structures. This is especially important when the new long cycle assembly work is realized. This paper summarizes some experience and methods from the development of long cycle time assembly systems, and reports general questions valid for the person or the company intending to humanize and make assembly work more efficient, and illustrates these questions with specific 'delimited' examples. Here we refer principally to the development of ways of describing products. Our results are based on several years' collaboration, characterized as action research, between researchers and practitioners mainly within the Volvo Car and Volvo Truck Corporations. The work has led to real implementation of research results in industrial environments. When designing alternatives to traditional line–assembly, we advocate a generally applicable product structure as a basis, which is then adapted to different local final assembly plants. When this approach is used it has proved to be important to define principles of differentiation between causal and taxonomic descriptions and the use of stereometric models of products and local assembly processes in relation to central design and pre–production departments is essential. In this paper, we report a number of practical examples to illustrate the empiric support for the reasoning applied, and these in turn have necessitated certain modifications to the original basic stereometric model that we have published previously. These modifications are also based on practical experience which is not explicitly reported in this paper.