Assembly-Oriented Product Structures and Assembly Work Instructions for Extensively Long Work Cycle Times. Insights and experiences from assembly of automobiles and heavy truck chassis

This article (or preprint not yet completed due to the retirement of one of the authors) deals with the research area of extensively long work cycle times performed by small parallel workgroups. As was carried out in two of the most socio-technically advanced assembly plants (namely, the Volvo automobile plant in Uddevalla and the so-called assembly docks in the heavy truck chassis plant in Tuve). The author will elucidate (1) how the assembly-oriented product structure was designed for the former assembly plant, (2) how the assembly work instruction was edited and organised within the same assembly plant. Moreover, will also (3) the theoretical frames of references thereby prospected be brought forward. Some interesting explanatory examples fetched from both these two (real-life) assembly systems will also be explicated. The reason for these efforts is that some of the author's colleagues, as well as the author, were heavily involved the births and lives of these two (real-life) assembly systems. Especially was this the case for the automobile plant (which with some time delay feed insights and experiences to the heavy truck chassis plant). Both companies were taking advantages of the author's experimental workshop in Gothenburg, and these two automotive companies were stationing employees to work within this workshop during the period of it's more than eight years life (1985–1993). Since mastering of the product architecture and product variation (as was defined by the product information) proved to be the crucial insight for making extensively long work cycle time possible for full-scale production purposes. That is, the author was in fact enforced to decompose both these (physical) products methodically. However, this approach was in coherence with the author's earlier research and development work regarding such assembly system designs, which were predating the experiences and insights reported here by more than twelve years. This work had previously (i.e. before we got heavily involved) underlined the fact that the content concealed in various information system was deceiving. This anomaly meant that the only way to earlier get correct information regarding components within single physical products proved to be inventories of materials located along some assembly lines (implying that disassembly of products had to be carried out). The work within the experimental workshop, as well as our examination of the present information systems, soon proved that the design-oriented product structure used within Volvo companies was not (for some different reasons) suitable to use as a point of departure for this type of assembly work discussed. The practitioners' misconception during this, as well as during earlier period just mentioned, was that the products in question were so complicated that they were impossible to comprehend by any human being and featuring an immense number of components that in turn were composed of an enormously number of different product variants. A myth that had to be addressed by us, and which in fact proved to emanate from the information systems. Among other things, designing assembly work instructions for long work cycle time work was considered to be almost impossible. Making assembly work instruction for extensively long work cycle time work was usually considered impossible (just the sheer size of such work instructions would make them unreadable for any blue-collar employee). ¬ In conclusion, recognition of the correct product architecture and product variation was, therefore, a necessity. That is, the present (central) design-oriented product structure, used to communicate product information to all Volvo assembly plants on an international level, had to be complemented by an assembly-oriented product structure valid for the (local) shop floor work. In other words, the physical product assembled must indisputably be the point of departure this type of assembly work (i.e. dissections of products were necessary). This insight was attended to by systematically disassembling both several automobiles and one heavy truck chassis in the experimental workshop, while the removed components were spread out on the shop floor. These spread out components were compared with the content in some information systems, which were represented by various paper printouts that were positioned on large elbow-high tables. A large puzzle was thereby at hand that the author's colleagues and the author, in fact, more or less were forced to make sense out of through moving components around on the shop floor and by cutting and pasting in the paper printouts. A process that was e.g. substantiated using e.g. textbooks on automotive matters as well as various other aids from the two automotive companies. This approach was our overarching methodology to deal with the research question already hinted above. This overarching methodology gave birth to two assembly-oriented product structures, this in turn also gave birth to the assembly work instruction.