A New Mixed-Model Assembly Line Planning Approach for an Efficient Variety Steering Integration

In order to deal with the extreme complexity occurring within mass customization production processes today, appropriate variety steering and formation concepts are mandatory. Those are responsible for a customer oriented variant definition which simultaneously reduces internal complexity. In order to achieve mass production at least costs, assembly lines are still attractive means. By avoiding transportation and storage as well as, in particular, by specifically training the employed workers, assembly lines yield substantial reductions of variable unit costs. However, by producing a mass customization variant program with billions of different constellations on the same line, an oscillating capacity use can be observed. Obviously, planning the structure of the used assembly lines and variety steering are strongly interdependent decision problems whose coping is decisive for efficient mass customization. Unfortunately, an integration of both decision levels currently fails because of lacking adequate approaches for mixed-model assembly line balancing. Since known concepts are still based on integral product architectures, they neither correspond to existing steering approaches nor do they cope with the extreme complexity of mass customization processes. Consequently, the present paper sketches a new balancing approach with a modular variant definition. In addition to this, the new model comprises a sophisticated personnel planning. In order to determine systematically appropriate line layouts, a randomized parallel Tabu Search algorithm is generated and analyzed. This approach was designed for the use in an ordinary LAN of personal computers which can be found in almost all companies today. In order to validate its utilizability in companywide networks, results measured for constellations with oscillating background loads are presented.