Complexity Framework of Adaptive Scalable Manufacturing Systems

Abstract

Manufacturing industries of today are under tremendous stress of providing high-quality products having suitable variety with relatively low cost compared to their competitors. The associated challenges with this include high initial cost, larger lead time and lack of throughput. These obstacles have forced companies to develop complex systems with deep adaptability. These systems also require an ability to adapt to changing demand in throughput. This ability more commonly known as scalability. It allows the manufacturers to scale up or scale down the manufacturing setup as per demand. Such systems along with the intrinsic parts that are being manufactured in turn come with their associated complexity. This ‘complexity’ is the result of multiple products being developed at the same time with varying development requirements. It has been seen in the past that severe issues can arise in manufacturing setups and their products if complexity is left unaddressed. More notably, in Mercedes E-series, electrical and starting issues were encountered due to more complex product development systems. To address these challenges, we need to first identify the level of complexity. This can be accomplished by considering certain attributes that affect the part and overall complexity. Our model was developed keeping this particular concept in mind. Existing models for part complexity have certain limitations. Some of them consider aspects that make the model too complicated. Others have too few and thus do not model complexity completely. Therefore, using these existing models, a complexity modeling system was developed. This aided us in not only categorizing products but also provide a clear road map for the developers in setting up production lines. Once this is accomplished, based on that, the model was extended to develop means to separate products that are more or less complex. A quantitative analysis was also performed on the model. Existing parts with varying complexities were used as a basis and prominent existing complexity models were applied on those parts. It was found that these models did not model the complexity of these parts correctly. Some produced complexity values of more simple parts greater than the more complex ones. Others gave equal values to a range of different parts. Our model produced a satisfactorily increasing trend as per the part complexity. The adaptive scalable manufacturing provides solution by reducing the lead time and provide economic development of varying products. Other models such as dedicated manufacturing systems do have an extremely low throughput time but are unable to cope with XIV the changing market demands. Systems that are flexible or reconfigurable (collectively called adaptive) do have a drawback of a higher initial cost but make up for it at higher production rates with a suitable level of product variety. These do not remain cost efficient below a certain production rate level due to the high initial investment. This thesis encompasses four basic models: A model for part complexity, assimilating part complexity with system scalability of the proposed framework, a modeling system for the said framework and product family formation model. To show the working of these models, several case studies are presented throughout the dissertation. The model is validated through a comparison with the existing models for complexity. It is shown that existing models do not show an increasing trend for certain parts with increasing complexity. Linkage between parts complexity and the effects on system scalability requirements is missing from literature. This has also been added in this dissertation. Significant contributions of this work include complexity model formulation having improved differentiability between similar parts, formation of product families using the complexity model along-with its implementation on a scalable system. Suggested models are also applicable separately if required.