Switchover to additive manufacturing: Dynamic decision-making for accurate, personalized and smart end-use parts

Additive manufacturing (AM) is rapidly developing into a general-purpose technology akin to electric drives and computers serving a plethora of applications. The advent and proliferation of the additive process triggering Industry 4.0 is encouraging academics and practitioners to establish new practices, designs, and modes of creating and supplying end-use parts. Contributing to this emerging stream of research on AM technologies, the overarching objective of this doctoral dissertation is to discover situations and ways in which companies can benefit from implementing AM in conjunction with conventional manufacturing technologies. This is addressed and limited by three sub-objectives. First sub-objective establishes a new operational practice—dynamic supplier selection using the build-to-model mode of manufacturing—for the provision of idiosyncratic spare parts to improve the after-sales operations of a case company. Second sub-objective estimates the combined uncertainty and the worst-case error in creating an end-use part, particularly a personalized implant made by radiologic images, thresholding, digital design, and AM. Third sub-objective develops process interruption-based embedding and creates prototypes of smart parts, in particular intelligent implants using four AM technologies. The work uses a multi-methods approach combining three case studies, experiments, and research methodologies to achieve the aim of theoretical insights, practical relevance, and innovation. The empirical evidence confirms that AM can radically shift the performance frontier for problematic parts in conventional supply. The dynamic supplier selection practice allows operations managers to choose a supplier or multiple suppliers for idiosyncratic parts both existing and new. The selection can be based on cost reduction, lead-time reduction, and trade-offs in cost and lead-time according to customer requirements without significant transaction costs. The generative mechanism of successful outcome is triggered by the simplicity in AM process instructions. Encapsulating the design and production-process instructions reduces mundane transaction costs and enables highly interactive model-based supplier relationships for decentralized manufacturing. The accuracy of AM technologies is predominant for establishing and substantiating appropriate practices. The process interruption-based embedding opens a direction for creating smart parts facilitating condition monitoring, machine learning, and preventive maintenance for Industry 4.0. This doctoral dissertation aids researchers and practitioners in switching parts over to AM technologies from large spare part repositories with a dynamic response as opposed to a static choice with conventional manufacturing involving increasing minimum order quantities, costs, and lead-times. It can allow a dynamic response for accurate, personalized, and smart end-use parts.