Digital twin-based shop-floor reconfiguration design for uncertainty management

Abstract

Reconfiguration has emerged as a paradigm that adjusts shop-floor operations based on production requirements in such a way as to obtain desired performance and adapt to changes quickly. Existing research puts more emphasis on optimising the layout and configuration of shop floors based on simplified mathematical models and algorithms. However, uncertainty and complexity of shop-floor operations could diminish the effectiveness of the reconfiguration solution. This paper presents a digital twin-based design method for shop-floor reconfiguration to solve the problems. Digital twin models with dynamic fidelity are used for reconfiguration design to provide effective information of shop-floor complexity. To analyse the impact of uncertainty on shop-floor performance and then guide the shop-floor reconfiguration design, methods of shop-floor performance fluctuation identification, uncertain event extraction, reconfiguration operation domain division are proposed. The proposed method is verified with the case study of a chemical fibre silk cake packing shop floor.