Rerouting and rescheduling of in-plant milk run based delivery subject to supply reconfigurability constraints

Today the concept of mass customization is becoming increasingly important for project-oriented companies manufacturing assembled products. Expectations imposed by mass customization challenges force producers to increase flexibility of exploited manufacturing systems. In turn, since systems flexibility assumes the possibility of its adaptation to the conditions and requirements set by implemented production orders, a reconfigurability understood as its ability to adjust the functionality and capacity at the correct level begins to play a key role. In this context our study considers a multi-item assembly system where in-plant transportation operations are organized in milk-run loops. Assuming that production flows both before and after the disturbance retain a cyclical nature, the considered problem boils down to whether there exist reroutings and reschedules enabling transition from one cyclical production run to a new one determined after the disturbance occurrence or not. In other words, assuming a given structure of a milk-run serviced multi-item production system as well as specifications of production orders implemented in it, the sufficient conditions guaranteeing transitions between different cyclic steady states following order sets flow are sought. The purpose of this research is to develop a declarative model and a heuristic approach which are used to define and evaluate the reconfigurability level of the considered production system. The feasibility of the proposed method of production flow re-planning is analyzed for its computational complexity and evaluated through many numerical experiments. Presented results of conducted experiments demonstrate the system-wide performance benefits of simultaneous logistic trains rerouting and rescheduling, compared to commonly used sequential trains fleet rerouting and rescheduling approaches. Outcomes from this study provide an approach to avoid time consuming computer simulation-based calculations of logistic trains routing and scheduling, aimed at updating the production flow plan adapting it to changes forced by introducing new orders and/or damage to the transport sectors.