Event-based modeling and simulation for optimizing order picking

Manually-performed order picking is a very common, yet very expensive process in warehouse operation. Multiple pickers work simultaneously in the picking area to respond to customer orders “as soon as possible”. To do so, company policies are devised to meet performance requirements, while guaranteeing minimum interference among pickers. Here we model a person-to-goods manual system by means of an event graph (EG). An EG-based representation covers the event-driven logic of the picking system, as well as the need of providing a highly-detailed description of the system logic. EGs also represent the common ground for integrating complementary analysis techniques, such as discrete-event simulation, digital twins and process mining, which allows to move towards a truly connected supply chain. We then resort to simulation to exploit the benefit of replacing a time-free order behavior with a time-window based organization for order collection. Numerical results are presented to support decision making in a real cooperative that provides wholesale distribution to customers in Central and Sothern Italy.