Material Flow Control (MFC) mechanisms control the movement of jobs through a set of stationery capacity resources on the shop floor. Although the objective of MFC is item-centric, i.e. to control the flow of individual jobs, most existing MFC mechanisms are resource-centric, i.e. focus on managing the capacity resources. While this was justified by technical constraints on real-time information feedback, advances in technology allow for new designs. In particular, smart products are cognizant of their local context and can communicate with one another through the Internet of Things, thereby enabling self-organized control of individual jobs. Despite this potential most application of smart products and the Internet of Things, including multi-agent systems for scheduling and holonic control, continue to focus on hierarchical, centralized data and control structures. In response, this study develops a simple item-centric MFC mechanism and uses simulation to proof the feasibility of self-organized control.