Papers

Global supply chain optimization to achieve better efficiency in respect of environmental constraints recently motivated several research works on the idea of the physical internet as a worldwide open logistics system intending to bring new models that make the current logistical systems more flexible and sustainable. The developed concept aims to profoundly change the way objects are handled, stored or moved, taking inspiration from the digital internet. Physical Internet reproduces many concepts from the digital Internet, which are based on packet data transmitting in the TCP-IP protocol using routing protocols.

The objective of this study is to evaluate the contribution of physical internet on reducing logistics costs and improving the quality of service in PI-cross docks. Given the significant difference between the digital and physical systems, such a study is necessary to evaluate the efficiency of investments linked to the implementation of physical Internet.

Two simulation models are proposed to compare performances of a classical cross dock and a PI-cross dock under the same flow.

The first model is for the simulation of a real and classic cross dock with an incoming flow of three different types of household appliances products (type 1, 2 and 3) coming from different suppliers. The model considers three incoming docks, three outgoing docks and a warehouse separated into three areas (one area for each type of product). The cross dock uses 9 forklifts for unloading, loading and storage. Each one of the three incoming docks handles one type of products and each one of the incoming trucks comes with one type of products. For example, an incoming truck carrying product type 1 will be unloaded in dock 1. When a truck arrives in an incoming dock, products are unloaded in the reception area of the dock using the available forklifts, and each one of the incoming docks has one reception area. In the other side of the cross dock facility, there are three outgoing docks. Trucks arrive with their orders related to one or several products. In order to respect the FIFO rule, forklifts pick up products from the warehouses first. But if the quantity in the warehouse is not sufficient, the requested quantity will be picked up directly from the reception areas, if there is a truck unloading the requested product.

The second model is for the simulation of the automated cross dock (PI-cross dock), keeping the same flow of incoming and outgoing trucks and the same level for receptions and orders. The characteristics of the cross dock facility are also the same as for the first model (the surface and the number of incoming and outgoing docks). However, instead of using forklifts, automated loading and unloading resources (PI-docks) connected to a sorting area (PI-sorters) using PI-conveyors are considered. In the PI-cross dock, manual forklifts for storing and retrieving products in the temporary warehouse are replaced by an automated storage and retrieval system (AS/RS) that is connected directly to the sorting area with three storage and retrieval machines, one machine for each kind of product. When a truck arrives to an incoming dock, the PI-dock unloads automatically the products which will be moved to the PI-sorters using PI-conveyors. Trucks’ orders are served using the available products in the warehouse. The required products are picked up using the AS/RS system and then they come through the sorting area (PI-sorters) and to the PI-dock through PI-conveyors. However, if the requested quantity is not sufficient to serve the truck and if there is a truck being unloaded in the incoming PI-dock then the products arrive directly from that PI-dock.

The objective of this simulation is to compare and evaluate the KPIs (Key Performance Indicators) of each one of those cross docks such as the total time spent by a product in the cross-dock, the waiting time of incoming and outgoing trucks, and the usage of resources under the same flow of incoming products and orders, keeping the same surfaces for the two cross docks.

After running the two simulation models under the same flow, the average total time (cycle time) spent by the three products in the cross dock is reduced by 31%. The waiting time in the docks is reduced by 90% and the resources usage is reduced by 87%. Further research is ongoing on more complicated configurations related to the system of an industrial partner. However, from those initial results; it is very clear that if the arrival and the departure of trucks are well synchronized, PI-cross docks can greatly improve actual quality of service and resources occupation.