In most companies involved in optimizing transport operations, various decision-support techniques are used. Many of these tools are parameterized within the IT systems operating in companies; however, certain aspects of transport activity still do not find a place, or at least are not common, in the functional scopes of IT systems. This state of affairs does not stem from a lack of methodological background or specific computational solutions. Rather, it is caused by the high cost of implementing these instruments into IT systems. This article presents one method for supporting the planning and organization of transport activities. Before discussing selected analytical techniques, however, it is worth remembering that transport activity encompasses a broader set of operations than just moving goods, as it also includes a certain group of handling operations. All activities accompanying transport processes occur in time and space, and handle streams of specific sizes and values. Such an approach already suggests the areas and dimensions that must be taken into account when building transport models.
The transport system is characterized by high complexity. We are usually dealing with more than one supplier and more than one recipient; additionally, transit points, which can be interpreted as transshipment points, may appear in the system. In analyzing complex problems, including those concerning network optimization, it is worth starting with a graphical model. We represent the network using a graph, and we can also present certain relationships in a tabular layout using a matrix.
Fig. 1 illustrates the graph and the concepts accompanying it, thanks to which this form of flow presentation can be properly interpreted.
If we assign certain features (e.g., capacity, costs, etc.) with appropriate values to the arcs or nodes of the graph, we thereby create a network model.
To discuss the method more broadly, let us create a hypothetical situation.
The owner of three warehouses h1, h2, h3, forming the demand side H, has the possibility of sourcing good x from five different producers: P1, P2, P3, P4, P5, forming the supply side. The owner decided that the warehouses should have the possibility of replenishing stocks as follows:
• warehouse h1 up to 38 full-truckload deliveries
• warehouse h2 up to 27 full-truckload deliveries
• warehouse h3 up to 25 full-truckload deliveries
His goal was to build a portfolio of deliveries of good x for the warehouses such that the total time of transport processes TD would be minimal. Based on information available to business intelligence, he quickly determined the times (man-hours, machine-hours) for completing one full-truckload delivery from individual producers (p=1,2,3,4,5) to the respective warehouses h (h=6,7,8), later denoted as tdph. The data regarding the times tdph were aggregated in the form of a matrix.
Additionally, the quantities available from individual producers are given, which amount to:
• producer p1 offers a quantity that can handle 25 full-truckload shipments
• producer p2 can provide 20 full-truckload shipments
• producer p3 constantly offers a volume corresponding to 18 full-truckload shipments of the given good
• the next producer, p4, has the equivalent of 15 full-truckload shipments
• the last of the considered producers, p5, possesses goods corresponding to 12 full-truckload shipments.
The article comes from the magazine “Logistyka” 4/2003.
