Logistical Operations

Introduction

Whether you want to get as much stones possible from one place to another to build your pyramid, or you want to transport supplies for your troops via the quickest line in order to outcompete the enemy, or simply are trying to imagine the capacities Santa Claus would need to be able to deliver his presents to all the children of the world, transportation and logistics are almost everywhere. Consequently, the associated fields are big business. Finding the cheapest way to get from point A to point B may seem like a walk in the park. But the playfield alters when one adds an extraordinarily amount of goods and multiple different possible routings. To help companies overcome these difficulties, and of course to make big money, DHL offers its services as an intermediate transport manager. This case, based on real data, focuses on one of such projects of DHL, namely on medicine giant Pfizer. Pfizer, mostly known for its Viagra, is responsible for the production and delivery of a large number of medicine and cosmetic products. After Pfizer has made these products, they have to be delivered to their customers on time. As we are dealing with medicine, there is more on the line than just money. A failure in delivery or not being able to meet the deadline could result in possible loss of lives. This case will therefore focus on delivering on time with the least costs possible.

This case represents a vehicle routing problem. The problem at hand, which we are about to solve in this report, focuses on delivering shipments that have their origin in Europe to the four locations of interest, namely ATL or Atlanta, Mex or Mexico, Caracas and Valencia contained in VE or Venezuela. The dimension of the problem is higher than simply finding cheapest ways because their deadlines for delivery differ. As those deadlines are meant to be made, this makes the problem more complex. Also, some products need to be heated or cooled down to certain temperatures during transport. So they need to be handled with care. Therefore, the problem can be best described as a multiple vehicle routing problem with a dynamic system of routings depending on time. These routings depend on time as we do not control the traffic of airplanes. We simply take the travel times of the airplanes as exogenous and model the problem conformingly.

Defining the Data

We are given the actual data on shipments of 2013 and 2014. These data contain the city of origin, the destination, the type of commodity, the chargeable weight (which takes into account the volume and weight of the package), the actual weight, the number of pieces, the corresponding dates of receiving the package and the deadline for delivery and the consignee. Also we have been given a file containing a set of exchange rates in order to be able to convert currency values and a file containing the distances between airports