Optimizing Real-Life Freight-Distribution Problems

Real-life freight-distribution problems are characterized by a high degree of complexity. Although this kind of problem has been addressed broadly by the operation research community, models proposed in the literature are not able to completely describe real distribution problems, and while attempting to simplify and generalize the problem, they neglect aspects that could be of crucial importance. In real-life applications a huge variety of constraints may be faced. Some concern vehicle features, such as multi-dimensional vehicle capacity (weight and volume) and vehicle characteristics (refrigerated vehicles for perishable food, armored vehicles for jewelry items, special vehicles for hazardous materials or for medicines, etc.). Some others deal with time restrictions. In fact, route duration and length restrictions may be found in drivers’ national contract agreements, customer time windows, and road accessibility time windows (traffic-limited zones). Others constraints address product-product, product-vehicle, and/or customer-vehicle compatibility. Moreover, real-life applications often present objective functions more complex that go beyond the classical distances minimization. Different factors may be taken into account, such as driving costs, environmental costs, risk level (for hazardous materials transport), customer satisfaction, and others. In several cases multiobjective function optimization is required. The aim of this article is to describe relevant issues arising in real-life freight-distribution problems and to discuss the varying degree of importance they assume in different contexts. The article provides a deep investigation of models and algorithms addressing real-life problems presented in the literature and a critical dissertation on still-open issues and gaps between models and real-life applications. Promising avenues to be followed to create powerful and Real-life freight-distribution problems present high degrees of complexity mostly derived from needing to respect a variety of constraints and addressing complex objective functions in which different factors are taken into account. The aim of this article is to point out relevant issues arising in real-life freight-distribution problems and to describe still-open issues and gaps between models and real-life applications. Furthermore, a real-life vehicle-routing problem is presented and a real instance related to regional freight distribution in Piedmont (northwest Italy) is solved by a fast heuristic method. The extreme portability of this method is then discussed.