A Decision Support Approach for Postal Delivery and Waste Collection Services

作者：Abbatecola Lorenzo^*; Fanti Maria Pia; Mangini Agostino Marcello; Ukovich Walter

来源：IEEE Transactions on Automation Science and Engineering, 2016, 13(4): 1458-1470.

DOI：10.1109/TASE.2016.2570121

• Decision support system (DSS)
• heuristic algorithms
• logistics
• vehicle routing problem (VRP)

This paper presents an urban-decision support system (U-DSS) devoted to manage, in a unified framework, the logistic services of the smart cities, such as postal delivery (PD) and waste collection (WC) services. The U-DSS architecture is proposed by describing its main components. In particular, this paper focuses on the core of the U-DSS, i.e., the model component that provides the solutions of a general vehicle assignment and routing optimization problem with the aim of minimizing the length of the routes and satisfying time and capacity constraints. In order to solve the vehicle routing problems in reasonable time, a two-phase heuristic algorithm is proposed based on a clustering strategy and a farthest insertion heuristic for the solution of a traveling salesman problem. The applicability of the proposed U-DSS is enlightened by comparing the proposed heuristic algorithm solutions with the mixed integer linear programming problem solutions of the PD and WC services. Moreover, the discussion of the real case studies of the city of Bari (Italy) assesses the proposed approach.
Note to Practitioners-The motivation of this paper is to design an urban-decision support system (U-DSS) for use by decision makers (DMs) in the offline optimal planning of the vehicle assignment and routing in different services, such as postal delivery and waste collection. The U-DSS provides solutions to the DMs with threefold important key features. First, the modeled and solved vehicle routing problem is general and can consider several landfills, set of vehicles with different capacities and speeds, several shifts and routes satisfying capacity, and time constraints. Second, the traveling times are obtained by the routes and the distances are determined by Google Map application programming interface, the traffic, and weather forecasts. Third, the solutions are obtained in extremely short times, even if the dimensions of the problem are very large. Future research aims at extending the U-DSS by considering dynamic routing and modifying in real time the planning of the routes with a suitable use of the modern information and communication technologies.

• 出版日期2016-10