THE QUAY CRANE DEPLOYMENT PROBLEM AT A MARITIME CONTAINER TERMINAL Pasquale Legato Dipartimento di Elettronica, Informatica e Sistemistica University of Calabria Via P. Bucci, Cubo 41C, 87036, Rende (CS), Italy E-mail: legato@deis.unical.it Daniel Gullì, Roberto Trunfio Centro di Supercalcolo per l’Ingegneria Computazionale CESIC – NEC Italia S.r.l. Via P. Bucci, Cubo 22B, 87036, Rende (CS), Italy E-mail: {daniel.gulli, roberto.trunfio}@eu.nec.com KEYWORDS Quay crane deployment problem, scheduling problem, maritime container terminals, discrete-event simulation, optimization. ABSTRACT Container unloading/loading at marine container terminals (MCTs) is a key logistic process, to which some research efforts have been addressed by using mathematical programming models formulated in a deterministic-static environment. Vice versa, DES models in a stochastic-dynamic environment are well capable of representing the entire process. Hence, simulation results to be an effective planning and control tool for decision making at all decisional levels. Here we remark that optimal decisions in MCTs may be practically pursued by modelling the whole MCT and focusing attention on the core logistic processes, while representing in a simplified manner the remainder. We focus on the operational management of the cranes deployed along the quay, during the container unloading/loading process at a given number of vessels according to a previously planned berth- schedule. We suggest a two-phase approach to the quay crane deployment problem: in the first phase an IP model is used to decide when and how many cranes must be assigned to each vessel; afterwards, we propose a heuristics to determine which specific crane should be assigned to a vessel. We indicate how this approach can be successfully integrated in a DES model, already available, to support dynamic assignment of cranes to berthed vessels. INTRODUCTION Freight transportation plays a key role in modern economies as it allows goods exchange between far-off countries. The most notable and steady technology for transporting freight, especially on long maritime routes, is containerization. A considerable growth in worldwide containerised freight transportation has been registered in recent years – approximately 90 percent of the world's cargo traffic moves by container (UNCTAD 2007). Competitiveness within the growth can be achieved by enforcing the introduction of decision support systems in the organization and management of core logistic processes involved in transportation. An efficient and effective management of logistic activities in a container terminal can decrease the operating costs and service times and increase the quality of services. A maritime container terminal is a complex facility organised around a set of logistic processes. The logistic activities at a container terminal often belong to more complex logistic processes. This fact is critical for a good management of the system and the choice of the system modelling approach. A firm classification of the decision problems in a maritime container terminal concerns to the following logistic processes (Vis and De Koster 2003; Steenken et al. 2004): i) arrival of the ship, ii) unloading and loading of the ship, iii) transport of containers from ship to stack and vice versa, iv) stacking of containers, and v) inter-terminal transport and other modes of transportation. Several interesting papers focusing on the previous logistic processes have been proposed (Legato and Mazza 2001; Park and Kim 2003; Legato and Monaco 2004; Cordeau et al. 2005; Cordeau et al. 2007; Canonaco et al. 2008). Among these works, only the ones that are based on a simulation modelling approach, to capture the dynamic and non deterministic framework, are able to evaluate large instances in a reasonable time, to conduct scenario analysis and overall performance evaluation. Besides, simulation offers the opportunity of highlighting congestion phenomena occurring at those (shared) resources resulting as bottleneck within a logistic process (e.g. the quay cranes). In this paper we propose a two-phase approach to the quay crane deployment problem (QCDP). The QCDP is a complex scheduling problem that arises when multiple vessels berths to a quay and a limited set of quay cranes must be assigned to the berthed vessels in order to respect vessels committed due-time of departure. The first phase consists of an integer programming model that produces the optimal number of cranes that must be assigned to each berthed vessels on the basis of a one-hour time-slot. In the second phase, this output data are used to assign the cranes to the vessels according to an ad hoc heuristics. We are currently integrating the proposed approach in a discrete-event simulator to support runtime crane assignment using a deterministic berth schedule. The approach has been validated using the Park and Kim Proceedings 22nd European Conference on Modelling and Simulation ©ECMS Loucas S. Louca, Yiorgos Chrysanthou, Zuzana Oplatková, Khalid Al-Begain (Editors) ISBN: 978-0-9553018-5-8 / ISBN: 978-0-9553018-6-5 (CD)