In case of block stacking systems, one of the most important objectives of the storage and retrieval operations is to minimize the number of relocations during the retrieval operation. This study suggests two methods for determining locations of relocated blocks. First, a branch-and-bound (B&B) algorithm is suggested. Next, a decision rule is proposed by using an estimator of the expected number of additional relocations for a stack. The performance of the decision rule was compared with that of the B & B algorithm. The comparison was performed by using a container yard, a typical example of storage system of the block stack.

The port of Gioia Tauro located in Southern Italy is the largest transshipment port in the Mediterranean Sea. It is mainly devoted to transhipment activity involving mother vessels and feeders. The berth allocation problem consists of assigning incoming ships to berths for container handling, loading, and unloading operations. The processing time of the ship is not the same for every berthing point. In fact it varies with the distance from this point to the pick-up/delivery area of container storage in the yard. The management of berth allocation is made up of two interrelated decisions:

• the assignment of the forthcoming ship to a berthing point;
• the sequencing of ships assigned to every berthing point.

The port of Gioia Tauro located in Southern Italy is the largest transshipment port in the Mediterranean Sea. It is mainly devoted to transhipment activity involving mother vessels and feeders. The yard management problem is critical in a maritime terminal container management because decisions about the yard affect berth allocation. The goal is to locate containers at various locations in order to minimize movements between pick up and delivery points on the berth and storage area in the yard. A "service" or "port rotation" is a sequence of ports planned by the navigation companies for their ships. A container arriving with a service is stored in the area of this service, and is then transferred to the area of the outgoing service. Management knows the historical data of the interactions between the services. The problem can be formulated as a quadratic assignment problem with capacity constraint on the slots, since every service has a space requirement less than the capacity of the slots in the yard. In our particular case, the problem reduces to a linear ordering problem which can be formulated as an integer linear program. Unfortunately solving this problem is impractical. Instead two heuristic approaches are presented:

• a genetic algorithm, derived from a recent method proposed by Drezner for the quadratic assignment problem;
• a tabu search heuristic, which can be used alone or in combination with the genetic algorithm as a post-optimizer.