Abstract The objective of the paper, knowing the number of ports and ship fleet, is to optimises maritime transport routing of a containership, based on demand scheduling to each port of call ,using the expert system approach with owner utility function (McFadden D. 2000). All that need the operative cost of ships employed and their technical characteristics. The problem solution will be given, for each ship of the fleet, by routing of the ships , container movement for each port of call and transport cost. This paper proposes the use of methodology based on an expert system computation program with a random utility function of a shipowner operating in a maritime network mapped by geographical information system GIS ( Catalani M. 2001).