To manage the overall design process of complex systems like ships, particular attention has to be deserved to the connections and overlapping between the design stages. Therefore, the paper addresses the complex aspect of quickly moving from concept to preliminary design, so yielding an initial design in a decreasing time span. Concept design is here treated as a MADM synthesis process, where a huge number of feasible designs are randomly generated by an adaptive Monte Carlo sampling and ships’ properties are assigned by means of metamodeling techniques. Non-dominated solutions are then filtered as a Pareto set identifying a short list of preferred solutions up to the “best possible” design characterized by top-level specifications (geometry, ship performance, capex, opex, etc.). This procedure has its core in the trade-off between technical and economic attributes also embracing uncertainty in a ship’s lifetime perspective. A set of ship parameters are oriented to preliminary hull geometry and general arrangement definition. Hierarchical dependencies are used to match the concept hull form with modular internal spaces. The process is here applied to a case study, consisting in the initial design of a Compressed Natural Gas (CNG) ship as member of a fleet optimised for gas shipping from the Zohr field (Egypt) to the Adriatic Sea.
