Abstract: This study is dedicated to exploring an advanced intelligent multi-objective shielding optimization method based on the third generation non-dominated sorting genetic algorithm (NSGA-Ⅲ). The program’s functionality was tested and validated to improve the efficiency and effectiveness of shielding design for spent fuel transport ship. The study employed a sophisticated multi-objective intelligent optimization procedure and designed a one-dimensional discrete ordinates scale computational model. The model played a key role in optimizing the shielding structure above the hatch cover, paying particular attention to the dimensions of the concrete and polyethylene layers. This meticulous process resulted in the selection of a series of improved shielding solutions. In addition, a realistic 3D Monte-Carlo computational model was built to rigorously evaluate the shielding effectiveness of these optimized solutions. This model was used for evaluating the effectiveness of the optimized solutions compared to conventional solutions (materials based on concrete, polyethylene or borosilicate-containing resins). The evaluation metrics are comprehensive and include parameters such as shield thickness, total weight, cumulative dose rate and economic cost. The results of the study show that the solutions optimized by the intelligent multi-objective shielding optimization method exhibit unique properties. These solutions stand out from the crowd, providing designers with richer and more varied options. The optimization technique skillfully navigates between ensuring optimal radiation protection and conserving material usage, culminating in a series of Pareto-optimal solutions that demonstrate the effectiveness of the method. The intelligent optimization procedure based on the NSGA-Ⅲ was successfully applied to the optimization of shielding solutions for the hatch region of a spent fuel transport ship. The derived optimization scheme proves a lightweight design concept and does not compromise the strict radiation protection standards. The multi-objective optimization algorithm outperforms traditional radiation shielding design methods in quickly finding a solution that meets the shielding requirements while simultaneously weighing multiple objectives, such as the dose rate, volume and weight of the shielding material. This approach significantly improves the efficiency of shielding design. The methodology described in this paper has broad applicability and is particularly suitable for scenarios involving optimization of photon shielding in major shielding geometries. However, the methodology encounters certain limitations when dealing with complex geometries such as localized shielding challenges with penetrating members, and it is recommended that 3D Monte-Carlo method is used to optimize the design in such cases.