Utilizing artificial intelligence in cutting plan generation of ship structure in ship recycling process

Abstract

Traditional ship dismantling in developing countries endangers workers, environment, and social safety as it involves intensive manual labour. While use of dry dock method is a safer method of vessel dismantling, due to higher costs of the green yards, ship owners deny selling their vessels to them. Use of innovative artificial intelligence algorithms in generation of cutting plan for ship hull which involves most of the steel mass of a vessel can aim to pave the way for green yards to lower their costs. That is because in current practice, cutting plan for a vessel is generated by human expertise which is costly, subjective, and time-consuming. Utilizing AI algorithms can lower the need for human expertise while ensuring that the plan is optimized and feasible. As the plan is optimized, cutting operations become more efficient and the emissions reduce. The feasibility of plan is considered by imposing constraints of the shipyard facility such as lifting and transportation capacity and consideration of stability while the cuts are performed. To implement AI algorithms, a set of reliable data from digital model of a vessel would be needed to ensure that the generated plan is not artificial. Since the needed data as input for AI algorithm is impossible to be achieved by interacting with user interface of CAD software, application programming interface is used instead to automatize the process. Different algorithms have different characteristics that the proposed problem needs to be defined such that the objectives can be realized. After deploying AI algorithms, their achieved solution can be validated by analysis of convergence and comparison of different algorithm solutions. Finally, a discussion on performance of algorithms and their adequacy for this specific problem is done.