Artificial intelligence for ship design process improvement: A conceptual paper.

Abstract

This paper explores the artificial intelligence (AI) concept for complex engineering design processes in the shipping industry. It is driven by the computer technologies advancement for fast and concurrent tasks processing, machine learnability, and data-centric approach. While AI has been adopted in many industries, it is still lacking the structured approaches for practical implementation. This is especially on the generality of the methodologies and explaining AI to the non-technical members and their preparedness. Therefore, this work proposed a conceptual framework to systematically extract, represent and visualize the ship design knowledge, to develop and deploy the machine learning (ML) models, and to demonstrate the AI-based ship design processes. Comparisons to the generic ship design model were made and discussed to highlight the improvements observed. It is found that while the conventional algorithmic approach procedures were faster in terms of execution time, the stepwise empirical models were often limited by the dataset and the design assumptions with restricted estimation capabilities for solving the nonlinear ship design problems. The findings presented the impact in improving the existing processes and effectively reducing its cycle. Additionally, the approach emphasised on the validated ship design data thus its generalization for fast and wide adoptions at scales.