Automated Layout Generation and Design Rationale Capture to Support Early-Stage Complex Ship Design

Abstract

Early-stage design of complex ships is of an iterative nature due to a coevolving problem definition and solution generation. In addition, design decisions are often made based on assumptions early on. To inform the stakeholder dialogue to support the definition of the design problem, designers need to gain insight into the technical feasibility, costs, and risks of requirements and potential design solutions. These aspects need to be addressed early on because of the lock-in of the concept design by, mostly, early decisions. In this situation, rework is considered a challenge because of the high cost of late design changes and might be reduced by providing designers with more accurate information on technical feasibility and risk. In this dissertation, the focus is on ship layout design. The level of detail of ship layouts is typically kept low to reduce rework during early-stage design. However, this limits how much insight into detailed design integration challenges can be obtained. Also, detailed design rationale, or design justification, is often not recorded, hampering designers in reconsidering past design decisions. In this dissertation, an automated layout design method (WARship GEneral ARrangement, WARGEAR) is developed and evaluated to support designers in quickly generating a series of detailed ship layouts. This helps designers obtain crucial insights into ship integration issues early on, potentially reducing costly rework later in the design process. Furthermore, a design rationale method (Ship Design Rationale Method, SDRM) is developed and coupled with a ship layout design tool to enable designers to capture design decisions in the context of the evolving concept design. The SDRM is consequently tested in small-scale design experiments, showing the benefits of explicitly considering design rationale during design work. The methods proposed in this dissertation allow designers, first, to identify potential sizing and integration issues in complex ship layouts, with less effort than required in current ship design practice. Second, the methods enable designers to capture their design reasoning in the context of the progressing concept design, which supports both current and future decision-making.