The development of autonomous ship operation technology is progressing in the marine domain, necessitating risk analysis implementation in the development of such technology. However, a risk analysis method tailored for autonomous ship operation is yet to be established at this stage. Autonomous ships rely on software to perform tasks such as recognition, judgment, and operation, which are conventionally performed by human operators. In addition, these functions are exclusively utilized within their predetermined Operational Design Domains (ODDs), which represent assumed operational conditions. Therefore, when conducting a risk analysis for autonomous ships, in addition to the conventional equipment-based perspective, considering tasks and deviations from ODD is crucial. This study utilized an extended SWIFT method, Task-Based HAZID (TB-HAZID), using Unified Modeling Language (UML) class diagrams for hazard identification. In this paper, we present the hazards identification process through TB-HAZID for a hypothetical autonomous ship, and provide the trial results. In this paper, we show an example of hazard identification using TB-HAZID for a hypothetical autonomous ship, with the task illustrated as the focus point. Our proposed method can be applied to risk analysis of autonomous ships, which are expected to be developed based on various concepts in the future, and that the hazard identification examples can be used as a reference.
抄録全体を表示