The dynamic autonomy resulting from control mode transitions in the Maritime Autonomous Surface Ships (MASS) poses a significant risk to its navigation safety. While Systems-Theoretic Process Analysis (STPA) has proven effective in the hazard identification of autonomous ships, the safety challenges inherent in the dynamic autonomy of MASS with seafarers onboard controlled by three possible controllers have not been addressed much. This study proposes a three-phase framework that combines STPA, state machines and Sequentially Timed Events Plotting (STEP) diagram aiming at identifying failure scenarios in the transition process. Findings from a case study highlight the need for such a framework to understand the triggering events for mode transition, the transition process, how transition failure may occur, and refine safety constraints to ensure smooth and safe transitions. The results lay the groundwork for future research into test case generation methods that focus on these identified transition failure scenarios, intending to ensure safe and secure navigation for the progressive introduction of MASS.

海上自主水面舰艇（MASS）控制模式转换产生的动态自主性对其航行安全构成了重大风险。虽然系统理论过程分析 (STPA) 已被证明在自主船舶的危险识别方面是有效的，但 MASS 的动态自主性固有的安全挑战尚未得到解决，船上的海员由三个可能的控制器控制。本研究提出了一个结合了 STPA、状态机和顺序定时事件绘图 (STEP) 图的三阶段框架，旨在识别过渡过程中的故障场景。案例研究的结果强调需要这样一个框架来了解模式转换的触发事件、转换过程、转换失败如何发生，并细化安全约束以确保平稳和安全的转换。结果为未来研究测试用例生成方法奠定了基础，这些方法侧重于这些已识别的转换故障场景，旨在确保安全可靠的导航，以逐步引入 MASS。