Maritime Autonomous Surface Ships: Automation Transparency for Nearby Vessels

Abstract

In the future, maritime autonomous surface ships (MASS) will become more advanced, and mariners will move from the bridge to remote operation centres. To avoid dangerous situations and frequent intervention from the operators, a MASS needs to perceive the environment and calculate the current state and future intention of nearby ships. Another challenge for a MASS is to explain the decisions of its algorithms and make the artificial intelligence's decisions transparent to not only the operators, but also external stakeholders, such as passengers, vessel traffic services, and nearby ships. This is called automation transparency. In this paper we focus on how MASS, through various means of maritime communication, can (1) express their own state and intention to other ships and (2) perceive other ships' state and intention. Towards that, we first analyse current maritime communication methods in terms of how well they can be expressed and perceived by today's MASS in comparison to traditional ships. Next, we analyse which gaps in their communication should be focused on to maximise safety at sea. Finally, we present multiple design recommendations for naval architects, human factors engineers, and interaction designers who are designing and building the next generation of MASS.