dc.description.abstract | “Necessity is the mother of all inventions” – this famous proverbial saying (unknown author) aptly suits the maritime shipping industry in the context of autonomous ships. Though shipping industry has had highly automated systems onboard specialised ships like DP vessels, fully autonomous shipping operations has eluded it for some time until recently.
The push towards increased innovation and testing of autonomous shipping has primarily begun due to the need for cutting operational costs, for increasing safety at sea, for increasing productivity and for reducing carbon-footprint to make shipping more sustainable to meet IMO’s Greenhouse gas emission targets. It has also been ably supported by the enabling environment created by government policies worldwide, research institutions, shipping companies and ship classification societies.
In order to achieve fully autonomous shipping (or unmanned) operations, the ship besides replicating human senses of an onboard operator like vision, hearing and communicating – will also need to have the situational awareness and decision-making skill of humans especially expert seafarers with long experience.
Hence, a risk analysis method is required which can acquire the virtue of expert seafarers and provide accurate decision-making support to the ships autonomous system enabling it to take navigational decisions of its own without human-intervention.
The real-time risk analysis method looks promising in this regard. The objective of this thesis report is to establish a sound body of knowledge about real-time risk analysis, and to apply it to build a real-time risk analysis model for autonomous ships. For this purpose, the Sundbåten autonomous passenger ferry project which is currently under way is taken as a case-study. Here the mission is to develop a real-time risk model which is capable of warning the captain to take the ship’s control when its autonomous system is incapable to do so. The real-time risk analysis model developed in this thesis is capable of identifying the critical risks from marine traffic analysis and expert judgements. The framework for risk model looks promising and its modular and flexible architecture makes it adaptable for a variety of ships & regions. | en_US |