The collision avoidance system is one of the core systems of MASS (Maritime Autonomous Surface Ships). The collision avoidance system was validated using scenario-based experiments. However, the scenarios for the validation were designed based on COLREG (International Regulations for Preventing Collisions at Sea) or are arbitrary. Therefore, the purpose of this study is to identify and systematize objective navigation situation scenarios for the validation of autonomous ship collision avoidance algorithms. A data-driven approach was applied to collect 12-month Automatic Identification System data in the west sea of Korea, to extract the ship’s trajectory, and to hierarchically cluster the data according to navigation situations. Consequently, we obtained the hierarchy of navigation situations and the frequency of each navigation situation for ships that sailed the west coast of Korea during one year. The results are expected to be applied to develop a collision avoidance test environment for MASS.
Accident analyse in marine industry is one of the critical issues for safety practitioners to prevent loss of life. Although considerable efforts were undertaken to prevent marine accident, numerous researches revealed that marine accidents are still on-going. In order to minimize accidents in the marine transportation, this paper presents a proactive decision- making tool which is integrating Decision-Making Trail and Evaluation Laboratory (DEMATEL) method with interval type-2 fuzzy sets (IT2FSs). As the DEMATEL enables to analyse cause and effect relationship in decision-making, the IT2FSs overcome ambiguity and vagueness of linguistic assessment of decision-makers through the DEMATEL. Thus, significant accident causal factors and their effects can be analysed on the basis of cause-effect diagram. The application of proposed approach is demonstrated with a real ship collision case. Beside its theoretical contribution, the proposed approach provides practical benefits to ship owners and operators to perceive cause and effect relationship and to avoid marine accident.
The development of soft computing techniques in recent years has encouraged researchers to study on the path planning problem in ship collision avoidance. These techniques have widely been implemented in marine industry and technology-oriented novel solutions have been introduced. Various models, methods and techniques have been proposed to solve the mentioned path planning problem with the aim of preventing reoccurrence of the problem and thus strengthening marine safety as well as providing fuel consumption efficiency. The purpose of this study is to scrutinize the models, methods and technologies proposed to settle the path planning issue in ship collision avoidance. The study also aims to provide certain bibliometric information which develops a literature map of the related field. For this purpose, a thorough literature review has been carried out. The results of the study have pointedly showed that the artificial intelligence methods, fuzzy logic and heuristic algorithms have greatly been used by the researchers who are interested in the related field.
In collision risk-based design frameworks it is necessary to accurately define and select a set of credible scenarios to be used in the quantitative assessment and management of the collision risk between two ships. Prescriptive solutions and empirical knowledge are commonly used in current maritime industries, but are often insufficient for innovation because they can result in unfavourable design loads and may not address all circumstances of accidents involved. In this study, an innovative method using probabilistic approaches is proposed to identify relevant groups of ship-ship collision accident scenarios that collectively represent all possible scenarios. Ship-ship collision accidents and near-misses recently occurred worldwide are collated for the period of 21 years during 1991 to 2012. Collision scenarios are then described using a set of parameters that are treated individually as random variables and analysed by statistical methods to define the ranges and variability to formulate the probability density distribution for each scenario. As the consideration of all scenarios would not be practical, a sampling technique is applied to select a certain number of prospective collision scenarios. Applied examples for different types of vessels are presented to demonstrate the applicability of the method.