Eiko Saito, National Maritime Research Institute, Japan

View Abstract Sign in for premium content


More than 70% of ships involved in marine accidents are small crafts such as pleasure boats and fishing boats. In this paper, we focused on collision avoidance support systems for the small crafts using collision alerts. There are two ways to support collision avoidance by using collision alerts; one is a guard-zone alert using positions and the other is a collision alert based on Closest Point of Approach (CPA) analysis using information of the positions, courses and speeds. Also, clarification of communication time in information processing is necessary to provide timely support regardless of the aforementioned ways. Automatic Identification System (AIS) and radar Target Tracking (TT) data are used for collision avoidance support. However, they are not suitable for the crafts because of their costs and sizes. Meanwhile, smartphones become popular in late years. When the smartphones provide sufficient accuracy of the information for collision avoidance support, they can be practical and useful onboard devices to support the crafts. In this paper, we clarified the communication time, accuracy of the positions, the courses and the speeds obtained by the smartphones that affect collision alerts using results of actual sea experiments. The results are below. (1) The longest communication time was 8 seconds. (2) A maximum position error was approximately 18m when the crafts were sailing. (3) The courses and the speeds were estimated using Doppler measurement in GNSS functions and past tracks. In past tracks, we estimated using a least-square method of past 4 points' positions. (4) The errors of the courses were within ±4.0 degrees when the crafts were intending to sail on a straight course. (5) The errors of the speeds were within ±1.1 m/s when the crafts were intending to sail at a constant speed. The errors were the difference between values obtained by GPS compasses and the smartphones.