According to the statistics of maritime accidents, most collision accidents have been caused by human factors. In an encounter situation, the prediction of ship’s trajectory is a good way to notice the intention of the other ship. This paper proposes a methodology for predicting the ship’s trajectory that can be used for an intelligent collision avoidance algorithm at sea. To improve the prediction performance, the density-based spatial clustering of applications with noise (DBSCAN) has been used to recognize the pattern of the ship trajectory. Since the DBSCAN is a clustering algorithm based on the density of data points, it has limitations in clustering the trajectories with nonlinear curves. Thus, we applied the spectral clustering method that can reflect a similarity between individual trajectories. The similarity measured by the longest common subsequence (LCSS) distance. Based on the clustering results, the prediction model of ship trajectory was developed using the bidirectional long short-term memory (Bi-LSTM). Moreover, the performance of the proposed model was compared with that of the long short-term memory (LSTM) model and the gated recurrent unit (GRU) model. The input data was obtained by preprocessing techniques such as filtering, grouping, and interpolation of the automatic identification system (AIS) data. As a result of the experiment, the prediction accuracy of Bi-LSTM was found to be the highest compared to that of LSTM and GRU.

The aim of this study is to apply a collision warning algorithm for a small fishing vessel in a fishing waterway to verify its alarm operation and to validate its feasibility. For this purpose, a scenario-based real ship test was conducted, and cases extracted from real sea data (Vpass data) were applied. Moreover, zones with frequent alarms and high-risk waters were compared. First, we installed millimeter-wave communication terminals in three small fishing vessels and applied our algorithm based on two scenarios. Furthermore, we applied the collision warning algorithm by extracting two cases encountered by multiple ships from the Vpass data. The results show that the algorithm triggered alarms continuously under risky situations. This study also compares waterway risk levels as assessed by maritime risk-assessment tools (potential assessment of risk model, environment stress model, and International Association of Marine Aids to Navigation and Lighthouse Authorities Waterway Risk Assessment Program MkII) and the locations having frequent alarms based on Vpass data collected for 7 days. Not only did the eastern sea of Yeongheung Island indicate that more alarms were triggered, but we found high-risk results from the risk-level assessment, indicating that the risky zones and the frequent alarm zones were identical. Additional research is necessary to develop an algorithm based on qualitative evaluation by actual ship operators. In addition, since fishing vessels navigate differently from general navigation methods during fishing, it is necessary to develop additional algorithms for this.

This study leveraged the millimeter wireless access in vehicular environments (mmWAVE) communication technology to reflect the maneuvering characteristics of small fishing vessels and constructed a collision prevention algorithm that can be applied relatively easily. The algorithm was verified through simulation and actual ship experiments. The algorithm had four components: detection of vessels within three miles; identification of dangerous vessels by applying the time to the closest point of approach (TCPA) and distance at the closest point of approach (DCPA) criteria; continuous monitoring of maritime traffic risk; and incremental alarm signaling. The simulations and experiments confirmed that the alarm was generated incrementally in accordance with the distance to a dangerous situation, with no false alarms. Thus, the proposed algorithm offers potential to enhance the safety of small fishing vessels.

In this study, the appropriate fairway width was designed through the maritime traffic safety evaluation model. By applying the concept of L-converted traffic, it was found that the current maritime traffic volume is 5 to 20 ships per hour in the sea area near Gun-san Port. High-density sections were identified through density analysis based on maritime traffic volume, which ranged from 5 to 10 km, and sections with an area of 10 km or more were also analyzed. Considering the general traffic flow and density, it is necessary to set a route width of 10 km or more in the target sea area, as well as to design a minimum fairway width that ensures safety for all types of ocean use. In this study, four traffic flows were analyzed through machine learning and a comprehensive environmental stress model was created for the traffic flows. The amount of safety change according to the increase in traffic volume was compared and reviewed for a fairway width of 1 km to 10 km, which can be generally formed according to the characteristics of the sea area. In the assessment, we found that when the fairway width was 1–4 km and the L-converted traffic volume was 15 or more, the ES ≥ 750 value exceeded 10% of the total, thus creating a burden on the ship operators. In particular, it was found that the ES ≥ 750 value was 20% at most when the L-converted traffic volume was 20 or more. In a comprehensive analysis, the ES > 750 was found to be less than 10% when the fairway width was 6 km or more in the sea area with one crossing passage at 5 to 20 ships per hour, which was the experimental condition of this study, and the ship operator’s burden was determined to be insignificant. Based on these results, when designing routes between offshore wind farms that will be established in the future, we expect that it will be possible to design the minimum fairway width to ensure safety according to the traffic volume.

Korean coastal waterways are potentially the most dangerous zones for marine accidents owing to a high volume of traffic, irregular traffic patterns of fishing vessels, frequent ship encounters, and narrowness of navigable area because of offshore construction and port enlargement construction. However, there are only few marine traffic assessment models that can be referenced from other countries, and none of them contain a risk assessment model suitable for the situation regarding Korean waterways. Therefore, it is necessary to develop a quantitative assessment of marine traffic environments based on ship navigator’s consciousness. To develop a new, suitable assessment model for marine traffic, the internal and external elements of vessels, based on a questionnaire survey and simulation experiment, are considered. The ship’s internal static parameters, vessel type, length, and width, as well as mariner’s experience, license, and rank on board are determined for the risk value. Furthermore, the external dynamic elements are calculated by the results of the questionnaire, which addresses the ship’s encounter situations and distances. Through this potential assessment of risk model, we used function as navigation equipment onboard to prevent a collision accident, applied it as a solution of a Vessel Traffic System (VTS) decision-making support system, and used the vessel evacuation route to recover the vessel’s normal situation from an emergency.