Adaptive Information Visualization for Maritime Traffic Stream Sensor Data with Parallel Context Acquisition and Machine Learning

Abstract

Excessive information significantly increases the mental burden on operators of critical monitoring services such as maritime and air traffic control. In these fields, vessels and aircraft have sensors that transmit data to a control center. Because of the large volume of collected data, it is infeasible for monitoring stations to display all of the information on monitoring screens that have limited sizes. This paper proposes a method for automatically selecting maritime traffic stream data for display from a large number of candidates in a context-aware manner. Safety is the most important concern in maritime traffic control, and special care must be taken to avoid collisions between vessels at sea. It presents an architecture for an adaptive information visualization system for a maritime traffic control service. The proposed system adaptively determines the information to be displayed based on the safety evaluation scores and expertise of vessel traffic service operators. It also introduces a method for safety context acquisition to assess the risk of collisions between vessels, using parallel and distributed processing of maritime stream data transmitted by sensors on the vessels at sea. It provides an information-filtering, knowledge extraction method based on the work logs of traffic service operators, using a machine learning technique to generate a decision tree. We applied the proposed system architecture to a large dataset collected at a port. Our results indicate that the proposed system can adaptively select traffic information according to port conditions and to ensure safety and efficiency.

The Impacts of the Applications of Artificial Intelligence in Maritime Logistics

“…Kim et al introduced an algorithm for automatically selecting maritime traffic stream data for the presentation from a vast quantity of data by utilising a ML technique to generate a decision tree. The proposed system appears to be capable of adapting information selection based on port conditions to assure safety and efficiency (Kim & Lee, 2019). Some of the studies are concerned with anomaly detection.…”

Revolutionizing Marine Traffic Management: A Comprehensive Review of Machine Learning Applications in Complex Maritime Systems

“…Enter Machine Learning (ML)-a burgeoning field of artificial intelligence (AI) that offers promising solutions for the numerous challenges in marine traffic management [5]. By leveraging its data analysis capabilities and predictive modeling techniques, Machine Learning can potentially transform marine traffic management, enhancing safety [6], efficiency, and sustainability in the maritime industry.…”

“…For instance, ML models can predict oil spill trajectories or emissions from vessels, helping in planning mitigating strategies [11]. 5.…”