Predicting Vessel Speed Using Ensemble MLR and RF Models on AIS Data

Abstract

Defence agencies like the DRDO, Indian Navy, Indian Army, Indian Air Force (IAF), and many other organizations require ongoing oversight of their applications and devices to detect anomalous behaviour. The purpose of this research is to track the ship’s speed to facilitate navigation. AIS data plays a substantial role in this investigation, as it provides critical information such as the speed, course, and position of a vessel. Monitoring vessel speed is essential to ensure smooth functioning, enabling collision avoidance, route optimisation, and accurate analysis of the ship’s schedule for transportation. In this study, an ensemble approach using Multiple Linear Regression (MLR) and Random Forest (RF) models to predict vessel speed more accurately is developed. The proposed ensemble model outperformed existing methods, showing significant improvements in prediction accuracy and robustness. This enhanced performance also aids in the precise Estimation of the Time of Arrival (ETA) of the vessel, contributing to more efficient operation procedures and environmentally friendly practices. Early oversight of vessel speed ensures maritime navigation and safety, promoting reliable and optimized routes for vessels.