Oct 9 – 10, 2025 FPEP
The Klagan Regency Hotel, 1 Borneo, Kota Kinabalu, Sabah
Asia/Kuala_Lumpur timezone
Call for Abstract - Due date: 31 August 2025

Poster - Forecasting inbound vessel movements in the Strait of Malacca using LSTM, Prophet and SARIMA

Not scheduled
1h 15m
The Klagan Regency Hotel, 1 Borneo, Kota Kinabalu, Sabah

The Klagan Regency Hotel, 1 Borneo, Kota Kinabalu, Sabah

TB 00-01, One Borneo Hyper Mall Jalan UMS, 88450 Kota Kinabalu, Sabah
Others AI-related business and economic fields Parallel Session 2: Research Paper Presentations

Speaker

Mr Ahmad Zulaidil bin Mohd Zainol (UNIVERSITI PERTAHANAN NASIONAL MALAYSIA)

Description

Marine traffic in Malaysia, especially in the Malacca Straits, has been increasing steadily over recent years, leading to challenges such as port congestion, safety concerns, and environmental impacts. (GoComet, 2024). Despite the availability of marine traffic data, it has not been fully utilized to predict and address these issues. The lack of accurate predictions makes it difficult for port authorities to manage traffic flow effectively, allocate resources, and ensure smooth operations.

Current methods used for predicting marine traffic trends are limited in their ability to produce accurate forecasts. Traditional models such as SARIMA and Prophet are commonly applied but often fail to provide reliable predictions for monthly traffic patterns. With the advancement of new technologies, there is a growing need to apply more advanced techniques like deep learning models, particularly Long Short-Term Memory (LSTM), to improve forecasting accuracy. (Neptune.ai., 2021) By combining these new approaches with traditional models, more accurate predictions for marine traffic can be achieved.

Additionally, the ability to predict marine traffic trends for the next few months is essential for effective planning and port management. Accurate short-term forecasts will help authorities better manage congestion, enhance safety, and improve resource allocation. By using LSTM deep learning models, this project aims to provide precise predictions, allowing stakeholders to make better-informed decisions for port operations.
Moreover, there is a lack of user-friendly platforms that present marine traffic predictions in a clear and interactive way. To address this, an interactive dashboard will be developed using Jupyter Notebook to visualize the forecasts. (InfoQ, 2017) This tool will help port managers, maritime authorities, and other stakeholders to better understand traffic trends and take appropriate actions. The overall goal of this project is to improve the prediction of marine traffic, contributing to safer, more efficient, and sustainable maritime operations in Malaysia.

Primary author

Siti Sarah binti Mohd Isnan (UNIVERSITI PERTAHANAN NASIONAL MALAYSIA)

Co-authors

Ms Ainul Husna Abdul Rahman (UNIVERSITI PERTAHANAN NASIONAL MALAYSIA) Mr Lt Ts. Mohamad Azrin bin Abd Azis RMN (Rtd) (UNIVERSITI PERTAHANAN NASIONAL MALAYSIA) Mr Capt Mohd Zaid Sapii (RMN) (UNIVERSITI PERTAHANAN NASIONAL MALAYSIA) Mr Ahmad Zulaidil bin Mohd Zainol (UNIVERSITI PERTAHANAN NASIONAL MALAYSIA) Lt. Mohammad Hanif Dihani Mohd Zaidi RMN (Rtd) (UNIVERSITI PERTAHANAN NASIONAL MALAYSIA)

Presentation materials

There are no materials yet.

Peer reviewing

Paper