Enhancing Operational Efficiency in Domestic Cargo Handling at Tanjung Priok Port

Abstract

Maritime transportation plays a crucial role in national development and population mobility in archipelagic countries like Indonesia. It is also key in encouraging Indonesia's economic growth, especially in frontier, outermost, and underdeveloped areas, as well as being a gateway to international trade. One of the important nodal points in sea transportation is the Port of Tanjung Priok in North Jakarta. This port is the largest and busiest, serving as the main gateway for the flow of export-import goods and the distribution of goods between islands. This research aims to analyze the loading and unloading activities of domestic goods at the Port of Tanjung Priok using secondary data from the official website of the Central Bureau of Statistics for the period from January 2007 to December 2023. The models used are Seasonal Autoregressive Integrated Moving Average (SARIMA) and Long Short-Term Memory (LSTM). SARIMA is employed due to the presence of seasonal patterns in the data. Subsequently, the SARIMA model will be compared with the Long Short-Term Memory (LSTM) model, which uses a machine learning approach to evaluate and determine the most accurate model for predicting domestic cargo handling activities at Tanjung Priok Port. Based on the RMSE analysis, the LSTM model has a lower RMSE compared to the SARIMA model, indicating that LSTM provides more accurate predictions for this time series data. However, it is important to note that a lower RMSE does not always mean that one model is generally better. Additional evaluations, such as residual analysis, other statistical tests, or prediction consistency through cross validation, should also be considered to validate the model's superiority comprehensively. This analysis is expected to provide deeper insights into port capacity planning and operational management, enabling more precise and effective decision-making in response to future demand dynamics and operational trends.