Forecasting Soekarno-Hatta Airport Rail Link Passengers using Linear Regression and Exponential Smoothing Methods

Abstract

This article investigated the passenger demand forecasting for the Soekarno-Hatta Airport Rail Link, utilizing linear regression and exponential smoothing techniques. Given the in-creased significance of accurate demand forecasts in urban transportation systems, this research aims to enhance operational efficiency and service quality. Historical passenger data was analyzed from January 2024 to April 2025, revealing an upward trend in rid-ership. Three forecasting methods—linear regression, single exponential smoothing, and double exponential smoothing—were employed and compared using key accuracy metrics such as Mean Absolute Deviation (MAD), Mean Squared Error (MSE), and Mean Absolute Percentage Error (MAPE). The results indicate that linear regression yields the lowest error rates, making it the most effective method for forecasting future passenger numbers. En-hanced forecasting models, as proposed in this study, can significantly contribute to urban transportation planning and service optimization. From the research, indicate that the line-ar regression method is better in terms of forecasting accuracy, making it the preferred choice for predicting passenger demand for the airport rail service. The forecasts indicate a consistent and gradual increase in passenger numbers, starting from 770,025 in period 17 and reaching 840,457 by period 24. This steady growth pattern aligns with the upward trend observed in previous historical data, suggesting that increasing access to and aware-ness of the airport rail service will continue to drive ridership. The forecasts assist in opera-tional decision-making, allowing for adequate planning of resources, scheduling adjust-ments, and infrastructure needs to accommodate the projected growth in demand.