Analysis of Consumption Patterns and Prediction of Subsidized Fuel Availability Based on Dashboard Fusion Data in The Retail Sales Area of Bengkulu City
DOI:
https://doi.org/10.70656/ijcse.v3i01.908Keywords:
Pertalite, Biosolar, Fuel Consumption Patterns, Fuel Availability Prediction, Comparison between Gas Stations, Bengkulu CityAbstract
Subsidized fuel oil (BBM) is an important strategic commodity in supporting transportation activities and ensuring the smooth mobility of the community. Fluctuations in consumption and imbalances in the distribution of subsidized fuel have the potential to cause problems such as shortages at some gas stations, leading to long queues and excess or insufficient stock at other gas station locations. This study aims to analyze the consumption patterns of subsidized fuel (Pertalite and Biosolar), compare the consumption levels between gas stations, and predict the fuel needs for the next period. The data used consists of daily sales data from six gas stations in Bengkulu City during the period from January to March 2026, obtained from the Fusion Sales Area Retail Dashboard. Data processing was carried out using Python with the pandas library for data cleaning and matplotlib for visualization. The prediction method used is a 7-day moving average, with model evaluation using Mean Absolute Percentage Error (MAPE). The research results show that the pattern of fuel consumption is fluctuating, with Pertalite dominating total sales compared to Biosolar. In addition, differences in consumption between gas stations are closely related to geographical location, traffic density, and the dynamics of local economic activities. The prediction results show a relatively stable consumption trend with a gradual increasing tendency, with a MAPE value of 9.20%, indicating a good model accuracy level. This research contributes by providing an analytical approach based on operational data to support the optimization of fuel distribution management and more effective decision-making. However, this research still has limitations in the scope of data and the simple methods used, so future research is recommended to adopt more complex models by including external variables.
Downloads
References
[1] N. Chairunnisa dan N. Kahar, “Analisis Data Penjualan Berbasis Visualisasi Python,” Seminar Nasional Informatika (SENATIKA), pp. 205–213, 2026.
[2] A. P. Candra, “Analisis Data Menggunakan Python: Memperkenalkan Pandas dan NumPy,” Journal of Information System and Education Development, vol. 3, no. 1, pp. 11–16, 2025.
[3] S. Ishak dan N. Talango, “Analisis Konsumsi Bahan Bakar Motor Menggunakan Campuran Bio Fuel dan Pertalite,” Jurnal Simetrik, vol. 14, no. 2, pp. 910–916, 2024.
[4] Z. Ruma, “Pengaruh Estimasi Permintaan Bahan Bakar Pertalite,” Jurnal Ilmu Manajemen, Bisnis dan Ekonomi, vol. 3, no. 4, pp. 131–137, 2025.
[5] P. P. Nduru dan G. Kholijah, “Peramalan Permintaan dan Optimasi Persediaan BBM Menggunakan Moving Average,” JISTech, vol. 10, no. 1, pp. 96–106, 2025.



.png)
.png)



.png)
.png
)


