Determining Demand Forecasting Methods Based on Demand Pattern Classification for Heavy Vehicle Spare Part Remanufacturing

Hermanu, Deandra Kaylatifa (2026) Determining Demand Forecasting Methods Based on Demand Pattern Classification for Heavy Vehicle Spare Part Remanufacturing. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5010221056-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (31MB) | Request a copy

Abstract

Proses remanufaktur alat berat melibatkan ketidakpastian yang tinggi, sehingga manajemen permintaan suku cadang menjadi menantang karena volume disposal yang berfluktuasi dan kebutuhan suku cadang yang tidak dapat diprediksi. Peramalan volume disposal sangat penting karena suku cadang yang dibuang tidak dapat digunakan kembali sehingga memerlukan pembelian suku cadang baru dari pemasok eksternal. Oleh karena itu, peramalan permintaan disposal yang akurat diperlukan untuk mendukung perencanaan pengadaan. Studi ini bertujuan untuk mengklasifikasikan pola volume disposal, menentukan metode peramalan yang paling sesuai untuk setiap pola, dan menghasilkan rekomendasi pengadaan berdasarkan prediksi disposal di masa mendatang. Data yang digunakan adalah data permintaan disposal mingguan dari 30 suku cadang teratas dari perusahaan remanufaktur alat berat pada tahun 2025. Pola permintaan diklasifikasikan menggunakan kerangka kerja SBC berdasarkan Interval Permintaan Rata-Rata (ADI) dan koefisien variasi kuadrat (CV2). Metode peramalan yang diterapkan adalah Single Exponential Smoothing (SES), Double Exponential Smoothing (DES), Croston, Teunter-Syntetos-Babai (TSB), dan Random Forest. Akurasi peramalan dievaluasi menggunakan MAD, MAPE, dan MSE. Hasil klasifikasi menunjukkan bahwa 26 suku cadang dikategorikan sebagai permintaan yang erratic, 2 smooth, dan 2 sebagai lumpy. Random Forest menghasilkan kinerja keseluruhan terbaik berdasarkan evaluasi. Random Forest kemudian digunakan untuk meramalkan permintaan disposal untuk minggu ke-49 hingga minggu ke-60. Berdasarkan perkiraan permintaan dan cakupan Bill of Material, Sparepart 01 dan Sparepart 02 diidentifikasi sebagai prioritas pengadaan tertinggi.
================================================================================================================================
The remanufacturing of heavy equipment involves high uncertainty, making spare part demand management challenging due to fluctuating disposal volume and unpredictable spare part requirements. Forecasting disposal volume is important because disposed parts cannot be reused, which may create shortages and require purchasing new parts from external suppliers. Therefore, accurate disposal demand forecasting is needed to support procurement planning and inventory control. This study aims to classify disposal volume patterns, determine the most suitable forecasting method for each pattern, and generate procurement recommendations based on future disposal predictions. The data used are weekly disposal demand data of the top 30 spare parts from a heavy equipment remanufacturing company in 2025. Demand patterns were classified using the Syntetos-Boylan-Croston (SBC) framework based on Average Demand Interval (ADI) and squared coefficient of variation (CV²). The forecasting methods applied were Single Exponential Smoothing (SES), Double Exponential Smoothing (DES), Croston, Teunter-Syntetos-Babai (TSB), and Random Forest. Forecast accuracy was evaluated using MAD, MAPE, and MSE. The classification results showed that 26 spare parts were categorized as erratic demand, 2 as smooth demand, and 2 as lumpy demand, with no intermittent demand identified. The forecasting evaluation showed that Random Forest produced the best overall performance. Random Forest was then used to forecast disposal demand for Week 49 to Week 60. Based on forecasted demand and Bill of Material coverage, Sparepart 01 and Sparepart 02 were identified as the highest procurement priorities.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Demand forecasting, Klasifikasi Pola Demand, Spare parts remanufaktur, Disposal demand, Demand forecasting, Demand pattern classification, Spare parts remanufacturing, Disposal demand.
Subjects:	H Social Sciences > HD Industries. Land use. Labor > HD30.27 Business forecasting
Divisions:	Faculty of Industrial Technology and Systems Engineering (INDSYS) > Industrial Engineering > 26201-(S1) Undergraduate Thesis
Depositing User:	Deandra Kaylatifa Hermanu
Date Deposited:	29 Jul 2026 06:28
Last Modified:	29 Jul 2026 06:28
URI:	http://repository.its.ac.id/id/eprint/139492

Actions (login required)

View Item