USULAN PERAMALAN KEBUTUHAN DAN SAFETY STOCK PERSEDIAAN CARBON BRUSH NCC G34 DENGAN CROSTON’S METHOD DAN SYNTETOS-BOYLAN APPROXIMATION (STUDI KASUS PLTU X)

Abstract

Abstrak

PT X adalah perusahaan pemegang hak operasi sebuah Pembangkit Listrik Tenaga Uap di Jawa Tengah.
Dalam kegiatan operasionalnya, terjadi keterlambatkan overhaul karena stockout spare part, salah satunya
adalah Carbon Brush NCC 634. Penelitian ini dilakukan pada divisi inventory dan didapatkan proses
perencanaan dan pengendalian persediaan spare part dilakukan secara manual dengan metode intuitif.
Oleh karena itu, penelitian ini bertujuan memberi usulan metode peramalan, jumlah peramalan, dan usulan
perencanaan serta pengendalian spare part Carbon Brush NCC 634. Dalam menentukan metode
peramalan, digunakan metode yang mempertimbangkan permintaan bernilai nol dan permintaan bernilai
tidak nol, yaitu Croston’s Method dan dibandingkan dengan metode Syntethos Boylan Approximation
(SBA). Penelitian dilakukan dengan menghimpun data historis, mengelompokkan jenis permintaan spare
part, perhitungan peramalan, dan penghitungan nilai error. Dari perbandingan error, didapatkan metode
SBA memiliki error terkecil dengan MAD 15,707 dan MSE 414,466. Hasil tersebut pun lolos uji validitas
moving range dan uji T, sehingga digunakan sebagai metode terpilih untuk melakukan peramalan.
Berdasarkan hasil peramalan, didapatkan safety stock sebanyak 24 unit dan reorder point pada 63 unit.

Kata kunci: Metode Croston; Peramalan; Persediaan; SBA; Spare part

Abstract

[Proposed Forecasting of Needs and Safety Stock of NCC 634 Carbon Brush Inventory with Croston's
Method and Syntetos-Boylan Approximation (Coal Power Plant X Case Study)] PT X is a company
holding the operating rights of a Steam Power Plant in Central Java. In its operational activities, overhaul
delays occur due to spare part stockouts, one of which is the Carbon Brush NCC 634. This research was
found that the planning and controlling process for spare part inventory was carried out with an intuitive
method. Therefore, this study aims to propose forecasting methods, the number of forecasts, and the
proposed planning and control for Carbon Brush NCC 634. In determining the forecasting method, a
method that considers demand is zero and demand is not zero, namely Croston's Method and compared
using the Syntethos Boylan Approximation (SBA) method. The research was conducted by collecting
historical data, classifying the types of spare part requests, forecasting calculations, and calculating error
values. From the comparison of errors, it is found that the SBA method has the smallest error with a MAD
of 15.707 and a MSE Error of 414.466. These results also passed the moving range validity test and the T
test, so they were used as the chosen method for forecasting. Based on the forecasting results, there are 24
units of safety stock and 63 units of reorder point.

Keywords: Croston’s Method; Forecasting; Inventory; SBA; Spare part