Pemodelan Prediktif Box Crane Hour dan Box Ship Hour Kapal Internasional di PT.XYZ Menggunakan Random Forest Regression

Nasrania, Sella (2026) Pemodelan Prediktif Box Crane Hour dan Box Ship Hour Kapal Internasional di PT.XYZ Menggunakan Random Forest Regression. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 2043221004-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (4MB) | Request a copy

Abstract

Aktivitas bongkar muat petikemas merupakan bagian esensial dalam operasional terminal, yang menentukan kelancaran alur distribusi logistik global. Namun, dalam pelaksanaannya proses ini masih menghadapi hambatan yang menyebabkan ketidakefisienan proses. Dalam konteks pelabuhan, kinerja layanan operasional diukur melalui Box Crane Hour (BCH) dan Box Ship Hour (BSH). Penelitian ini bertujuan untuk menganalisis kinerja operasional dermaga internasional serta mengidentifikasi kontribusi variabel operasional terhadap capaian BCH dan BSH sebagai dasar perumusan strategi peningkatan produktivitas. Hasil pemodelan menggunakan Random Forest Regression menunjukkan bahwa sebesar 96,703% kapal memiliki nilai BCH di bawah target, sedangkan 63,736% kapal memiliki nilai BSH di bawah target. Model yang dibangun memiliki kemampuan prediksi yang baik, dengan rata-rata kesalahan prediksi BCH dan BSH masing-masing sebesar 2,907 box/jam dan 4,705 box/jam. Berdasarkan periode pengamatan, seluruh dermaga internasional belum mampu mencapai target produktivitas BCH, sementara pada indikator BSH hanya Dermaga 4 yang mampu mencapai target dengan efisiensi sebesar 116,304%. Analisis Shapley Additive exPlanations (SHAP) menunjukkan bahwa variabel lama bongkar muat dan panjang kapal merupakan faktor dengan kontribusi tertinggi pada prediksi BCH, sedangkan prediksi BSH dipengaruhi oleh kepadatan crane, jumlah muatan, lama bongkar muat, panjang kapal, serta jumlah petikemas empty
=================================================================================================================================
Container loading and discharge activities are a critical component of terminal operations, as they directly affect the efficiency of global logistics flows. However, these processes still face operational constraints that lead to inefficiencies. In port operations, service performance is commonly evaluated using Box Crane Hour (BCH) and Box Ship Hour (BSH). This study aims to analyze the operational performance of international berths and to identify the contribution of operational variables to BCH and BSH as a basis for formulating productivity improvement strategies. The Random Forest Regression results indicate that 96.703% of vessels recorded BCH values below the target, while 63.736% recorded BSH values below the target. The developed model demonstrates good predictive performance, with average prediction errors of 2.907 boxes/hour for BCH and 4.705 boxes/hour for BSH. During the observation period, none of the international berths achieved the BCH productivity target, whereas only Berth 4 met the BSH target with an efficiency level of 116.304%. Shapley Additive exPlanations (SHAP) analysis reveals that handling time and vessel length are the most influential variables in BCH prediction, while BSH is primarily influenced by crane density, cargo volume, handling time, vessel length, and the number of empty containers.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Bongkar Muat, Box Crane Hour, Box Ship Hour, Kinerja, Random Forest Regression, Box Crane Hour, Box Ship Hour, Loading and Discharge, Performance, Random Forest Regression.
Subjects:	Q Science > Q Science (General) > Q325.5 Machine learning. Support vector machines.
Divisions:	Faculty of Vocational > 49501-Business Statistics
Depositing User:	Sella Nasrania
Date Deposited:	26 Jan 2026 10:15
Last Modified:	26 Jan 2026 10:15
URI:	http://repository.its.ac.id/id/eprint/130596

Actions (login required)

View Item