Optimasi Perencanaan Load Project Maintenance Pesawat Narrow-body Menggunakan Pendekatan Machine Learning Dan Time Series

Hidayatullah, Gilang Maulana (2025) Optimasi Perencanaan Load Project Maintenance Pesawat Narrow-body Menggunakan Pendekatan Machine Learning Dan Time Series. Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 6032241019-Master_Thesis.pdf Restricted to Repository staff only until 21 January 2026. Download (1MB) | Request a copy

Abstract

Industri Maintenance, Repair, and Overhaul (MRO) menghadapi tantangan signifikan dalam merencanakan beban proyek perawatan pesawat akibat fluktuasi permintaan, ketidakpastian tren operasional, dan keterbatasan pendekatan peramalan konvensional. Penelitian ini bertujuan mengembangkan model peramalan data-driven untuk memprediksi jumlah proyek scheduled maintenance pesawat narrow-body menggunakan pendekatan time series dan machine learning. Dua model dibandingkan, yaitu ARIMA sebagai metode statistik tradisional serta Artificial Neural Network (ANN) sebagai representasi pendekatan non-linear.Data historis proyek perawatan periode 2020–2024 digunakan sebagai basis analisis. Hasil pengujian menunjukkan bahwa model ARIMA terbaik (1,1,1) menghasilkan tingkat akurasi sebesar MAPE 36,25%. Sementara itu, model ANN lag-9 memberikan performa terbaik dengan MAPE 31,90%, sekaligus menunjukkan stabilitas hasil melalui nilai standar deviasi terendah dibanding variasi lag lainnya. Berdasarkan model optimal, diperoleh prediksi sebanyak 77 proyek perawatan dalam satu tahun ke depan yang kemudian ikonversi menjadi kebutuhan sumber daya. Perhitungan selanjutnya menunjukkan estimasi kebutuhan total manhours sebesar 241.700 jam serta potensi pendapatan tahunan sebesar 109,3 MUSD, yang dilengkapi dengan interval prediksi untuk merepresentasikan ketidakpastian peramalan dan mendukung perencanaan kapasitas serta proyeksi kinerja perusahaan. Temuan ini menegaskan bahwa model machine learning lebih adaptif dalam menangkap pola data operasional MRO yang bersifat fluktuatif dan non-linear. Secara manajerial, hasil penelitian memberikan dasar kuat bagi GMF AeroAsia untuk mengoptimalkan erencanaan kapasitas, alokasi manpower, target penjualan tahunan, hingga mitigasi risiko operasional. Implementasi model prediktif ini berpotensi memperkuat fondasi tata kelola data serta mendorong penerapan pengambilan keputusan berbasis nalitik dalam proses bisnis inti perusahaan MRO.
==========================================================================================================================
The aviation Maintenance, Repair, and Overhaul (MRO) industry faces major challenges in forecasting aircraft maintenance projects due to fluctuating demand, operational uncertainties, and limitations of conventional forecasting approaches. Thi s tudy aims to develop a data-driven forecasting model to predict the workload of scheduled maintenance for narrow-body aircraft using time series and machine learning techniques. Two forecasting models, ARIMA as a traditional statistical method and Artificial Neural Network (ANN) as a nonlinear approach were evaluated and compared. Historical maintenance project data from the period 2020–2024 were used as the basis for analysis. The evaluation results indicate that the best- erforming ARIMA model, ARIMA(1,1,1), achieved an accuracy level of 36,25% in terms of Mean Absolute Percentage Error (MAPE). In contrast, the ANN model with a lag of 9 demonstrated superior performance, achieving a lower MAPE of 31,90% and exhibiting greater result stability, as reflected by the lowest standard deviation among the evaluated lag variations.Based on the optimal model, a forecast of 77 maintenance projects for the upcoming year was obtained and subsequently converted into resource requirements. The subsequent calculation indicates an estimated total manpower requirement of 241.700 man- ours and an annual revenue potential of 109,3 MUSD, complemented by prediction intervals to represent forecasting uncertainty and to support capacity planning and corporate performance projections. The findings highlight that machine learning is more capable of capturing the fluctuating and nonlinear characteristics of MRO operational data. From a anagerial perspective, the results provide a strategic foundation for GMF AeroAsia in capacity planning, manpower allocation, annual revenue targeting, and operational risk mitigation. The implementation of predictive models also supports stronger data governance and promotes analytics-driven decision-making within core MRO business processes

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	MRO, Forecasting, Time Series, Machine Learning, ARIMA, ANN, Aircraft Maintenance, MRO, Forecasting, Time Series, Machine Learning, ARIMA, ANN, Aircraft Maintenance
Subjects:	Q Science > Q Science (General) > Q325.5 Machine learning. Support vector machines. Q Science > QA Mathematics > QA276 Mathematical statistics. Time-series analysis. Failure time data analysis. Survival analysis (Biometry) Q Science > QA Mathematics > QA280 Box-Jenkins forecasting Q Science > QA Mathematics > QA76.87 Neural networks (Computer Science) T Technology > T Technology (General) > T174 Technological forecasting
Divisions:	Interdisciplinary School of Management and Technology (SIMT) > 61101-Master of Technology Management (MMT)
Depositing User:	Gilang Maulana Hidayatullah
Date Deposited:	20 Jan 2026 01:11
Last Modified:	20 Jan 2026 01:11
URI:	http://repository.its.ac.id/id/eprint/129750

Actions (login required)

View Item