Penjadwalan Probabilistik Proyek Konstruksi Berbasis Hybrid Bayesian Network-PERT

Rizqi, Safira Yulia (2022) Penjadwalan Probabilistik Proyek Konstruksi Berbasis Hybrid Bayesian Network-PERT. Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 6012202015-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (3MB) | Request a copy

Abstract

Penjadwalan proyek berbasis probabilistik bisa dilakukan menggunakan metode Program Evaluation Review Technique (PERT). Namun, penjadwalan berbasis PERT ini masih memiliki kelemahan. Diantaranya yaitu durasi aktivitas tidak dapat diperbarui secara real time, mengingat durasi PERT diperoleh berdasarkan data historis pada proyek sejenis. Pada kenyataannya, perubahan kondisi proyek, terkait sumber daya memiliki sifat yang sangat dinamis. Ketersediaan material, produktivitas tenaga kerja dan peralatan yang fluktuatif, akan sangat menentukan waktu penyelesaian proyek. Penelitian ini bertujuan untuk mengusulkan model penjadwalan probabilistik berbasis Hybrid Bayesian Network-PERT. Model ini menggabungkan metode penjadwalan probabilistik PERT dan Bayesian Network (BN). Metode BN ini digunakan untuk mengakomodasi perubahan real time dari kondisi sumber daya seperti ketersediaan material, produktivitas tenaga kerja dan peralatan pada setiap aktivitas proyek. Pengumpulan data dapat diperoleh melalui pengamatan langsung di lapangan, menyebarkan kuesioner ataupun wawancara langsung dengan pihak pihak terkait seperti konsultan pengawas, manajer proyek ataupun project control, serta para pakar/ahli yang berpengalaman di bidangnya. Validasi model dilakukan dengan mengaplikasikan model usulan pada penjadwalan proyek konstruksi jembatan. Hasil model usulan dan rencana jadwal proyek studi kasus kemudian dibandingkan untuk melihat performa/akurasi model usulan. Hasil model usulan hybrid Bayesian-PERT yang telah dibangun dalam penelitian ini menunjukkan bahwa 1) Diperoleh hasil yang akurat dengan nilai probabilitas durasi aktual proyek selesai yaitu 150 hari adalah 82,89% sehingga dapat digunakan. 2) Dapat mengakomodasi updating kejadian real time sesuai dengan kondisi pelaksanaan pada masing-masing proyek jembatan sehingga dapat membantu bagi semua pemangku kepentingan, praktisi maupun profesional dan untuk menyusun strategi dan dapat meningkatkan produktivitas kerja dengan faktor-faktor yang mempengaruhinya.
==================================================================================================================================
Probabilistic-based project scheduling can be done using the Program Evaluation Review Technique (PERT) method. However, PERT-based scheduling still has weaknesses. One of them is that the activity duration cannot be updated in real time, considering that the PERT duration is obtained based on historical data on similar projects. In reality, changes in project conditions, related to resources, are very dynamic. The availability of materials, labor productivity and equipment that fluctuates, will greatly determine the project completion time. This study aims to propose a probabilistic scheduling model based on Hybrid Bayesian Network-PERT. This model combines the PERT and Bayesian Network (BN) probabilistic scheduling methods. This BN method is used to accommodate real-time changes in resource conditions such as material availability, labor productivity and equipment in each project activity. Data collection can be obtained through direct observations in the field, distributing questionnaires or direct interviews with related parties such as supervisory consultants, project managers or project control, as well as experts/experts who are experienced in their fields. Model validation is done by applying the proposed model to the bridge construction project scheduling. The results of the proposed model and the case study project schedule plan are then compared to see the performance/accuracy of the proposed model. The results of the proposed Bayesian-PERT hybrid model that have been built in this study indicate that 1) Accurate results are obtained with the probability value of the actual duration of the project being completed, which is 150 days, is 82.89% so that it can be used. 2) Can accommodate updating real time events according to the implementation conditions for each bridge project so that it can help all stakeholders, practitioners and professionals and to develop strategies and can increase work productivity with the factors that influence it.

Item Type:	Thesis (Masters)
Additional Information:	RTS 658.404 Riz p-1 2022 3100022096862
Uncontrolled Keywords:	Bayesian network, penjadwalan proyek, PERT, produktivitas kerja, proyek konstruksi, project scheduling, work productivity, construction projeccts
Subjects:	H Social Sciences > HD Industries. Land use. Labor > HD69.T54 Time management. Scheduling T Technology > TA Engineering (General). Civil engineering (General) > TA660.P55 Underground pipeline design, construction and management
Divisions:	Faculty of Civil, Planning, and Geo Engineering (CIVPLAN) > Civil Engineering > 22101-(S2) Master Thesis
Depositing User:	Anis Wulandari
Date Deposited:	14 Oct 2025 01:55
Last Modified:	14 Oct 2025 02:11
URI:	http://repository.its.ac.id/id/eprint/128591

Actions (login required)

View Item