PERBANDINGAN METODE EARNED VALUE, EARNED SCHEDULE, DAN KALMAN FILTER EARNED VALUE UNTUK PREDIKSI DURASI PROYEK

Main Article Content

Andree Sugiyanto
Onnyxiforus Gondokusumo

Abstract

In the world of construction, control is needed at the implementation stage, which is prediction or forecasting duration project schedule. Estimated project schedule is an important part for project management making decisions that affect the future of the project. Forecasting method commonly used by practitioners in this case the construction project contractor in evaluating prediction of duration is deterministic forecasting method Earned Value Method (EVM), Earned Schedule Method (ESM). Kalman Filter Earned Value Method (KEVM) as probabilistic forecasting method is carried out to produce more accurate predictive value. The purpose of this study to compare the accuracy of three methods. This research was conducted by calculating duration of the project from EVM, ESM, and KEVM on maintenance and reconstruction projects of Jakarta-Cikampek and Jakarta-Tangerang toll roads. The data used from the project control data S-curve. The control data is processed with EVM, ESM, KEVM to determine the comparison between three methods of predicting duration. Prediction results of three methods were tested with Mean Absolute Percentage Error (MAPE). The results of this study indicate that KEVM can reduce errors after Kalman Filter is performed on estimated duration using EVM. ESM duration prediction yields the smallest MAPE value of the three methods.

 

Abstrak

Dalam dunia pembangunan konstruksi dibutuhkan pengendalian pada tahap pelaksanaan yaitu prediksi atau peramalan durasi jadwal proyek. Perkiraan jadwal proyek adalah bagian penting untuk manajemen proyek membuat keputusan yang mempengaruhi masa depan proyek. Metode peramalan yang umum digunakan para praktisi dalam hal ini kontraktor proyek konstruksi dalam mengevaluasi prediksi durasi adalah metode peramalan deterministik Earned Value Method (EVM), Earned Schedule Method (ESM). Kalman Filter Earned Value Method (KEVM) sebagai metode peramalan probabilistik dilakukan untuk menghasilkan nilai prediksi yang lebih akurat. Tujuan penelitian ini membandingkan akurasi dari ketiga metode. Penelitian ini dilakukan dengan menghitung durasi proyek dari EVM, ESM, dan KEVM pada proyek pemeliharaan dan rekonstruksi jalan tol Jakarta – Cikampek dan Jakarta – Tangerang. Data yang digunakan dari proyek tersebut adalah data-data pengendalian berupa kurva S. Data pengendalian tersebut diolah dengan EVM, ESM, KEVM untuk mengetahui perbandingan antara ketiga metode prediksi durasi tersebut. Hasil prediksi dari ketiga metode diuji dengan Mean Absolute Percentage Error (MAPE). Hasil dari penelitian ini menunjukkan bahwa KEVM dapat mengurangi kesalahan setelah dilakukan Kalman Filter pada perkiraan durasi menggunakan Earned Value Method. Prediksi durasi ESM menghasilkan nilai MAPE yang paling kecil dari ketiga metode.

Article Details

Section
Articles

References

Awwad, H. M., Valdes, J. B., dan Restrepo, P. J. “Streamflow forecasting for Han River basin, Korea.” Journal of Water Resources Planning and Management 120.5 1994, 651–673.

Barraza, G. A., Back,W. E., dan Mata, F. “Probabilistic forecasting of project performance using stochastic S curves.” Journal of Construction Engineering and Management 13.1 2004, 25–32.

Kalman, Rudolph Emil. “A new approach to linear filtering and prediction problems.” Journal of Basic Engineering 82.1 1960: 35-45.

Kim, Byung-Cheol, dan Kenneth F. Reinschmidt. “Probabilistic forecasting of project duration using Kalman filter and the earned value method.” Journal of Construction Engineering and Management 136.8 2010: 834-843.

Kim, Byung-Cheol. “Probabilistic performance risk evaluation of infrastructure projects.” Int. Conf. on Vulnerability and Risk Analysis and Management (ICVRAM), ASCE, Reston, VA 2011.

Kim, Byung-Cheol, dan Hyung-Jin Kim. “Sensitivity of earned value schedule forecasting to S-curve patterns.” Journal of Construction Engineering and Management 140.7 2014: 04014023.

Lipke, W. 2003. “Schedule is Different.” The Measurable News, Summer 2003:31–34.

Vaudrey, Toby. M163: Kalman Filter Tank Filling. September, 2008.

Wauters, Mathieu, dan Mario Vanhoucke. "Study of the stability of earned value management forecasting." Journal of Construction Engineering and Management 141.4 2014: 04014086.