Penerapan Machine Learning untuk Analisis Life Cycle Costing (ML-LCC) dalam Optimasi Pemeliharaan Jaringan Pipa Air Minum PDAM Kutai Timur

Bohari⁽¹⁾, Suharman Hamzah^(1)*, Syarif Burhanuddin⁽²⁾

¹Departemen Teknik Sipil, Universitas Hasanuddin, Makassar, INDONESIA

*Corresponding author: suharmanhamzah@gmail.com

INTISARI

Penelitian ini dilatarbelakangi oleh tingginya risiko kegagalan jaringan pipa air minum di PDAM Kutai Timur akibat kondisi geoteknik berupa tanah gambut dan lempung yang korosif Penelitian ini menyajikan pendekatan terintegrasi Machine Learning – Life Cycle Costing (ML-LCC) untuk mengoptimalkan strategi pemeliharaan jaringan pipa air minum PDAM Kutai Timur.
Menggunakan dataset simulasi realistis sebanyak 5.000 observasi (500 segmen pipa × 10 tahun, 2014–2023) dengan 23 variabel input yang dikelompokkan dalam lima klaster, penelitian ini memperkenalkan indeks korosivitas tanah sebagai variabel novel yang mencerminkan karakteristik tanah gambut dan lempung khas Kalimantan Timur. Delapan algoritma ML dilatih dan dibandingkan secara sistematis; LightGBM menghasilkan akurasi regresi tertinggi (R² = 0,9813, RMSE = 2,028 Rp
juta), sementara ANN/MLP unggul pada klasifikasi risiko (F1 = 0,7699). Output prediksi ML kemudian diintegrasikan ke dalam model LCC berbasis NPV 20 tahun (discount rate 7%) menghasilkan lima strategi pemeliharaan. Strategi optimal ML-LCC menghasilkan penghematan LCC Total (finansial + sosial) agregat sebesar +Rp 9,3 miliar dibandingkan strategi Corrective eksisting, setara menghindarkan 1.609 kejadian kegagalan pipa. Expected social cost NPV selama 20 tahun
mencapai Rp 30,87 miliar, menegaskan urgensi pemeliharaan preventif berbasis data. Analisis SHAP mengidentifikasi n_bocor, failure_rate, dan indeks_korosivitas sebagai prediktor biaya paling dominan, mendukung prioritas kebijakan pemeliharaan yang konkret dan terukur.

CO2_Bohari_Penerapan-Machine-Learning-untuk-Analisis-Life-Cycle-Costing-ML-LCC-dalam-Optimasi-Pemeliharaan-Jaringan-Pipa-Air-Minum-PDAM-Kutai-Timur-Bohari.docx

REFERENSI

Bent, C., Cs, A., & Mart, G. (2019). A Comparative Analysis of XGBoost arXiv : 1911 . 01914v1 [ cs . LG ] 5 Nov 2019. 1–20.

Delnaz, A., Nasiri, F., & Li, S. S. (2023). Asset management analytics for urban water mains: a literature review. Environmental Systems Research, 12(1). https://doi.org/10.1186/s40068-023-00287-7

Elmasry, M., Hawari, A., & Zayed, T. (2017). Cost benefit analysis for failure of sewer pipelines. MATEC Web of Conferences, 120, 1–11. https://doi.org/10.1051/matecconf/201712008006

Ghobadi, F., Jeong, G., & Kang, D. (2021). Water pipe replacement scheduling based on life cycle cost assessment and optimization algorithm. Water (Switzerland), 13(5). https://doi.org/10.3390/w13050605

Journal, I., & Isbn, C. P. (2013). Application ANN: A Review. 180–184. Ke, G., Meng, Q., Finley, T., Wang, T., Chen, W., Ma, W., Ye, Q., & Liu, T. Y. (2017). LightGBM: A highly efficient gradient boosting decision tree. Advances in Neural Information Processing Systems, 2017-Decem(Nips), 3147–3155.

Matthews, J. C., Allouche, E. N., & Sterling, R. L. (2015). Social cost impact assessment of pipeline infrastructure projects. Environmental Impact Assessment Review, 50, 196–202. https://doi.org/10.1016/j.eiar.2014.10.001

Pelegrina, G. D., Duarte, L. T., & Grabisch, M. (2023). A k-additive Choquet integral-based approach to approximate the SHAP values for local interpretability in machine learning. Artificial Intelligence, 325, 1–31. https://doi.org/10.1016/j.artint.2023.104014

Poudel, A., Prasad Dahal, K., K. C., D., & Bhattarai, J. (2020). A Classification Approach for Corrosion Rating of Soil to Buried Water Pipelines: A Case Study in Budhanilkantha-Maharajganj Roadway Areas of Nepal. World Journal of Applied Chemistry, 5(3), 47. https://doi.org/10.11648/j.wjac.20200503.12

Shaik, N. B., Benjapolakul, W., Pedapati, S. R., Bingi, K., Thien Le, N., Asdornwised, W., & Chaitusaney, S. (2022). Recurrent neural network-based model for estimating the life condition of a dry gas pipeline. Process

Safety and Environmental Protection, 164(January), 639–650. https://doi.org/10.1016/j.psep.2022.06.047

Suharman, H. (2012). Development of A Practical Model for Pavement Management Systems. 京都大学. https://doi.org/10.14989/doctor.k16818

Tanjung, R. A., Yusariarta, A. W. P. P., & Wulandari, M. (2021). Brief overview on corrosion behaviour of buried structure at Kariangau industrial complex. IOP Conference Series: Earth and Environmental Science, 623(1). https://doi.org/10.1088/1755-1315/623/1/012097

Testa, F., Iraldo, F., Frey, M., & OConnor, R. (2011). Life Cycle Costing, a View of Potential Applications: from Cost Management Tool to Eco-Efficiency Measurement. Supply Chain Management. https://doi.org/10.5772/15332

Winkler, D., Haltmeier, M., Kleidorfer, M., Rauch, W., & Tscheikner-Gratl, F. (2018). Pipe failure modelling for water distribution networks using boosted decision trees. Structure and Infrastructure Engineering, 14(10), 1402–1411. https://doi.org/10.1080/15732479.2018.1443145