Digital Twin Rantai Pasok UMKM Berbasis GNN untuk Prediksi Permintaan dan Optimasi Persediaan
Abstract
Ketidakpastian permintaan dan keterbatasan visibilitas antaraktor menyebabkan UMKM sering mengalami kelebihan atau kekurangan persediaan. Penelitian ini mengembangkan Digital Twin rantai pasok UMKM berbasis Graph Neural Network (GNN) yang mengintegrasikan representasi graf heterogen, prediksi spasial-temporal, dan simulasi generatif untuk mendukung keputusan persediaan secara closed-loop. Eksperimen terkontrol menggunakan data simulasi 150 minggu pada graf berarah yang terdiri atas 36 simpul dan 78 relasi pemasok-gudang-UMKM-produk. Model residual GCN-GRU dibandingkan dengan naive forecast, Random Forest, dan LSTM menggunakan MAE, RMSE, MAPE, serta R². Prediksi selanjutnya dihubungkan dengan kebijakan order-up-to yang mengoptimalkan safety factor per simpul melalui simulasi Monte Carlo. Hasil menunjukkan bahwa GCN-GRU memperoleh MAE 4,995, RMSE 6,749, MAPE 9,887%, dan R² 0,818, lebih baik daripada LSTM dengan MAE 5,318 dan R² 0,802. Integrasi Digital Twin menurunkan biaya persediaan menjadi 26.814,62 unit biaya, atau 9,64% lebih rendah daripada kebijakan berbasis LSTM, sekaligus menurunkan rata-rata persediaan 10,85% dengan service level 99,96%. Uji sensitivitas hingga 20% gangguan pasokan mempertahankan keunggulan biaya Digital Twin. Kebaruan penelitian terletak pada penyatuan pembelajaran graf spasial-temporal dan simulasi generatif untuk mengoptimalkan kebijakan persediaan node-specific dalam satu siklus Digital Twin yang dapat diperbarui secara dinamis
Downloads
References
Chang, S., Lin, Z., Yan, B., Bembde, S., Xiu, Q., Wong, C. H., & Leskovec, J. (2025). Learning production functions for supply chains with graph neural networks. Proceedings of the AAAI Conference on Artificial Intelligence, 39(27), 27878–27886. https://doi.org/10.1609/aaai.v39i27.35004
Cho, K., van Merriënboer, B., Gulcehre, C., Bahdanau, D., Bougares, F., Schwenk, H., & Bengio, Y. (2014). Learning phrase representations using RNN encoder–decoder for statistical machine translation. Proceedings of EMNLP, 1724–1734. https://doi.org/10.3115/v1/D14-1179
Cimino, A., Longo, F., Mirabelli, G., Solina, V., & Veltri, P. (2024). Enhancing internal supply chain management in manufacturing through a simulation-based digital twin platform. Computers & Industrial Engineering, 198, 110670. https://doi.org/10.1016/j.cie.2024.110670
Corsini, R. R., Costa, A., Fichera, S., & Framinan, J. M. (2024). Digital twin model with machine learning and optimization for resilient production–distribution systems under disruptions. Computers & Industrial Engineering, 191, 110145. https://doi.org/10.1016/j.cie.2024.110145
Douaioui, K., Oucheikh, R., Benmoussa, O., & Mabrouki, C. (2024). Machine learning and deep learning models for demand forecasting in supply chain management: A critical review. Applied System Innovation, 7(5), 93. https://doi.org/10.3390/asi7050093
Kipf, T. N., & Welling, M. (2017). Semi-supervised classification with graph convolutional networks. International Conference on Learning Representations.
Kingma, D. P., & Ba, J. (2015). Adam: A method for stochastic optimization. International Conference on Learning Representations.
Li, J., Fan, L., Wang, X., Sun, T., & Zhou, M. (2024). Product demand prediction with spatial graph neural networks. Applied Sciences, 14(16), 6989. https://doi.org/10.3390/app14166989
Niu, T., Zhang, H., Yan, X., & Miao, Q. (2024). Intricate supply chain demand forecasting based on graph convolution network. Sustainability, 16(21), 9608. https://doi.org/10.3390/su16219608
Sifuentes-Domínguez, S., Mejia-Muñoz, J.-M., Cruz-Mejia, O., Pizarro-Gurrola, R., Domínguez-Flores, A.-S., & Ortega-Máynez, L. (2026). Predicting demand in supply chain management: A decision support system using graph convolutional networks. Future Internet, 18(1), 26. https://doi.org/10.3390/fi18010026
Tao, F., Zhang, H., & Zhang, C. (2024). Advancements and challenges of digital twins in industry. Nature Computational Science, 4, 169–177. https://doi.org/10.1038/s43588-024-00603-w
Zaidi, S. A. H., Khan, S. A., & Chaabane, A. (2024). Unlocking the potential of digital twins in supply chains: A systematic review. Supply Chain Analytics, 7, 100075. https://doi.org/10.1016/j.sca.2024.100075
JISTI
