Leveraging deep learning for risk prediction and resilience in supply chains: insights from critical industries
Abstract Supply chain resilience (SCR) is crucial for firms and organizations to respond swiftly and effectively to operational disruptions, ensuring smooth transitions from raw materials to final products. To enhance SCR and mitigate risks, this paper proposes a deep learning (DL) framework that he...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
SpringerOpen
2025-04-01
|
| Series: | Journal of Big Data |
| Subjects: | |
| Online Access: | https://doi.org/10.1186/s40537-025-01143-4 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Abstract Supply chain resilience (SCR) is crucial for firms and organizations to respond swiftly and effectively to operational disruptions, ensuring smooth transitions from raw materials to final products. To enhance SCR and mitigate risks, this paper proposes a deep learning (DL) framework that helps predict supply chain risks and provides companies with actionable insights to overcome them. We applied five DL models—recurrent neural networks (RNN), long-short-term memory (LSTM), gated recurrent units (GRU), convolutional neural networks (CNN), and artificial neural networks (ANN)—across four key case studies: pharmaceuticals, automotive, agriculture, and energy (General Electric). In the pharmaceutical case, DL models were used to optimize drug inventory and logistics, reducing wastage and stock-outs. For agriculture, the models predicted food demand for each country annually, helping ensure efficient supply management. In the automotive industry, the models predicted car demand, while in the energy sector, they forecasted electricity needs. Our results showed that the CNN model achieved the highest accuracy, with 99.3% in the pharmaceutical case study. In the food and automotive sectors, the GRU model had the lowest mean absolute error and mean squared error, while the ANN model performed best in predicting electricity demand. When compared to traditional machine learning models, our DL models demonstrated superior accuracy and performance. This research provides firms and organizations with a valuable tool to predict future disruptions in their supply chains, improving resilience by identifying potential risks and taking proactive measures to address them. |
|---|---|
| ISSN: | 2196-1115 |