This paper presents a solution to a challenge faced in the supply chain management of a spare parts distributor with a dispersed global supply network and local distribution network in Morocco. The problem is a lack of accurate lead time information, leading to difficulties in meeting customer demand. The proposed solution is a framework using an LSTM (Long Short Term Memory) model for lead time forecasting and anomaly detection. The framework combines CNN (Convolution neural network) -Bidirectional LSTM model for forecasting and an LSTM autoencoder with One-Class Support Vector Machine for anomaly detection. The data was obtained from a legal ERP system of a major automotive distributor in Morocco. The results show that the framework is effective in overcoming the lead time information issue, and the relevance of the methods used has been verified via precise performance indicators, such as the RMSE (Root of the mean of the square of the errors) testifying to the accuracy of the results and also through comparison with other models.

@article{oscm-2023-86,
  title={Improving Lead Time Forecasting and Anomaly Detection for Automotive Spare Parts with A Combined CNN-LSTM Approach},
  year={2023},
  author={Asmae Amellal and Issam Amellal and Hamid Seghiouer and Mohammed Rida Ech-Charrat},
  journal={Operations and Supply Chain Management: An International Journal},
  volume={16},
  number={2},
  pages={265--278},
  doi={10.31387/oscm0530388}
}

Asmae Amellal, Issam Amellal, Hamid Seghiouer, Mohammed Rida Ech-Charrat (2023). Improving Lead Time Forecasting and Anomaly Detection for Automotive Spare Parts with A Combined CNN-LSTM Approach. Operations and Supply Chain Management: An International Journal, 16(2), 265-278. https://doi.org/10.31387/oscm0530388