Utilizing Machine Learning Techniques in Inventory Planning and Control of Automotive Spare Parts
Loading...
Date
2024-11-25
Authors
Baha’a Dweikat
Galina Zagha
Leen Atout
Yasser Isteitieh
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Abstract
The automotive industry facing significant challenges in inventory planning and control activities, particularly in managing the availability of spare parts. In this project, machine learning techniques were utilized to develop and analyze several models to enhance the inventory planning and control of automotive spare parts, and their performance was compared with traditional methods to show their applicability. The focus was placed on employing Artificial Neural Networks (ANN) and Autoregressive Integrated Moving Average (ARIMA) models, and their effectiveness was compared against the traditional Moving Average method.
The primary objective was to enhance the accuracy of inventory forecasts and improve overall inventory management efficiency. Historical data were collected and analyzed, models were trained and tested, and their performance was evaluated. Specific constraints were encountered, such as the lack of comprehensive historical data and limited software availability. The unavailability of suitable software restricted the development and analysis of support vector machine models, leading to a concentration on ANN and ARIMA models.
It was demonstrated that machine learning models, particularly ANN, provided more accurate forecasts compared to the traditional Moving Average method. Superior performance in capturing complex patterns in the data was shown by the ANN model, whereas the ARIMA model offered robust time series forecasting capabilities. The traditional Moving Average method, while simpler, lacked the sophistication needed to handle the details of automotive spare parts inventory.
Overall, the potential of machine learning techniques in revolutionizing inventory planning and control in the automotive industry was underscored by this project. By leveraging advanced models, significant improvements in forecast accuracy and inventory management can be achieved, ultimately contributing to more efficient operations and reduced costs.