Reel Tower Control Using Machine Learning

In this paper, we propose a machine learning-based approach to streamline the loading and unloading processes of electronic components in reel towers. The conventional method typically involves loading electronic components into nearby empty spaces without considering their frequency of use, resulting in increased robot movements and longer processing times. Our method utilizes machine learning to predict the interval between component requests and strategically load components based on their anticipated usage frequency. Experimental results demonstrate that our approach significantly reduces the number of robot movements compared to the conventional method. Future work will involve implementing our model in real reel tower settings to further validate its effectiveness and optimize its performance. Through these efforts, we aim to introduce more efficient management practices for electronic component storage and retrieval in reel towers, ultimately enhancing overall productivity and reducing operational costs.