| Issue |
RAIRO-Oper. Res.
Volume 59, Number 5, September-October 2025
|
|
|---|---|---|
| Page(s) | 3285 - 3307 | |
| DOI | https://doi.org/10.1051/ro/2025132 | |
| Published online | 29 October 2025 | |
Optimizing smart warehousing with Genetic Algorithm and Deep Q-Learning
Department of Industrial Engineering, Faculty of Engineering and Natural Sciences, KTO Karatay University, Konya, Turkey
* Corresponding author: esra.boz@karatay.edu.tr
Received:
17
March
2025
Accepted:
24
September
2025
This study uses Genetic Algorithm (GA) and Deep Q-Learning (DQL) methods to optimize the Order Picking Problem (OPP) and Storage Location Assignment Problem (SLAP) for Automatic Storage and Retrieval Systems (AS/RS). Dynamic storage policies based on order frequency are explored in the study, and the best optimization method is determined. The results show that the GA has lower order picking times and costs than the random assignment technique, while the DQL model increases the efficiency of operations through creating dynamic location policies. The study also shows that dynamic dwell points improve order picking time by 44% compared to fixed dwell points. These findings imply that logistics performance can be enhanced considerably by optimizing the location and dwell policies for warehouse managers.
Mathematics Subject Classification: 90B06 / 90B90
Key words: Storage Location Assignment Problem / Deep Q-Learning algorithm / dwell point policy / Order Picking Problem
© The authors. Published by EDP Sciences, ROADEF, SMAI 2025
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.