WSEAS Transactions on Systems and Control
Print ISSN: 1991-8763, E-ISSN: 2224-2856
Volume 20, 2025
Robotic Agents through Scalable Multi-agent Reinforcement Learning for Optimization of Warehouse Logistics
Authors: , , ,
Abstract: Optimizing warehouse logistics is a daunting challenge, especially in today's ever-changing industrial environment. As warehouses become complex labyrinths with robots dedicated to various tasks, heuristic methods become indispensable. Based on experience and intuition, these methods offer shortcuts to solving complex challenges, enabling rapid decisions without resorting to exhaustive research. The recent research, based on pre-established rules and previous experience, has found that heuristic methods are difficult to adapt quickly to frequent changes in logistical fields. The heuristic methods are weak when faced with new situations requiring frequent change. These constraints allow us to migrate to reinforcement learning, which introduces a dynamic and continuous path in logistics environments, unlike static heuristic methods. In this environment, this paper aims to provide agents with intelligent and structured strategies to manage navigation efficiently in such dynamic logistic environments, and to meet the challenges of modern warehouses and their respective targets in real time. To achieve this, we have hybridized the BAT algorithm of Meta heuristics and reinforcement learning algorithms, which will yield remarkable results.
Search Articles
Keywords: Multi-agent reinforcement Learning, Warehouse-Logistics, Robotic Agents, artificial intelligence, Bat algorithm, Q-learning Algorithm, Sockets
Pages: 42-49
DOI: 10.37394/23203.2025.20.5