Designing Self-Adaptive Automation Systems Using Reinforcement Learning for Real-Time Decision Optimization in Manufacturing

Abstract

The integration of artificial intelligence (AI) in industrial manufacturing has enabled substantial advancements in autonomous decision-making and operational optimization. This paper explores the development of self-adaptive automation systems using reinforcement learning (RL) for real-time decision optimization in manufacturing environments. It focuses on how RL can be leveraged to enhance adaptability, efficiency, and scalability of manufacturing systems under dynamic conditions. Building upon existing research, the study proposes a framework incorporating RL-based agents capable of learning and adjusting operational strategies in response to changing production demands and environmental uncertainties. A comparative analysis of different RL algorithms is conducted, and simulation results demonstrate the effectiveness of the approach in reducing cycle time, improving resource utilization, and minimizing energy consumption.