Optimization Planning Scheduling Problem in Industry 4.0 Using Deep Reinforcement Learning

Industry 4.0 provides us with more precise real-time information about each factory element. Reinforcement Learning gives us new opportunities to improve old methodologies to resolve planning and scheduling problems using this information. Reinforcement Learning models can learn about old Master Plans and correct the mistakes that traditional algorithms cannot predict. This proposal improves the form to create a plan reducing the backlogged cost using Reinforcement Learning, specifically by means of a DQN Agent.