No abstract

No abstract

Machine Learning (ML) techniques and algorithms, which are emerging technologies in Industry 4.0, present new possibilities for complex scheduling methods. Since different rules can be applied to different circumstances, it can be difficult for the decision-maker to choose the right rule at any given time. The purpose of the paper is to build an "intelligent" tool that adapts its choices in response to changes in the state of the production line. A Deep Q-Network (DQN), a typical Deep Reinforcement Learning (DRL) method, is proposed for creating a self-optimizing scheduling policy. The system has a set of known dispatching rules for each machine's queue, from which the best one is dynamically chosen, according to the system state. The novelty of the paper is how the reward function, state, and action space are modelled. A series of experiments were conducted to determine the best DQN network size and the most influential hyperparameters for training.

No abstract

Deep Reinforcement Learning (DRL) has been included into the production system for multiple objectives, including control, scheduling, and maintenance planning. Maintenance must be planned sensibly and economically in order to preserve the usable life of the production systems while not sacrificing productivity and so minimising costs and losses. In this work a hybrid simulation-based and DRL approach is employed to develop an agent that can autonomously determine when to do preventative maintenance by considering the failure probability at a particular instant and the length of time since the last maintenance operation has been performed. The novelty of this approach is the configuration of the DRL setting, in particular the reward function. Results are promising comparing the approach with a heuristic from the literature, as they show that the frequency of machine failures is dramatically reduced.