Systematic Literature Review of Artificial Intelligence in production scheduling problems in real cases

Industry 4.0 has revolutionized the scheduling problem by providing real-time data availability, predictive maintenance capabilities, optimization algorithms, flexibility, integration of supply chain data, and collaborative platforms. These advancements empower schedulers to make data-driven decisions, adapt to changes more efficiently, optimize resource allocation, and improve overall scheduling effectiveness. As a result, companies can enhance their operational efficiency, reduce downtime, and gain a competitive edge in the market. Consequently, scheduling AI algorithms can play a significant role in enhancing a company competitiveness in today fast-paced business environment with their related benefits. In this context and for this purpose, the European project AIDEAS ‘Fabrication Optimiser’ tool wants to realise an AI-based scheduling tool. Therefore, the first step to its realisation was to define the state of the art. In literature, there are several contributions on using artificial intelligence to solve production scheduling problems. The first search was conducted on Scopus database selecting the years 2011 to 2022 and considering only publications in the field of engineering. Since Neural Networks and Particle Swarm Optimization are the most widely used strategy for solving such problems, this study aims at analysing how authors solve real production scheduling problems, in what context and what benefits they have obtained. The results of this study show how neural network and particle swarm optimization allow for solving different types of multi-objective and single-objective programming problems in dynamic production environments while generating benefits for the company according to their needs. © 2023, AIDI - Italian Association of Industrial Operations Professors. All rights reserved.