A reference architecture for the application of machine leaning algorithms in production planning and control
Year: 2024
Type: article
Abstract: The organization of complex production systems in enterprises poses major challenges for decision-makers in production planning and control (PPC). Especially the volatility of planning scenarios and the associated demand for short-cycle adaptation of decisions requires methods that enable fast and reliable evaluations. This article presents a system architecture that addresses these challenges holistically. The system architecture uses real production data to generate scenario-based value stream simulations, which are subsequently used to train machine learning algorithms. The trained models enable a holistic evaluation of new production scenarios. Finally, decisions can be made for the real production environment based on reliable knowledge.
Source: Procedia CIRP
Authors: Jonas Mielke, Herwig Winkler
Language: English
Subfield: Industrial and Manufacturing Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W4404789596