The presentation will demonstrate the potential integration of numerical analysis and Artificial Intelligence (AI) in heavy metal adsorption, marking a significant advancement toward achieving the principles of Industrial production. Come to join us on Wednesday, December 10, 2025 at 2:30 p.m. ZOOM link: https://cesnet.zoom.us/j/96723343909

Hybrid RSM–Machine Learning Framework for Optimizing Cobalt Adsorption onto Magnetic Graphene Oxide

Mohammad Gheibi1,2*, Martin Palušák1, Michal Salava1,2, Yehor Sydorenko1,2, Mohammad Eftekhari3, Daniele Silvestri1, Miroslav Černík1, Stanisław Wacławek1,2

1 Institute for Nanomaterials, Advanced Technologies and Innovation, Technical University of Liberec, Studentská 1402/2, 461 17 Liberec 1, Czech Republic

2 Faculty of Mechatronics, Informatics, and Interdisciplinary Studies, Technical University of Liberec, Liberec, Czech Republic

3 Department of Chemistry, Faculty of Sciences, University of Neyshabur, Neyshabur, Iran

*Corresponding authors:  Mohammad.gheibi@tul.cz

Abstract: The integration of numerical analysis and Artificial Intelligence (AI) into heavy metal adsorption signifies a major advancement toward fulfilling the principles of Industry 4.0. This research aims to develop a predictive model employing Response Surface Methodology (RSM) to optimize the preconcentration of Co2+ ions in aquatic systems. Subsequently, a range of machine learning algorithms is applied to enhance the operational efficiency of the process. In addition, the sensor network architecture is structured based on the Chen-Entity Relationship Model (Chen-ERM), which is seamlessly integrated into an intelligent control framework. The study’s novelty lies in the implementation of a comprehensive Decision Support System (DSS) for Cobalt preconcentration through adsorption, achieved by combining RSM optimization with machine learning-based predictive modeling. Magnetic graphene oxide (mGO) is utilized as an adsorbent due to its superior metal-binding characteristics. When graphene oxide is functionalized with magnetic nanoparticles, it exhibits enhanced recovery efficiency for metal ions from aqueous environments. The research thus focuses on optimizing this material’s performance for applications in environmental purification and wastewater treatment. Findings indicate that the system’s pH, adsorption duration, and adsorbent dosage are the key parameters 

influencing Cobalt recovery efficiency. Optimal operational conditions were determined to be a pH of 7, an adsorption time of 188 seconds, and an adsorbent concentration of 1.67 mg·mL-1, resulting in high detection efficiency. Machine learning assessments further reveal that the Adaptive Neuro-Fuzzy Inference System (ANFIS) provides superior predictive accuracy compared to Lazy learning methods, achieving over 94% accuracy in adsorption performance prediction. Moreover, the integration of the Chen-ERM approach significantly improves the efficiency and reliability of the sensor network within the water resource management framework.

Keywords: Magnetic nanocomposites; Cobalt ions; Graphene Oxide; Artificial Intelligence; Control system design.