A smart decision support system for Co2+ preconcentration using RSM and AI-Driven adsorptive modeling

Using AI-based numerical analysis advances Industry 4.0 goals by developing a Response Surface Methodology (RSM) model to optimize Co 2+ preconcentration in water. Subsequently, several machine learning techniques were employed to enhance process efficiency. Furthermore, the sensor network was designed based on the Chen Entity-Relationship Model (Chen-ERM) integrated with a control system. The main novelty of this study was the implementation of an integrated Decision Support System (DSS) for the preconcentration of Co 2+ through an adsorptive process by combining RSM and machine learning techniques. Because of its remarkable adsorption capabilities, magnetic graphene oxide (mGO) was employed in this investigation. Combining graphene oxide with magnetic nanoparticles enables efficient ion recovery, aiming to enhance its use in environmental and wastewater treatment. The results of this study demonstrated that, in the order of significance, pH of the system, duration of the adsorption, and adsorbent mass were the essential elements determining the recovery (%) of Co 2+ through the adsorption process. The optimal pH, contact time, and adsorbent mass were 7, 188 s, and 1.6 mg mL −1 , respectively, for achieving maximum efficiency of the process. Machine learning evaluations indicated that the Adaptive Neuro-Fuzzy Inference System (ANFIS) model outperformed lazy learning techniques, achieving approximately 95% predictive accuracy in adsorption estimation. Similarly, the Chen-ERM model improved system efficiency in water quality monitoring.