Aqueous cadmium (II) removal using integrated technique of phytoremediation with artificial intelligence and machine learning models: a review

The major environmental concern today is the proper conveyance and complete treatment of wastewater. Several novel treatment techniques are under constant research. Among emerging methods, phytoremediation is treated as the most promising alternative for the remediation of toxic heavy metals from aqueous systems. The prominent ill effects of aqueous cadmium and the elimination of this aqueous cadmium (II) (Cd(II)) using the technique of phytoremediation have been the main motivation for this research. This review is an attempt to present the potential to critically examine the applications of artificial intelligence (AI) and machine learning (ML) models in the phytoremediation process. This amalgamation would help make the process of phytoremediation more feasible due to an improved control, reduced cost, and increased efficiency of the process, which can be integrated into this process to improve the performance, predictability, and scalability. The sources of literature comprised of Scopus, Web of Science, and PubMed, databases from the year 2000 to 2024. The keywords used were ‘cadmium (II),’ ‘Wastewater treatment,’ ‘Phytoremediation’ ‘Artificial intelligence,’ and ‘Machine learning.’ Studies related to aqueous Cd (II) removal have been reviewed and the use of AI and ML models for the mentioned purpose has been evaluated. The novelty of this manuscript is in the integration of phytoremediation with AI and ML techniques, with a detailed discussion of the merits and demerits of this wastewater remediation appoach, in the context of cost and applicability. The mechanisms of the combined treatment techniques have been studied by evaluating several models, such as Artificial Neural Networks (ANN), Support Vector Machine (SVM), Random Forest (RF), regression, and other hybrid models, to aid the research. The challenges in adopting these methods, like overfitting, data security, and limited scalability when used on-site, have been discussed. An examination of ecological risks has been made at large. This review targets the gaps, limitations, and also suggests possible changes for decentralized application of this Explainable Artificial Intelligence (XAI) along with low-energy sensor technique at the user end.