The evolving landscape of AI-driven risk management in the biogas production: A systematic and bibliometric review

This review presents the first combined systematic and bibliometric review synthesising artificial intelligence (AI)-driven approaches to risk management in biogas production within wastewater treatment plants (WWTPs), with emphasis on decision-optimisation and operational safety. Seven academic databases: Scopus, Web of Science, IEEE Xplore, ScienceDirect, SpringerLink, Taylor & Francis, and Google Scholar were systematically searched from 2015 to March 2025, and screening followed PRISMA 2020 guidelines. Of 3,716 retrieved records, 109 studies met the inclusion criteria. Bibliometric mapping (VOSviewer) and qualitative synthesis identified five thematic clusters: (i) biogas process safety, (ii) IoT integration and renewable-energy, (iii) optimisation and supply-chain resilience, (iv) AI-driven decision-support frameworks, and (v) advanced machine-learning techniques. The analysis reveals a marked increase in publications since 2020, reflecting a shift from conceptual modelling toward applied digital risk solutions. Europe and China remain leading contributors, although collaboration networks are fragmented and methodological heterogeneity persists. Full-scale validation of AI models in operational WWTP-based biogas plants remains limited, with most studies relying on laboratory experiments, simulations, or pilot-scale data. Constraints include publication bias, database coverage, English-language restrictions, inconsistent performance metrics, and limited access to long-term Supervisory Control and Data Acquisition (SCADA) Systems datasets. The review demonstrates that AI-driven methods have significant potential to improve safety, operational efficiency, and regulatory assurance in biogas facilities. However, achieving practical and scalable implementation will require rigorous multi-site validation, standardised evaluation indicators, integration of explainable AI, and alignment with plant-level risk-governance frameworks.