ARTIFICIAL INTELLIGENCE FOR BATTERY SAFETY, DEGRADATION PREDICTION AND CIRCULAR ECONOMY IN NET-ZERO TRANSPORT AND GRID ENERGY STORAGE: A SYSTEMATIC REVIEW AND UK RESEARCH ROADMAP

Abstract

The shift to a net zero energy system in the UK has helped to boost the use of lithium-ion batteries in EVs and grid-scale energy storage solutions. However, battery safety, degradation and end-of-life issues remain key challenges to system reliability, sustainability and economic performance. These issues are generally studied separately, with little connectivity between the various stages of the battery lifecycle and few studies addressing policy and industrial priorities that are relevant in the UK. The study is a systematic search of literature from peer-reviewed articles published from 2019 to 2025, following PRISMA framework. Literature was gathered from Scopus, Web of Science and IEEE Xplore using the following keywords: Artificial Intelligence (AI), Battery Safety, Thermal Runaway, State of Health (SOH), Remaining Useful Life (RUL), Second-life Batteries, and Battery Recycling. Studies were analysed under four themes: battery safety, degradation prediction, circular economy applications and future UK research needs after screening, and eligibility assessment. According to the results, machine learning and deep learning are applied in a wide variety of ways for fault diagnosis, thermal runaway prediction, and battery management systems. While physics-informed and hybrid AI models show promising results in SOH estimation and RUL prediction, there are still challenges such as data scarcity, model transferability, and explainability. The applications of AI in second-life battery assessment, optimisation of recycling processes, and resource recovery are emerging, but need further validation and standardisation. The most significant aspect of the study is the creation of an integrated framework that connects battery safety, degradation, and circularity via AI technologies. A UK Research Roadmap (2026–2035) is proposed that highlights key areas of explainable AI, digital twins, federated learning, AI-driven recycling and closed-loop battery systems. The roadmap offers strategic input to researchers, industry and policy makers who are helping the UK transition to a net-zero future.