Advances in soil moisture measurement techniques and prediction using artificial intelligence: An extensive and systematic review

Soil moisture is a crucial parameter influencing agricultural productivity, Irrigation scheduling and climate modelling. Accurate measurement and prediction of soil moisture are essential for ensuring sustainable agriculture and mitigating environmental challenges related to water scarcity and land degradation. In this study, we carried out a Systematic Literature Review (SLR) to examine techniques and approaches used in soil moisture sensing technologies and prediction between 2018 and 2025. Following the PRISMA guidelines, we screened and analyzed studies based on defined inclusion and exclusion criteria, resulting in a comprehensive and unbiased review of the most relevant research. Particular emphasis is placed on identifying the strengths, limitations, and application domains of each method, with a focus on their relevance to environmental sustainability and precision agriculture. The review further synthesizes key challenges such as data heterogeneity, sensor calibration and the generalizability of ML models under diverse soil and climatic conditions. Future research directions are also highlighted, including the integration of eXplainable AI (XAI), hybrid physical data-driven frameworks and Federated Learning (FL) for collaborative model training across distributed datasets to ensure data privacy. Moreover, the adoption of low-cost IoT-based soil sensors powered by renewable energy and coupled with predictive models is underscored as a pathway to scalable and environmentally sustainable solutions. By consolidating individual insights and outlining research gaps, this review provides a roadmap for advancing soil moisture sensing and prediction technologies toward enhanced environmental monitoring, sustainable agriculture and resilient ecosystems.