Soil is one of Earth's most vital resources, playing a central role in sustaining terrestrial ecosystems, food production, and climate regulation. It regulates water flow and quality, stores carbon, and supplies essential nutrients for plant growth. Understanding key soil properties, including moisture, texture, organic matter, and nutrient availability, is therefore fundamental for soil health assessment, effective land management, and sustainable agriculture. As the need for robust soil protection strategies grows, new monitoring capacities such as hyperspectral sensing are significantly expanding our data horizons. Consequently, the integration of remote sensing with AI and machine learning has emerged as a powerful framework for tracking soil degradation, providing critical support for policy-making and carbon credit systems. This Reprint presents recent advances in methodologies, workflows, and sensing technologies for estimating and monitoring soil properties using spaceborne and proximal data. The contributions address digital soil mapping, AI-based prediction frameworks, uncertainty-aware modelling, soil organic carbon estimation, soil moisture assessment, and the monitoring of soil degradation and erosion. Overall, this Reprint illustrates the expanding role of Earth observation in soil science and demonstrates how advances in AI, machine learning, and sensor integration are supporting more robust, scalable, and operational soil monitoring frameworks for research and daily applications.