We reviewed the literature on soil health definition, indicators and assessment frameworks, highlighting sensing technologies' significant potential to improve current time-consuming and costly assessment methods. We proposed a soil health assessment framework from an ecological perspective free from human bias, that leverages proximal sensing, remote sensing, machine learning, and sensor data fusion to enable objective, rapid, cost-effective, scalable, and integrative assessments.