A Review of AI and IoT Integration Frameworks for Sustainable Soil Health Management

Abstract

The rising global population and increasing food demand, exacerbated by extreme climate changes in Sub-Saharan Africa, have accelerated soil degradation on arable lands critical for crop production. This has prompted calls for continental and national interventions to mitigate the trend. Among the proposed solutions are digitally enabled tools leveraging Artificial Intelligence (AI) and the Internet of Things (IoT), which have demonstrated significant success in developed regions. Recent advancements—such as cheaper, more accurate sensors, Edge AI, lightweight machine learning (ML) models running on edge devices, and open-source microcontrollers with enhanced computational capabilities—present promising opportunities for Sub-Saharan Africa, where small-scale farmers dominate agriculture. This study conducts a systematic literature review to evaluate AI and IoT integration frameworks for sustainable soil health management, examining the types of tools deployed and their contributions. The review analyses peer-reviewed publications from 2014 to 2024, sourced from Scopus, Google Scholar, the ACM Digital Library, and AJOL-CGspace. Findings indicate that while a layered approach framework is commonly adopted, it is often modified to suit specific contextual needs. However, many tools remain in the experimental stage, with proof-of-concept demonstrations but limited quantifiable evidence of their impact on sustainable soil health management. A key limitation of this study is the scarcity of long-term, large-scale field trials in Sub-Saharan Africa, restricting the generalisability of the findings to real-world farming conditions. Addressing this gap, calls for partnerships between researchers, governments, and local farming communities to implement pilot projects that test and refine these technologies in diverse agroecological zones across the region.