A review of augmented reality in embedded systems for smart farming: innovations, applications and future directions

Abstract

The integration of augmented reality (AR) into embedded agricultural systems is reshaping precision farming by enabling real-time visualizations and interactions with complex environmental data. In the face of mounting global pressures, from climate variability to resource constraints and food system demands, AR-enhanced platforms present a promising pathway toward more efficient farming practices. However, existing research has predominantly treated AR as a standalone tool, overlooking its potential to link and enable the functional integration of diverse embedded technologies. Therefore, the objective of the present review is to investigate how AR visualization can integrate data from Internet of Things devices, unmanned aerial vehicles, farming machinery, robotics, edge computing platforms and artificial intelligence (AI) to enable their coordinated, field-level deployment inprecision agriculture. The article offers three primary contributions: a structured synthesis of AR applications across embedded systems, a conceptual architecture for AR-centered smart farming and an integrated analysis of research gaps and future directions. Key research gaps include the lack of studies addressing model interpretability and system interoperability, insufficient exploration of real-time edge AI processing and gesture-based AR controls, and the absence of globally representative data sets for AI image analysis. Future research directions include the development of low-latency data pipelines, explainable AI interfaces, swarm-capable Drone-AR systems, energy-efficient edge AI models, federated learning for data privacy and participatory design strategies tailored for resource-limited contexts. These findings offer valuable insights for researchers, technology developers, policymakers and farmers working to implement scalable, secure and accessible AR-powered agricultural solutions.