CROSS-LAYER RESILIENCE IN UAV-ENABLED DIGITAL AGRICULTURE: A SYSTEMATIC MAPPING REVIEW

Abstract

Unmanned aerial vehicles (UAVs) have become a common part of digital agriculture, yet much of the literature still treats them as aerial cameras at-tached to crop-monitoring workflows. That view is useful, but it misses several conditions that decide whether a field system can be trusted in prac-tice. A farm-deployed UAV workflow has to keep producing useful decisions when illumination changes, wind and dust affect image quality, rural links become unstable, onboard devices have limited compute, and battery capacity constrains both sensing and communication. This review asks: how can UAV-enabled agricultural systems remain reliable when sensing quality, edge intelligence, communication, energy, and decision requirements interact in the field? We address this question through a systematic mapping review of UAV-enabled digital agriculture and closely related edge, communication, and deployment literature. The screening process identified 1,248 records, screened 897 records by title and abstract, assessed 238 full texts, and retained 45 sources that met the final criteria for direct cross-layer synthesis. Each full text was coded using a common scheme covering the primary system layer, computation location, communication dependency, energy treatment, deploy-ment evidence, reported metrics, and decision-support role. In this review, resilience means the ability of a UAV-agriculture workflow to maintain useful sensing, inference, communication, mission execution, and decision delivery under field and system disturbances. The synthesis shows that the literature is strong in sensing and task-level perception, but much less consistent in re-porting latency, energy, connectivity, uncertainty, and farmer-facing decision value together. The paper contributes a seven-layer deployment taxonomy, a reporting-quality appraisal, a failure-mode and mitigation landscape, a com-pact quantitative reporting summary, and a minimal benchmark specification for future cross-layer UAV-agriculture evaluation.