Smart Fence for Real-Time Farm Protection in Sub-Saharan Africa Using Artificial Intelligence: A Systematic Review and Research Roadmap

• Omokhafe J. TOLA

  Author

• James G. AMBAFI

  Author

• Umar S. DAUDA

  Author

Agricultural insecurity from wildlife encroachment and crop raiding threatens food production in Sub-Saharan Africa (SSA), where 60% of smallholder farmers rely on agriculture. Conventional barbed wire and electric fences are passive, static, and poorly suited for rural SSA's resource-constrained environments. This paper presents a systematic review of smart fence technologies for real-time farm protection, adhering to PRISMA 2020 guidelines. From an initial 1,247 documents across IEEE Xplore, ScienceDirect, SpringerLink, and Google Scholar, 56 peer-reviewed studies (2008–2026) were selected for synthesis following title and abstract screening, duplication, and full-text eligibility assessment. The review classifies smart fences into four categories: sensor-based, vision-based, IoT-based, and hybrid AIoT-based. Under controlled conditions, YOLOv8 models outperformed others with over 92% detection accuracy. For edge deployments, MobileNetV3 achieved 85.7% accuracy on commodity hardware, consuming only 3.1 W. Critically, the study found no smart fence deployments in the literature validated specifically for SSA, highlighting a major gap given the region's unique infrastructural challenges. Key contributions of this paper include: (i) synthesizing AI and smart fence performance benchmarks; (ii) establishing a 12-factor matrix quantifying SSA deployment constraints by severity, prevalence, and technology readiness; (iii) proposing an SSA-specific edge-AI reference architecture optimized for solar power, low connectivity, and a sub-$400 deployment cost; and (iv) outlining a structured 3-level research roadmap aligned with TRL 4-9 milestones. This review provides a solid foundation for developing cost-effective, context-aware smart fencing tailored to SSA's agricultural realities.

• 51_Tola et al. (2026)

Copyright (c) 2026 Omokhafe J. TOLA, James G. AMBAFI, Umar S. DAUDA (Author)

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.